THE RELATIONSHIP BETWEEN CERTAIN FAMILY VARIABLES AND THE PSYCHOLOGICAL WELL-BEING OF BLACK ADOLESCENTS CAROLINA MARIA HENN Thesis submitted in accordance with the requirements for the degree of PHILOSOPHIAE DOCTOR in the Faculty of Humanities Department of Psychology at the UNIVERSITY OF THE FREE STATE Bloemfontein 31 May 2005 Promoter: Prof. D.A. Louw 1 STATEMENT I, Carolina Maria Henn, declare that the thesis submitted by me for the Philosophiae Doctor degree at the University of the Free State, is my own independent work and has not previously been submitted by me at another university/faculty. I furthermore cede copyright of the thesis in favour of the University of the Free State. ______________________ _________________ Signed Date 2 READER’S ORIENTATION In accordance with the regulations of the University of the Free State, this thesis is presented in article format. Consequently articles should be viewed as independent, yet related. This may translate into a perception of overlap and repetition between articles. In the scientific community, however, this situation is not unusual, as several articles based on a single study, are usually published independently in different scientific journals. It is recommended that the reader view the five articles independently. The thesis consists of the following five articles: Article I The black family in South Africa: An overview Article II Selected family variables in a South African township: An empirical study Article III The influence of selected family variables on depression in black adolescents Article IV The influence of selected family variables on perceptions of social support among black adolescents Article V The influence of selected family variables on self-esteem and life satisfaction among black adolescents 3 ACKNOWLEDGEMENTS My sincere gratitude and appreciation go to the following people: § Prof. D.A. Louw, an exceptional promoter and academic whom I hold in very high regard § Prof K.G.F. Esterhuyse for statistical analysis of data and patiently answering all my e-mails and telephone calls § The National Research Foundation (NRF) for financial assistance granted to this project § Magriet and Justus for assistance with editing of the manuscript § Esther for endless hours of babysitting § My husband, Carel, for his loyal support and sacrifices, as well as proofreading of the manuscript § My two children, Louis and Catharina, who were always good while I was studying § My parents whose encourage ment, support and prayers never ceased § Everybody who inquired regularly about my progress Ps. 117 “Praise the Lord, all nations! Praise Him, all peoples! His love for us is strong and His faithfulness is eternal.” 4 Any opinions, findings and conclusions or recommendations expressed in this material are those of the author and therefore the National Research Foundation (NRF) does not accept any liability in regard thereto. 5 For my parents, because no -one loves me as much as they do 6 ARTICLE ONE THE BLACK FAMILY IN SOUTH AFRICA: AN OVERVIEW 7 TABLE OF CONTENTS INTRODUCTION.............................................................................. 2 THE TRADITIONAL BLACK FAMILY.......................................... 4 SOCIO-POLITICAL FACTORS THAT IMPACTED ON TRADITIONAL BLACK FAMILIES............................................... 6 Colonialism................................................................................................................7 Industrialization and urbanization.........................................................................8 Westernization ..........................................................................................................8 Apartheid ...................................................................................................................9 CONSEQUENCES OF SOCIO-POLITICAL FACTORS FOR BLACK FAMILIES......................................................................... 11 Marriage and divorce .............................................................................................11 Family structures ....................................................................................................13 The decline in parental authority in black families.............................................15 Poverty and cycles of disadvantage.......................................................................17 Mental health ..........................................................................................................20 STRENGTHS OF BLACK FAMILIES........................................... 21 CONCLUSION................................................................................ 22 LIST OF REFERENCES................................................................. 24 Abstract This article presents an overview of the traditional South African black family. The impact of South Africa’s socio-political history on these families is also explored. The traditional black family was predominantly patrilinear in nature; extended families were the norm; and childbearing was of the utmost importance. Colonialism, industrialization and urbanization caused fathers to leave their families behind in rural areas to find employment in cities. Apartheid marginalized black people in all spheres of living and numerous laws limited their freedom of movement. All these factors contributed to the disintegration of black families. As a result divorce rates increased, parental authority declined and most blacks ended up living in poverty. New family structures, such as single-parent and nuclear structures, are also emerging. Despite all these negative influences, the black family proved its resilience and did not founder completely. A strong sense of community is a particular strength in this regard. However, black families need much support in overcoming the atrocities of the past. INTRODUCTION Several authors point out that the family is one of the main agents of socialization (de Visser & le Roux, 1996; Mboya, 1998; Popenoe, Cunningham & Boult, 1998). The family prepares the individual for contributing to society in a useful, active manner. More specifically, the family is considered to be vital in the raising of children who are well-socialized, mentally healthy and emotionally strong (Andrews & Morrison, 1997; Maforah, 1987; Mboya, 1998; Pretorius, 1996). It is therefore understandable that the family is viewed as one of society’s most important pillars (Burman, 1996), with stable family life enhancing social stability. It is generally accepted that the family is a dynamic, ever-changing institution which reflects societal changes. However, it is equally true that nowadays, more than ever, the modern family is subjected to extreme pressures which have led to a worldwide decline in the quality of family life. The results of this include, amongst others, increasing divorce rates and family violence, as well as a breakdown of parental authority (Steyn, 1996; Thekisho, 1990). In fact, many professionals have come to the conclusion that the family as an institution is in a state of constant decline, and even possible disintegration (cf. Maforah, 1987; Popenoe et al., 1998; Thekisho, 1990). However, authors such as Campbell (1994) do not share this pessimistic 2 viewpoint, although they agree that exacting demands are made on present-day families. The structure of the family has changed and diversified so much over the last few generations (Popenoe et al., 1998) that it is becoming increasingly difficult to define the family. For example, families are no longer only nuclear (mother, father and children), but are comprised of varieties such as female-headed single-parent, male- headed single-parent and female-headed extended families. Furthermore, marriage can no longer be viewed as a minimum requirement for the establishment of a family, as many men and women opt for non-marriage while still forming households which have many of the traditionally accepted characteristics of a family (Preston-Whyte, 1993). To a large extent, the aforementioned changes are even more pronounced in the average South African black family. This is understandable if all the socio-politica l factors that the black family in South Africa has been exposed to, especially during the last century, are taken into account. Unfortunately the same socio-political climate also resulted in an almost total lack of research on black families. Although research dating as far back as the 1930s can be found, it is extremely meagre in extent and scope. Furthermore it is mainly comprised of projects of a descriptive nature; for example the changes brought about by urbanization and industrialization were investigated by some of these projects (Steyn & Viljoen, 1996). In Siqwana -Ndulo’s (1998) view, these projects were merely aimed at comparing the black family with what was considered the “ideal” family, namely the westernized, white family. Thus, the research did not succeed in actually analyzing the black family. The black family was often viewed as primitive and inferior to the western ideal, and was consequently not judged according to its own merit. It is obvious that such research probably did more harm than good. Research aiming at investigating the full impact of South Africa’s socio-political history on the black family is fairly recent (extending more or less over the last decade). This article presents a review of the traditional black family and the factors that have contributed to its current status. 3 THE TRADITIONAL BLACK FAMILY Pretorius (2005) notes that the first signs of human settlement in South Africa date as far back as 25 000 years, and that the first pastoral people inhabited the country from as far back as 2 000 years ago. The first European settlers arrived in 1652. Appolis (1996) points out that definitive socializations existed in South Africa long before the arrival of these settlers. These socialized groups were the San, the Khoikhoi, the Nguni (including groups such as the Xhosa and the Zulu), the Sotho, the Venda and the Tsonga. These groupings had distinct cultures, including their own sets of values and norms, which found expression in religious and social practices. As a detailed discussion of the cultures of each of the mentioned groups falls beyond the scope of this article, only some of the most prominent characteristics of the traditional South African black family will subsequently be highlighted. The family structure of black families was predominantly patrilinear in nature; and polygamy was widely practised and accepted. Extended families were the norm, with three or more generations sharing homes. Hallmarks of the black extended family included an emphasis on loyalty towards the group, large households and mutual support (Steyn, 1993, 1994). To these, Maforah (1987) adds co-operation between relatives, no assistance from the father with household chores, children being reared by many relatives (in fact, all adults in a community could correct a child’s behaviour), and large families. According to Bester (1994), the following values and norms directed behaviour: respect for the elderly; conformity to tradition; strict discipline administered by the father as undisputable head of the family; and social control through, for example, religious practices such as ancestral worship. In traditional black culture, the individual is virtually non-existent (except in corporal terms) (Siqwana-Ndulo, 1998). In fact, individuals define themselves within the context of others. The economy in which the family functioned was that of a self-supporting type of agriculture. Division of labour was clear: men defended the homestead, hunted and cared for the cattle while wome n carried out domestic chores, raised children and cultivated land to produce crops (Maforah, 1987). 4 The institution of African marriage formed the core of a clearly defined yet complex system, which extended to the entire community (Siqwana -Ndulo, 1998). Parents were closely involved in their children’s choice of marriage partners, courtship was non-existent and women tended to marry at a fairly young age (Maforah, 1987). The role of love in choosing a spouse was under-emphasized. Marriage was seen as a transaction between two kinship groups rather than as an agreement between two individuals (Foster, Makufa, Drew & Kralovec, 1997; Kanjo, 1994). Vorster (1993) explains how African law operated. Firstly, the death of a spouse did not necessarily mean that a marriage was dissolved, and referring to a woman as a “widow” was inappropriate. Secondly, a fertile woman whose spouse passed away was allowed to appoint another man (usually from her husband’s family) to take over the role of her husband. Any children born from this relationship were viewed as belonging to the deceased husband’s household. The bride-price, commonly known as “lobola”, was of the utmost importance in traditional African culture (Kanjo, 1994; Maforah, 1987). “Lobola” refers to a “price” that was paid by the groom’s family to the bride’s family. This payment was usually made in the form of a mutually agreed-upon number of cattle (Foster et al., 1997). It is important to note that this transaction was not meant to be a procedure whereby the bride became a mere possession that could be sold. The bride-price, in fact, emphasized the community’s perception of women as valuable and worthy and was a gesture of appreciation from the groom’s family (Thekisho, 1990). In-laws were closely involved in the marriage. Maforah (1987) also comments that extended family members (usually the respective parents of the couple) even intervened in marital conflicts. The most common form of polygamy in African families was polygyny (in terms of which a man may have more than one wife). Polygyny was a way of protecting a man from childlessness. It also ensured that there would be enough hands available to assist with all the work that needed to be done to sustain the family (Maforah, 1987). 5 Divorce was extremely rare and was predominantly instituted by men, who were not eager to divorce their wives in any case, since a reduction in the number of their wives lowered their status (Chiwome, 1994). Under the white National Party government that came into power in 1948, customary marriages were acknowledged. However, if these marriages did not follow the westernized procedures prescribed in the Marriages Act, No. 25 of 1961, they were not officially recognized (Statistics South Africa [STATSSA], 1998). A customary marriage was viewed as an association rather than as an official marriage (Maithufi, 1994). Consequently customary marriages did not enjoy the legal protection granted to civil marriages. Many authors emphasize the importance of childbearing in African culture (cf. Kanjo, 1994; Maforah, 1987; Preston-Whyte, 1993; Vorster, 1993). In fact, having children was so important that a man could marry more than one wife to protect himself from becoming “childless” (Maforah, 1987). The Batswana saying, tseo ke go tsala bana (to marry is to have children), illustrates the importance of children in the black family (Vorster, 1993). Failure to have children was considered a legitimate cause for divorce (Kanjo, 1994) and having as many children as pos sible was highly desirable (Maforah, 1987). The reason for this was mainly twofold. Firstly, as already mentioned, children were needed in order to assist parents in numerous tasks. Secondly, having large families served to ensure that parents would be taken care of when they were elderly. SOCIO-POLITICAL FACTORS THAT IMPACTED ON TRADITIONAL BLACK FAMILIES The history of the black family during the last three centuries or so is inextricably linked to the socio-economical and political history of South Africa. Factors that exerted extreme pressure on the black family were, in particular, colonialism, industrialization and urbanization, as well as the political system of apartheid. 6 Colonialism Instructed by a Dutch company, Jan van Riebeeck arrived in South Africa in 1652 to establish a halfway station to supply food to Dutch ships passing on their way to and from the east. As explained by Appolis (1996), this arrival of Dutch settlers ushered in colonialism in South Africa. The settlers invaded the land and dispossessed the black communities of land that had been theirs for centuries. As more European settlers arrived, more land was needed; and the settlers moved further inland, claiming ground as far as they went. They also excluded the black people from their government, and the indigenous people did not have the necessary means to stand up to these infringements of their rights. They lost not only their freedom, but to a large extent, also their customs and traditions. Black labour was considered cheap and in the eighteenth century, slave labour was viewed as the most effective source of labour (Appolis, 1996). Fragmentation of the black family started with the removal of fathers, and even young children, from their families to work for the colonialists. This permanently changed the division of labour in black families. Wives and children had to take over the tasks typically performed by fathers, while the father’s absolute authority over his family was greatly diminished by his absence (Maforah, 1987). To force black people into a capitalist economy, hut tax was introduced (Viljoen, 1994). This meant that blacks had to pay to live on land that actually belonged to them historically. The discovery of gold and diamonds in 1870 led to the establishment of a mining industry. The latter further increased the need for cheap labour. To force blacks to supply their labour, the hut tax was raised and made payable in coins, and not produce as before (South Africa, c.1900). Black people had no choice other than to leave their homesteads and move to the mining areas where they could sell their labour. Thus, the advent of the mining industry worsened the fragmental effect that the removal of family members had on traditional family life. As pointed out by Appolis (1996), the abolishment of slave labour in 1834 was merely a theoretical exercise. The lives of the black people still did not change since, 7 being dispossessed of land, and therefore also of crops and livestock, they were by then almost completely dependent on the colonialists for their survival. It is therefore clear that colonialism not only introduced an era of submission and domination of black people in South Africa (Viljoen, 1994), but also disrupted their family lives significantly. Industrialization and urbanization Colonialism expanded and led to industrialization. In practical terms, industrialization entailed the erection of factories which, in turn, created a huge demand for labour. This led to many fathers leaving their families in the rural areas and moving to the cities to find employment in factories and mines. Industrialization thus resulted in the urbanization of black people. A migrant labour system came into existence (Pretorius, 2005): men left the rural areas, stayed in single quarters in the cities, worked in industry and visited their families infrequently. In South Africa urbanization was marked by what Popenoe et al. (1998) call circular migration. People, goods and money moved between urban and rural areas and most urbanized blacks maintained their rural linkages. Migrant labour affected black family life in more than one way. For example, husbands and fathers were absent from their families for long periods of time, while, although on a smaller scale, women join ed the exodus to the cities, where they mostly found employment in the domestic sector. Grandparents who remained behind in the rural areas took over absent parents’ childrearing responsibilities. Westernization Westernization was the inevitable result of industrialization and urbanization. Although westernization did not necessarily imply the full incorporation and acceptance of a western lifestyle and value system (Bester, 1994; Carstens, 1995), it did impact significantly on the traditional African culture. According to Thekisho (1990), such acculturation can cause people to suffer from an inability to be 8 themselves. This puts pressure on people that in turn manifests itself in negative forms such as stress, substance abuse and job absenteeism. The kisho (1990) also maintains that people may feel pressurized to keep up a certain standard of living that they cannot afford. The racism which became an integral part of the political system in South Africa, conveyed the message that blacks were inferior to whites and prevented black people from developing positive self-esteem. Many showed a yearning to become like white people (De Haas, 1989) and subsequently adopted Western beliefs and a Western lifestyle which led to the further erosion of the traditional African culture. Apartheid The unequal distribution of power between black and white people and the supremacy of white people had been an integral part of South African society long before the official institution of racial segregation as a policy w ith the coming into power of the Nationalist Party in 1948. This policy of apartheid (“separateness”) introduced numerous laws which severely impacted on the lives and families of black people. Appolis (1996) points out the most prominent of these laws: • Influx control policies, which refer to both legislation and its subsequent administrative measures, were aimed at restricting the movements of blacks in South Africa. Only a person with a valid pass, which contained information such as the person’s identity number and place of legal residence, was allowed to stay in the designated residential area indicated on the pass. This directly resulted in fragmentation of black families. For instance, a father was allowed to stay in the city, but not his wife and children. Although black people were allowed to work in urban areas, they were not allowed to obtain permanent residence there. • Security laws were implemented when black people (especially the youth) became increasingly politicised and involved in attempts to overthrow the apartheid government. Any attempt to bring about political, social or economic changes was declared unlawful and blacks were forbidden to take part in so-called “illegal gatherings”, which could have been any gathering of more than two persons. 9 • Taxation laws determined, inter alia , that black people had to pay specific taxes directed at them only, namely tribal levies, education tax and local tax. • Laws aimed at the segregation of residential areas earmarked specific areas for black and white people respectively. Black families who were, at the time, living in so-called white areas had to relocate voluntarily, or were forcibly removed. 10 CONSEQUENCES OF SOCIO-POLITICAL FACTORS FOR BLACK FAMILIES From the foregoing, it is clear that South Africa’s socio-political history has impacted severely on black families. Marriage and divorce In the traditional black culture the emphasis was never on the individual, but on the group. Western culture brought with it an emphasis on the individual. This included values such as romantic love and companionship, which took on more prominence in the choice of a marriage partner (Rautenbach & Kellerman, 1990). Traditions such as polygamy declined as a result of the Christian influence that propagated monogamy. In the latter, individual satisfaction was emphasized over the traditional African values such as procreation (Chiwome, 1994). Next to Christianity, economic considerations also caused a decline in polygamy (Maforah, 1987; Pretorius, 2005), which has become very expensive in modern-day South Africa. Furthermore, polygamy clashes with the recent emphasis on women’s rights; and women are especially intolerant of polygamy nowadays (Viljoen, 1994). The institution of bridewealth (lobola ) is also declining; and in instances where it is still practised, cash has replaced cattle (Kanjo, 1994). Unfortunately, in some instances, it has lost its original significance as a token of gratitude and has become an opportunity to negotiate for financial gain (Thekisho, 1990). In such cases women are reduced to mere “sales goods”. Many modern black women also perceive bridewealth as reducing their freedom, since their husbands may consider them to be property that they have bought (Kanjo, 1994). Preston-Whyte (1993) notes that there has been a steady increase in the number of non-married black women in South Africa and provide the following reasons, among others: delays in payment of bridewealth that can lead to indefinite postponement of marriage; lack of sanctions against premarital pregnancies; and increased independence from men. She further comments that the positive value attached to 11 fertility means that non-marriage will not necessarily prevent black women from having children. The breaking down of the traditional black marriage is reflected in the increased divorce rates: the specific divorce rate (number of divorces per 1000 married couples) rose from 1,37 in 1994 to 1,75 in 1998 (STATSSA, 1998). Unfortunately, accurate divorce statistics are very difficult to obtain, for the following two reasons. Firstly, customary marriages that end in divorce are not always known of and are therefore not incorporated in statistics. Secondly, many black marriages culminate in desertion without a legal divorce ever being obtained. Given the above, the divorce rate for black marriages is probably significantly higher than statistics indicate. According to Burman (1991), more than fifty percent of black marriages end in de facto divorce (non-formalized dissolutions of marriage). Steyn (1993, 1994) reports similar findings. There are a number of reasons for the high prevalence of divorce among black South Africans. As explained by de Haas (1989), the migrant labour system, in particular, put a great deal of strain upon their marriages. Male migrant labourers were mostly housed in single-sex hostels whereas the majority of women stayed in domestic servants’ quarters in private households. Husbands and wives were not allowed to live together. Wives could visit their husbands in the hostels, but prior reservation of an appointment was required and facilities lacked privacy in the true sense of the word. Many men resorted to having girlfriends who took care of their needs in the absence of their wives. The wive s in the rural areas were totally dependent on the allowances sent by their husbands and there was an ever -present fear that their husbands might neglect or desert them. In this “system which facilitates irresponsibility” (De Haas, 1989, p. 7), many men succumbed to the temptations of available women and never returned to their wives in the rural areas. Another explanation for the increase in divorce is found in the emancipation of black women. Legislation that was adopted in recent years recognized the equality of women and gave them the bargaining power that they lacked traditionally. Under the new Constitution, either partner can choose to end a marriage (Pretorius, 2005). Furthermore, customary marriages are now granted full recognition and are also 12 officially registered (STATSSA, 1998). Women who have been married under customary law are also entitled to a share of the property in the case of a divorce. Improvements in their educational qualifications afford black women the opportunity to obtain well-paying employment. This decreases their dependence on their husbands, thus making it easier for them to obtain divorces. It is therefore understandable that modern black women challenge the traditional ideas of male dominance and sole decision-making by men. A study conducted by Maforah (1993) among professional black women revealed that many of these women had negative attitudes towards marriage and the traditional role played by women in marriages. Maforah’s findings were confirmed by De Haas (1989). Family structures As mentioned previously, the extended family comprised the typical family structure in traditional black families. In Western societies, on the other hand, the nuclear family was perceived as the ideal family structure. In a research project conducted by Steyn (1993, 1994) it was found that 36,9% of black families were nuclear, 28,8% were extended and 14% were single-parent families. At first glance, it appears as if the nuclear family structure has the highest frequency among black families. However, Russell (1994) criticized Steyn’s research, claiming that Steyn had shown insensitivity towards kinship systems different from her own. A major criticism was that the category, “nuclear family with additional members,” identified by Steyn, is really an extended family. A further criticism was that Steyn omitted to take into account that any given information regarding a multigenerational family is only a reflection of the position of the family at that precise time. Siqwana-Ndulo (1998) agrees with Russell in this regard. For example, a family that appears to be female-headed during the course of the research may become male-headed again when a grown son succeeds his mother as head of the household. When Russell’s concerns are ta ken into account, the percentage of extended black families rises to 59%, with 33% of the families being nuclear. 13 In a study conducted among black families in rural areas, van Vuuren (1997) found that the majority (34,4%) of families in the sample were extended. She created a separate category for families who were originally nuclear, but became extended by rendering support to kinship members. The question that arises is whether these families are not really extended families, regardless of how they started out. If this is assumed to be the case, then extended families account for 58,8% of her sample. Both sets of research findings are confirmed by Rautenbach and Kellerman (1990) who, in their analysis of market research data, found that the majority of urban black families were extended (62,3%). In research conducted by Siqwana-Ndulo (1998) amongst rural African families in the Eastern Cape province, it was found that 34,3% of families in the sample were of the husband-and-wife type. However, only 15% of these couples resided with only their children. According to Siqwane-Ndulo, even the latter 15% of the families are probably not nuclear, but merely appeared to be so at the time of the interviews. The results of this study, then, also indicate that the majority of black people still form part of extended families. In support of the above findings, de Visser and le Roux (1996) determined that the majority of respondents in their study were part of extended families. It is clear that the extended family is still the dominant family structure among black families, although there is sufficient evidence that a variety of other family structures are emerging (Ziehl, 2002). The research of Steyn, Van Vuuren, Siqwana-Ndulo, and Rautenbach and Kellerman reflects the emergence of family types such as female- headed, male-headed and single-parent families, to name but a few. Boult and Cunningham (1992) view the female-headed, single -parent family as the fastest- growing family structure in society. Siqwana-Ndulo’s (1998) finding that 62,5% of households in her survey were female -headed, appears to support this viewpoint. Factors that led to the formation of new family types among blacks include the adoption of Western values and norms, which led to people opting for nuclear families. In de Visser and le Roux’s (1996) study, conducted among teenage mothers, the majority of respondents indicated that they would prefer to live in nuclear households, citing reasons such as grandparents’ interference. On the other hand, 14 many blacks would prefer to live in extended families, with the mutual support offered by such families, but the demands of modern society do not allow this. Economic constraints prevent people from providing and receiving support in the way that this was done in the traditional black family (Van Vuuren, 1997). The extended family has shown its resilience (cf. Møller, 1995), but it is not immune to the influences of modern society. On the other hand, it is also true that black families adapt to new conditions in accordance with their traditional values; in other words, the latter have not been eroded (Siqwana -Ndulo, 1998). For example, when migrant labour removes fathers from their homes, black families are often able to compensate by having an uncle take over the father’s duties. They are therefore not completely at the mercy of changes in society. However, in the final analysis, it seems to be a given fact that the structure of the African family is changing. The decline in parental authority in black families In the traditional black family, parental authority and the dominance of the father figure were non-negotiable. However, there seems to be consensus among most authors that this picture is changing and that there is a definite decline in parental authority (cf. Campbell, 1994; Thekisho, 1990; Viljoen, 1994). There are several possible explanations for this trend. After the uprisings in the seventies, the black youth played a crucial role in the fight against apartheid. The older generations disapproved of the youngsters’ challenging of authority. The youth, in turn, blamed the older generations for failing to fight apartheid (Viljoen, 1994). Phewa (1992) mentioned, for example, that teenagers complained about their parents being uninvolved in political affairs. These differences created a communication gap between the generations, with the young people showing less respect for the older people’s authority. Children are in a much better position than their parents were to obtain, and further, their education. This has also widened the gap between them. In schools, children are exposed to new ideas and values and this results in parents and children who have very little in common. The youth view their parents as knowing nothing and 15 consequently do not take their parents’ authority seriously (Bester, 1994). Portes and Zady’s (2002) observation with regard to immigrants to America also holds true for black youth in South Africa: children may acquire English faster than their parents do. This leads to a role reversal of sorts, in terms of which children become parents and vice versa. Children may also find it easer to reconcile traditional culture and western influences. Parents, however, place strong emphasis on the importance of education for their children; and they realize that nowadays, they are no longer fully equipped to advise their children on all issues (cf. Gilbert, van Vlaenderen & Nkwinti, 1995). Viljoen (1994) explains how poverty and poor housing led to a decline in parental authority. More specifically, poverty forced parents to find employment elsewhere, thereby leaving children unattended. Structural factors, such as the long distances that parents had to travel between their homes and places of employment, left them with even less time to spend with their children. A shortage in housing often led to situations where parents and children had to share bedrooms. There was no privacy for the married couple, and children were exposed to private discussions and sexual intimacy (Phewa, 1992). Social changes have also contributed to a decline in parental authority. The transition from a traditional lifestyle to a Western urbanized lifestyle was probably a major social factor that affected parental authority in black families. According to Maforah (1987), the influence of agents of socialization, such as the media and the school (which exposed children to Western values), caused children to become less submissive. Phewa (1992) comments that mass media communication replaced parent-child communication and often conveyed questionable values that opposed parental values. In addition, the increase in female -headed single-parent families, in particular, contributes to the decline in parental authority, since children are more prone to challenge their mother’s authority. Long absences of fathers, caused by migratory labour, resulted in children openly flouting their mother’s authority (Pretorius, 2005). Phewa (1992) accuses black fathers of withdrawing from their pa renting role, thereby leaving the full burden of raising children to the mothers. Without the presence of a 16 strong father figure, the mothers themselves were at a loss in terms of the “what” and “how” of conveying values to their children, and therefore often left the teaching of values to the teachers at school. Children also do not show the desired respect towards the elderly, as the latter have not been involved in the struggle (Phewa, 1992). In her study among township youth, Campbell (1994) found that the young people experienced strong bonds with their parents, and felt an obligation to take care of them when they were elderly. However, the same youth also viewed other groups (such as their politically orientated comrades, as well as educated, prof essional people) as offering them greater chances of success than their families. Parents were often under-educated, rural and politically inactive, which made association with them less valuable. The black family will probably never return to its traditional form in terms of absolute parental authority. The older generations mourn the loss of values related to parental authority (Viljoen, 1994). On the other hand, many parents acknowledge that times have changed and that they are limited in terms of what they can provide for their children. They are also aware, however, that they can still direct their children towards what is generally perceived as correct and desirable in terms of behaviour towards others (cf. Gilbert et al., 1995). Poverty and cycles of disadvantage The origin of poverty among black South Africans is rooted in the country’s socio - political history. The marginalization of black people in all spheres of life, and their exclusion from governance, prevented them from maintaining even a marginally decent standard of living. In Appolis’s (1996) view, poverty among black South Africans was not of their own making, but was structurally imposed on them. Deliberate measures by the apartheid government caused them to remain poor. For example, job reservation for whites limited blacks to the less highly -paid jobs. Blacks were also forced to settle in allocated areas on the outskirts of towns and cities, which greatly increased travelling costs. 17 In their report on poverty, Hirschowitz, Orkin and Alberts (2000) used data on monthly household expenditure obtained during the 1996 Census, in order to determine poverty levels in South Africa. Their findings indicated that more than 50% of Africans had a monthly expenditure indicative of poverty. Data compiled from STATSSA, amongst other sources, showed a poverty rate of 56,9% for South Africa in 1996. In addition, 1999 calculations indicated that about seven out of every ten blacks are considered poor, compared to one out of every ten whites (Pelser, 2002). Unemployment is an obvious indicator of poverty; and the Labour Force Survey conducted in September 2001 (STATSSA, 2003) re vealed an unemployment rate of 35,9% for black people. Black men showed an unemployment rate of 32,2% while black women had the highest unemployment rate of all groups, namely 39,8%. These figures seem particularly high when compared to the same survey’s indicated unemployment rate of 29,5% for the country as a whole. In terms of other indicators of poverty, such as access to formal dwellings, access to electricity, running water inside dwellings and chemical or flush toilets, vast improvements have been made since the African National Congress (ANC) came into power in 1994. For example, in 1996 (Hirschowitz et al., 2000) almost half of all blacks lived in traditional or informal dwellings. By 2001 (STATSSA, 2003) this percentage had shrunk to approximately 40%. These improvements are promising, but when the situation of blacks is compared to that of the other population groups in South Africa, they are clearly still behind. For example, STATSSA (2003) reports that by 2001, 80% of African households had access to piped water, either in the dwelling, on site or from a communal tap, compared to almost 100% of households in the coloured, Asian/Indian and white populations. The effects of poverty stretch far and wide, and unfortunately the consequences are also the factors that maintain the poverty, thus perpetuating the so-called cycle of disadvantage. These consequences include overcrowding, poor health and poor education. The lack of proper housing for blacks led to the establishment of overcrowded squatter camps. Concomitant with overcrowding are a lack of proper sanitation, multiple sharing of washing and toilet facilities (if available), ineffective methods of 18 food storage and a lack of proper ventilation (Mabetoa, 1994). All these factors have hazardous effects on people’s health. Water - and sanitation-related diseases in South Africa, such as gastrointestinal illnesses, are more prevalent in squatter camps than in other residential areas. Poverty also leads to malnutrition, which in turn lowers resistance to infections. One of the most common causes of death owing to disease in developing countries, such as South Africa, is respiratory infections (C. Swanepoel, personal communication, 30 November 2001). Pelser (2002) further emphasizes this relationship between poverty and disease by pointing out that the lack of basic services such as refuse removal and proper sanitation, is directly related to the high prevalence rates of diseases such as diarrhoea and tuberculosis. Surveys conducted in recent years have shown that some of the highest HIV infection rates often occur in socially unstable informal settlements (Pelser, 2002). This puts people in squatter camps at a higher risk of contracting HIV/Aids. Shisana (2002) found that the national prevalence rate of HIV in South Africa is 11,4%. Furthermore, her research also established that the HIV prevalence rate for African adults (15-49 years) is 18,4%. This is considerably higher than the national prevalence rate. Compounding this problem is the fact that the prevalence of HIV-related tuberculosis is increasing (Wilkinson & Davies, 1997). In a compilation of various sources, Pelser (2002) projects an infant mortality rate of more than 60 by 2008-2010. This high rate is attributable to the HIV epidemic. His projection of children orphaned by aids (under 15 years old), postulates a figure of one million by the year 2005. Pelser (2004) indicates the factors that increase the vulnerability of the poor to illnesses such as tuberculosis. Firstly, poverty leads to a lack of resources such as piped water, proper sanitation and electricity. This, in turn, leads to an increased prevalence of preventable diseases such as respiratory illnesses and tuberculosis. Secondly, the poor are often undernourished, which lowers their resistance and renders them more susceptible to these diseases. Compounding the problem is the fact that poor people often lack proper education in terms of, for example, hygiene. Good education puts people in a position to earn a higher income, which can provide them with an escape route from poverty. Unfortunately, education for black people 19 left much to be desired during the National Party government’s reign. Black schools were not provided with proper facilities and resources, and black youth who became politicized neglected their own schooling and disrupted that of others. High levels of illiteracy among the older generations (cf. Ramarumo, 1994) and long working hours meant that parents were not in a position to assist with and supervise children’s schoolwork (Mabetoa, 1994). Owing to poverty, parents could not afford to pay for better schooling, as provided by private schools, for example. Tertiary education was a luxury that only a small minority could afford. Census 2001 (STATSSA, 2003) showed that 22% of Africans aged twenty years and above had received no education whatsoever. About 17% had completed their high school career and only about 5% had obtained some form of post-school education. Poor education, together with factors such as unemployment, low levels of socio - economic development and illiteracy, is related to higher fertility rates and lower contraceptive use (Pelser, 2002). African teenage birth rates increased by 32,7% between 1991 and 1998 (Pelser, 2002). Teenage parents living in poor communities are not equipped either financially or emotionally to raise children, which increases the burden of poverty. The high value placed on children in African culture further increases the risk of teenage pregnancy. Corcoran and Franklin (2002) found that pregnancy and parenting are sources of self-esteem among poor teenagers. Apart from the increase in teenage pregnancies, these young people are also at a much higher risk of becoming HIV-infected. Mental health It cannot be denied that the psychological well-being of blacks has been significantly affected by South Africa’s socio-political history. Hickson and Kriegler (1991) indicate two important psychological effects. Firstly, a racist society led to poor self- esteem and lack of confidence among blacks. Instead of being a source of pride, their race became something to be ashamed of. Secondly, exposure to violence during the apartheid struggle led to depression and anxiety, including emotional numbing and a sense of hopelessness and meaninglessness. As in the case of most other resources, mental health care was grossly unequally distributed under the National Party 20 government. Such services, although now more accessible, are still not adequate. For example, most mental health centres are located in urban areas; which means that people in rural areas cannot access these services, even though they are much needed. It also seems that more acute basic needs for amenities such as housing are given preference above the “less observable” needs such as mental health services. STRENGTHS OF BLACK FAMILIES When considering all the factors that have impacted negatively on the black family in South Africa, one might expect that the black family structure would have completely disintegrated by now. However, black families have shown themselves to be resilient, and in the face of adversity, they have managed to maintain themselves, at least to some extent. This is probably owing to the fact that the apparent weaknesses and problems in black families were never inherent to them, but were imposed on them by socio-political and economic factors (Billingsley, 1992). A major strength of black people is their keen emphasis on kinship and community support. The burden of children, and the benefits of income and access to housing, are not only shared by kinship members (Burman, 1996), but also among community members. In Steyn’s (1993, 1994) research it was found that many families included additional members, who were not necessarily kin. The African family is highly dynamic and fluid, which means that there is much activity in terms of the coming and going of household members. This implies that the African family is very flexible and able to adapt to different circumstances, which is clearly a strength. Black people attach a very high value to “ubuntu”, which means that community members support one another and accept responsibility for one another (for example, unemployed people often provide child -care for those who are employed) (cf. Burman, 1991; Van Vuuren, 1997). Viljoen (1994) cites the following strengths in black family life: a high degree of commitment to the family, the importance attached to children and the family, and kinship networks. 21 CONCLUSION South Africa celebrated a decade of democracy during 2004. The country now affords equal opportunities to all her people; yet the long-term effects of apartheid, prejudice and discrimination cannot simply be swept under the carpet (Burman, 1996). Although many advances have been made in terms of basic human rights, such as freedom of movement, the right to vote and access to basic facilities and amenities, these improvements are still not enough. When compared to other population groups, it is clear that blacks are still lagging behind in almost all spheres of living. Hundreds of years of marginalization led to numerous problems such as poverty, high physical- and mental- health risks and a lack of proper education. Discriminatory legislation and the concomitant repression of black people stripped them of their human dignity and led to the disintegration of black families (Sonn & Louw, 1989). In assisting black families to overcome the atrocities of the past, their inherent strengths (such as their strong sense of community) should be utilized and promoted. There are many government projects, as well as projects run by non-governmental organizations, that are aimed at improving the quality of life of blacks, but understandably these projects take time to have an effect. Different organizations should pool their resources and co-operate, as this will ensure that the measures taken are more inte nsive and thus more effective. As education is often considered to be the best, and sometimes even the only, way out of poverty (cf. Nkabinde & Ngwenya, 1996), it should receive more attention (and especially more funding) than it currently does. This holds true for both formal and informal education. Entrepeneurial-skills training, for example, will enable people to create work for themselves. People also need education in terms of proper hygiene, contraceptive use, child-rearing practices, and prevention of infectious illnesses. Black people should develop a sense of pride in who they are, and in their environment; but to reach this point, the psychologically harmful effects of apartheid and poverty (e.g. low self-esteem) need to be addressed. Psychotherapy and 22 counselling in terms of aspects such as, for example, interpersonal problem-solving, life-skills training, accepting responsibility for personal actions, and dealing with shame, guilt and loss of self-esteem should be made easily accessible (cf. Hickson & Kriegler, 1991). In the process, the youth should not be neglected, as they are tomorrow’s nation. Addressing inequalities, however, does not only mean rectifying the past, but also the present. Several persisting disparities that form part of a centuries-old culture should also be addressed. At the forefront in this regard is gender discrimination, which closes many doors for females. By addressing this prejudice the whole nation, and not only one sector, could be bolstered and fortified. A needs analysis at grassroots level will assist in determining the psychological, physical and social needs of black families. Incorporating these findings when designing intervention strategies, will enhance the suitability and effectiveness thereof. Black families have been able to adapt to dire circumstances. Despite the burden and hardship foisted upon them by apartheid and its disastrous consequences, they have not only shown the necessary resilience to survive, but have also managed to a large extent to retain their innate philosophy and values. If they could continue building on their sources of strength, there is no doubt that they could not only outgrow and overcome their past, but could also approach the future with confidence. 23 LIST OF REFERENCES Andrews, A.B. & Morrison, B. (1997). The changing American family: Impact on South Carolina’s children. Columbia: Institute for Families in Society. Appolis, K.U.C. (1996). From fragmentation to wholeness: The black South African family under siege. New York: University Press of America. Bester, C.W. (1994). The consequences of urbanization and westernization on black families in South Africa. Social Work, 30(2), 101-106. Billingsley, A. (1992). Climbing Jacob’s ladder: The enduring legacy of African- American families. New York: Simon & Schuster. Boult, B.E. & Cunningham, P.W. (1992). Black teenage pregnancy: An African perspective. International Journal of Adolescence and Youth, 3, 303-309. Burman, S. (1991). Capitalising on African strengths: Women, welfare and the law. South African Journal on Human Rights, 7 (2), 215-221. Burman, S. (1996). Intergenerational family care: Legacy of the past, implications for the future. Journal of Southern African Studies, 22(4), 585-598. Campbell, C. (1994). Township families and youth identity: The family’s influence on the social identity of township youth in a rapidly changing South Africa. Pretoria: Human Sciences Research Council. Carstens, C. (1995). Die invloed van televisie op die verwesteringsproses by die swart adolessent. Ongepubliseerde doktorale proefskrif. Pretoria: Universiteit van Suid-Afrika. 24 Chiwome, E. (1994). The role of the church in the shift of the marriage and the family institutions from tradition to modernity: The case of Zimbabwe. In T. Sono (Ed.), African family and marriage under stress. Pretoria: Centre for Development Analysis. Corcoran, J. & Franklin, C. (2002). Multi-systemic risk factors predicting depression, self-esteem and stress in low SES and culturally diverse adolescents. Journal of Human Behavior in the Social Environment, 5(2), 61-76. De Haas, M. (1989). The socio-political roots of the breakdown of marriage amongst African South Africans. The Social Work Practitioner/Researcher, 2(4), 3-11. De Visser, J. & le Roux, T. (1996). The experience of teenage pregnancy in Knoppieslaagte. South African Journal of Sociology, 27(3), 98-105. Foster, G., Makufa, C., Drew, R. & Kralovec, E. (1997). Factors leading to the establishment of child-headed households: The case of Zimbabwe [Electronic version]. Health Transition Review, 7, 155-168. Gilbert, A., van Vlaenderen, H. & Nkwinti, G. (1995). Planting pumpkins: Socialization and the role of local knowledge in rural South Africa. South African Journal of Psychology, 25(4), 229-235. Hickson, J. & Kriegler, S. (1991). Childshock: The effects of apartheid on the mental health of South Africa’s children. International Journal for the Advancement of Counseling, 14(2), 141-154. Hirschowitz, R., Orkin, M. & Alberts, P. (2000). Key baseline statistics for poverty measurement. In P. Alberts et al., Measuring poverty in South Africa . Pretoria: Human Sciences Research Council. Kanjo, G. (1994). Western influences on African marriage in Malawi. In T. Sono (Ed.), African family and marriage under stress. Pretoria: Centre for Development Analysis. 25 Mabetoa, M. (1994). Cycles of disadvantage of African families in South Africa. In T. Sono (Ed.), African family and marriage under stress. Pretoria: Centre for Development Analysis. Maforah, N.F. (1987). The changing role of the black family and its adaptation to the present needs. Social Work, 23, 262-268. Maforah, N. (1993). Black, married, professional and a woman: Role conflict? Agenda, 18, 5-7. Maithufi, I.P. (1994). Marriage and succession in South Africa, Bophuthatswana and Transkei: A legal pot-pourri. TSAR, 2, 262-277. Mboya, M.M. (1998). Family relations and the self-concept of African adolescents: Gender-related differences. Journal of Comparative Family Studies, 29 (1), 201-213. Møller, V. (1995). Home environment and educational achievement among high- school pupils living in three-generation urban black households. South African Journal of Sociology, 26(3), 87-97. Nkabinde, Z.P. & Ngwenya, T.P. (1996). Interviews with African parents of children with disabilities in South Africa. Paidonomia, 19(1), 34-42. Pelser, A. (2002). The social landscape of sexual and reproductive health in South Africa with particular reference to HIV/AIDS. In D. van Rensburg et al., Strengthening local government and civic responses to the HIV/AIDS epidemic in South Africa. Bloemfontein: Centre for Health Systems Research and Development. Pelser, A.J. (2004). Health, environment and development in South Africa. In H.C.J. van Rensburg (Ed.), Health and Health Care in South Africa . Pretoria: Van Schaik Publishers. 26 Phewa, G.V. (1992). Problems experienced by black parents in relationship with their adolescent children: A psychopedagogic perspective. Unpublished master’s dissertation. Durban: University of Zululand. Popenoe, D., Cunningham, P. & Boult, B. (1998). Sociology: First South African edition. South Africa: Prentice Hall. Portes, P.R. & Zady, M.F. (2002). Self-esteem in the adaptation of Spanish-speaking adolescents: The role of immigration, family conflict, and depression. Hispanic Journal of Behavioral Sciences, 24(3), 296-318. Preston-Whyte, E. (1993). Women who are not married: Fertility, “illegitimacy”, and the nature of households and domestic groups among single African women in Durban. South African Journal of Sociology, 24(3), 63-71. Pretorius, E. (2005). Family life in South Africa. In J.L. Roopnarine & U.P. Gielen (Eds), Families in global perspective. Boston: Pearson Education. Pretorius, T.B. (1996). The family environment of students self-appraised as effective and ineffective problem solvers. Psychological Reports, 79, 915-921. Ramarumo, M.M. (1994, October). Family literacy. Paper presented at the World Symposium on Family Literacy, Paris, France. Rautenbach, E.A. & Kellerman, A.P.R. (1990). Die swart middelklasgesin in die stad. Maatskaplike Werk, 26, 228-240. Russell, M. (1994). Do blacks live in nuclear family households? An appraisal of Steyn’s work on urban family structures in South Africa. South African Sociological Review, 6 (2), 56-67. Shisana, O. (2002). The current situation of HIV/AIDS in South Africa. Retrieved August 28, 2004, from http://www.hsrcpublishers.ac.za 27 Siqwana-Ndulo, N. (1998). Rural African family structure in the Eastern Cape province, South Africa. Journal of Comparative Family Studies, 29 (2), 407- 417. Sonn, F.C.T. & Louw, D.A. (1989). Sosiokulturele perspektiewe van abnormale gedrag. In D.A. Louw (Red.), Suid -Afrikaanse Handboek van Abnormale Gedrag. Kaapstad: Southern Boekuitgewers. South Africa c. 1900 . (n.d.). Retrieved December 6, 2001, from http://library.thinkquest.org/27629/themes/society/sa000.html Statistics South Africa (1998). Marriages and divorces. Pretoria: Author. Statistics South Africa (2003). Census 2001: Census in brief (2nd ed) [Electronic version]. Pretoria: Author. Steyn, A.F. (1993, 1994). Family structures in the RSA. Pretoria: Human Sciences Research Council. Steyn, A.F. (1996). Values that support quality marital and family life. South African Journal of Sociology, 27, 143-147. Steyn, A.F. & Viljoen, S. (1996). The state of family theory and research in South Africa. In M.B. Sussman & R.S. Hanks (Eds), Intercultural variation in family research and theory: Implications for cross-national studies. New York: Haworth Press. Thekisho, G.N. (1990). Family disorganization among blacks. Social Work Practice, 1 (90), 6-7. Van Vuuren, A. (1997). The strengths and problems of black families in informal settlements. Pretoria: Human Sciences Research Council. 28 Viljoen, S. (1994). Strengths and weaknesses in the family life of black South Africans. Pretoria: Human Sciences Research Council. Vorster, L.P. (1993). Substitusie in die inheemse reg. TSAR, 4, 552-560. Wilkinson, D. & Davies, G.R. (1997). The increasing burden of tuberculosis in rural South Africa – impact of the HIV epidemic [Electronic version]. South African Medical Journal, 87(4), 447-450. Ziehl, S. (2002). Black South Africans do live in nuclear family households – a response to Russell. Society in Transition, 33 (1), 26-49. 29 ARTICLE TWO SELECTED FAMILY VARIABLES IN A SOUTH AFRICAN TOWNSHIP: AN EMPIRICAL STUDY 30 TABLE OF CONTENTS INTRODUCTION............................................................................ 33 LITERATURE REVIEW ................................................................ 34 The traditional black family..................................................................................34 Socio -political factors that impacted on the traditional black family................35 The effect of socio -political factors on the family life of blacks..........................37 Strengths of black families.....................................................................................40 METHODOLOGY .......................................................................... 40 RESULTS AND DISCUSSION........................................................ 41 CONCLUSION................................................................................ 51 LIST OF REFERENCES................................................................. 54 LIST OF TABLES Page Table 1 Educational status of parents of black adolescents 41 in Mangaung Table 2 Employment status of parents of black adolescents 43 in Mangaung Table 3 Family structures of black families in Mangaung 44 Table 4 Marital status of parents of black adolescents 45 in Mangaung Table 5 Household size of black families in Mangaung 46 Table 6 Time spent with parents per week 47 Table 7 Reasons why adolescents do not spend enough 49 time with parents Table 8 Households taking care of members with 50 severe disabilities or serious chronic illnesses Abstract As there is a scarcity of local information regarding black South African families in Mangaung, a township in the Free State province, a study was carried out to obtain data regarding several family variables among these families. Data were collected from 2 505 high-school students by means of a biographical questionnaire. Results revealed that parents had attained reasonably high levels of education. The extended family was found to be the predominant family structure, although only marginally so. Divorce rates are low, suggesting a commitment to marriage. The average household size is large in relation to national trends in the country. Furthermore, adolescents spend considerably more time with their mothers than with their fathers. A majority of adolescents do not spend sufficient time with their fathers. The most common reason cited for not spending sufficient time with both parents, was parents’ work. A large proportion of participants indicated that they were taking care of household members with severe disabilities or serious chronic illnesses. Results reflect changes in black families, but also reveal the strengths and resilience of these families. Future research should expand on the current findings and should look into the influence of phenomena such as the HIV/AIDS epidemic and the increase in single-parent families. INTRODUCTION The family is viewed as one of society’s most important pillars (Burman, 1996). More specifically, functional families provide children with love, stability and guidance that enable them to make responsible choices and to eventually contribute to society in a useful manner. There is no doubt, however, that more demands are being placed on families today than ever before. Factors such as marital status, family violence, and financial constraints put enormous strain on families. The level of functionality, or dysfunctionality, in a family will affect the psychological well-being of its children. Any exploration of families and their influence on children should therefore contain information regarding aspects such as family structure and size, as well as financial stability. Very little information exists regarding black families in South Africa. Earlier research, dating as far back as the 1930s, was mostly descriptive in nature (Steyn & Viljoen, 1996). Also, these studies mostly judged African families in terms of what was considered to be the “ideal” western family, and not on the basis of its own merits. Black families were presented as uncivilized, primitive and in need of 33 correction. Recently much more attention has been given to black families in research, yet in many areas even the most basic information on black families does not exist. An example of this lack of information is found in Mangaung, a township in the Free State province. This is the main reason why it was decided to investigate the family structure, and related aspects, of black families residing in Mangaung. This article aims to present and discuss the findings obtained regarding issues such as family structure, employment and educational status of parents, the marital status of parents and the amount of time spent with parents. Owing to the nature of the data that will be presented, the article will have a strong socio-psychological character. LITERATURE REVIEW In order to provide a background against which the empirical findings should be interpreted, a brief overview of the traditional black family is presented hereunder. The traditional black family According to Appolis (1996), definitive socializations existed among black people in South Africa long before European settlers arrived in 1652. Although different groupings existed, each with its own set of values and norms, the black families in all these groupings shared certain characteristics. A summary of the most prominent characteristics of traditional black families follows. Extended, multigenerational families were the typical family structure. Traditionally, families were patrilinear and polygamy was widely practised and accepted. Emphasis was placed on mutual support and loyalty to the group (Steyn, 1993, 1994). According to Bester (1994), the most important values and norms included respect for the elderly, conformity to tradition and strict discipline from the father as the undisputable head of the family. The economy was typically comprised of a self-supporting type of agriculture. Women were responsible for household chores, raising children and tending crops, 34 whereas men hunted, cared for cattle and provided protection for the family and homestead (Maforah, 1987). Parents played a determining role in their children’s choice of marriage partners. Marriage was seen as an agreement between kinship groups rather than between individuals (Kanjo, 1994; Vorster, 1993). Bridewealth, commonly known as lobola , was a “price” that the groom’s family had to pay to the bride’s family. Payment was usually made in the form of a mutually agreed-upon number of cattle (Foster, Makufa, Drew & Kralovec, 1997). In-laws were closely involved with their married children, and even intervened in marital conflicts. Polygyny was widely practised. It served two functions: firstly it protected men from childlessness and secondly, it ensured that there would be enough hands to assist with tasks (Maforah, 1987). Childbearing was a very important aspect of African culture (Kanjo, 1994; Ma forah, 1987; Preston-Whyte, 1993; Vorster, 1993). Failure to have children was considered a legitimate cause for divorce (Kanjo, 1994), and having as many children as possible was highly desirable (Maforah, 1987). Socio-political factors that impacte d on the traditional black family The history of the black family during the last three centuries or so is inextricably linked with the socio-economic and political history of South Africa. Three dominant factors that exerted particular pressure on black families were colonialism; industrialization and urbanization; and the political system of apartheid. The arrival of Dutch settlers in South Africa in 1652 ushered in colonialism in South Africa. Appolis (1996) explains how colonialism impacted on black people. Settlers dispossessed the natives of land that had been theirs for centuries. As more and more settlers arrived, they moved further inland, claiming ground as far as they went. Black people were excluded from government, and they did not have the skill or resources to defend themselves against these infringements of their rights. As black labour was considered cheap, fathers and even young children were removed from their families to work for the colonialists. In an attempt to obtain more blac k labour 35 for the mining industry, the level of taxation was raised and taxes were made payable in coins and not produce, as before (South Africa, c. 1900). Black people had no choice other than to sell their labour in the mining industry. Colonialism caused fragmentation in family life. It changed the division of labour in black families (women and children took over what were traditionally men’s tasks) and it diminished the father’s authority in his family (Maforah, 1987). The erection of factories created a huge demand for labour and many fathers left their families behind in rural areas while they relocated to cities to find employment in factories and mines. Industrialization thus resulted in the urbanization of black people. A migrant labour sys tem came into existence: men stayed in single quarters in cities and visited their families in the rural areas infrequently (cf. de Haas, 1989; Mabetoa, 1994). Although blacks did not necessarily fully incorporate and accept a westernized lifestyle, the latter did impact significantly on the African culture and way of life (Bester, 1994; Carstens, 1995). Racism conveyed the message that blacks were inferior to whites; and many black people assumed a westernized lifestyle and value system in an attempt to become like white people (De Haas, 1989). Unequal distribution of power and wealth between black and white people, and the supremacy of whites, were a part of South African society long before the introduction of racial segregation as a government policy. The Nationalist Party government, which came into power in 1948, introduced numerous laws that impacted negatively on black people. Appolis (1996) explains the most prominent of these laws. Influx control policies required blacks to carry passes with them at all times. A pass had to contain the bearer’s identity number, as well as his or her place of legal residence. Black people could work in urban areas, but could not obtain permanent residence there. Security laws determined that any attempt to bring about political, social or economic change was unlawful. Furthermore, black people were prohibited from participating in so-called illegal gatherings, which were defined as any gatherings of more than two persons. Taxation laws required black people to pay taxes directed exclusively at them, namely tribal levies, education tax and local tax. 36 The effect of socio-political factors on the family life of blacks South Africa’s socio -political history impacted negatively on African families particularly in terms of marriage, family structures, parental authority and poverty. Traditionally, African culture under-emphasized the individual in the marriage. Westernization led to a shift in emphasis, from the kinship to the individual. Values such as love and romance increased in importance (Rautenbach & Kellerman, 1990) and individual satisfaction was emphasized over traditional values such as that of procreation (Chiwome, 1994). Polygyny declined as a result of, firstly, the influence of Christianity, which propagated monogamy; and secondly, economic considerations (Maforah, 1987). The prevalence of lobola is also declining and, in instances where it is still practised, cash has replaced cattle (Kanjo, 1994). Increased divorce rates (Amoateng, 2004) reflect the breakdown in black marriages. Accurate divorce statistics for black marriages are difficult to obtain, firstly because customary marriages that end in divorce are often not incorporated in statistics, and secondly because many marriages culminate in desertion without a legal divorce ever being obtained. According to Burman (1991), more than fifty percent of black marriages end in de facto divorce (non-formalized dissolution of marriage). Explanations for the increase in divorce can be found in the migrant labour system and the emancipation of women. The migrant labour system caused husbands to leave their wives behind in rural areas while they went to work and live in urban areas. Men often resorted to finding girlfriends in the cities, thereby deserting their families in the rural areas (De Haas, 1989). Under the new Constitution, women have obtained bargaining power that they previously lacked. For example, a wife, and not only a husband, can now choose to end a marriage (Pretorius, 2005). In addition, customary marriages are now granted full recognition (Statistics South Africa [STATSSA], 1998), which means that women who are married under customary law are also entitled to a share of the property in the case of a divorce. Increased access to education and occupations of their choice means that black women can be financially independent. This makes it easier for them to obtain a divorce if they so wish. 37 The extended family is still the dominant family structure among black families. Steyn (1993, 1994) found that approximately 40% of blacks still live in extended families. Russell (1994) estimates that this figure is probably closer to 59%. This is confirmed by Rautenbach and Kellerman (1990), who also indicated that the majority of urban black families (62,3%) are extended. De Visser and le Roux (1996), Siqwana-Ndulo (1998) and Van Vuuren (1997) also found the extended family to be the dominant family structure among blacks. Other family structures, however, are also emerging. The research of Steyn (1993, 1994), Van Vuuren (1997) and Rautenbach and Kellerman (1990) reflects the emergence of family configurations such as female -headed, male-headed and single-parent families, to name but a few. Boult and Cunningham (1992) view the female-headed, single-parent family as the fastest-growing family structure in society. In the traditional black family, the father was the undisputable family head and parental authority was non-negotiable. Several authors indicate that there is a definitive decline in parental authority in black families (Campbell, 1994; Thekisho, 1990; Viljoen, 1994). One of the main reasons for this decline can be found in the political uprisings of the seventies (commonly known as the “struggle” against apartheid). The older generations disapproved of the youngsters’ challenging of authority; and the youngsters in turn accused their elders of passivity (Viljoen, 1994). These differences created a communication gap between the generations, and youngsters started showing less respect for their elders. Children in the post-apartheid era are in a much better position to obtain an education than their parents and grandparents were during the height of apartheid. Although parents encourage their children to pursue an education, they are also acutely aware of the fact that they cannot advise them on all issues (Gilbert, van Vlaenderen & Nkwinti, 1995). The migrant labour system and poverty also contributed to the decline in parental authority. Parents left their children behind while searching for employment elsewhere, which diminished their authority over their children and limited the time available to spend with them (Viljoen, 1994). Exposure to Western values through the media and school also led to children becoming less submissive to their parents, and more challenging of parental authority (Maforah, 1987). 38 Data compiled from Statistics South Africa (STATSSA), amongst other sources, showed a poverty rate of 56,9% for South Africa in 1996; in addition, calculations in 1999 indicated that about seven out of every ten blacks can be considered poor, compared to one out of every ten whites (Pelser, 2002). The Labour Force Survey conducted in September 2001 (STATSSA, 2003) revealed an unemployment rate of 35,9% for black South Africans. Black men showed an unemployment rate of 32,2%, while black women showed the highest unemployment rate of all groups, namely 39,8%. These figures seem particularly high when compared to the same survey’s unemployment rate of 29,5% for the country as a whole. The effects of poverty are unfortunately also the factors that maintain it, perpetuating the so-called cycle of disadvantage. These consequences include overcrowding, poor health and poor education. The lack of proper housing led to the development of overcrowded squatter camps. These are marked by a lack of proper sanitation, multiple sharing of washing and toilet facilities, ineffective methods of food storage and a lack of proper ventilation (Mabetoa, 1994). All these factors have hazardous effects on people’s health. For example, water - and sanitation-related diseases in South Africa, such as gastrointestinal illnesses, are more prevalent in squatter camps than in other residential areas. Pelser (2002) points out that the highest HIV infection rates often occur in socially unstable informal settlements. According to Shisana (2002), the national prevalence of HIV in South Africa is 11,4%, whereas it is 18,4% for black South Africans. The marginalization of blacks during the apartheid era was expanded to include the realm of education. So-called Bantu education was marked by a lack of proper facilities and resources. Black youth who became politicized neglected their own schooling and disrupted that of others. Owing to poverty, parents could not pay for proper schooling, such as that provided by private schools (Mabetoa, 1994). Tertiary education was a luxury to which only a small minority could gain access. The lack of proper education limited people in terms of occupational options, which meant that they had no way out of the poverty. 39 Strengths of black families The picture painted of black families appears bleak and hopeless. Fortunately, there is also another side to this coin. Black families have shown themselves to be resilient, and in the face of adversity, have managed to maintain themselves, at least to some extent. One of the main strengths in black families is their keen sense of community: the burden in respect of children, and benefits such as income and access to housing are not only shared by kinship members (Burman, 1996), but also among community members. Viljoen (1994) cites the following strengths in black families: a high degree of commitment to the family, the importance attached to children and the family, and kinship networks. The extended family, by virtue of its nature, is also a considerable strength in black families. In fact, it enabled them to withstand the hardships to which they were subjected throughout the period of apartheid (Møller, 1995). METHODOLOGY The participants in the study were comprised of 2 505 black male and female adolescents from six different secondary schools in the Mangaung municipal region. However, in many instances the number of responses did not add up to this total. This discrepancy can probably be attributed to the fact that many questionnaires were not completed in full. As the questionnaire was in English, the language barrier might have caused a lack of understanding concerning certain questions. It is also possible that participants did not know the answers to certain questions, such as those relating to their parents’ educational status. The majority of participants were Southern Sesotho-speaking (54,4%), although mother-tongue speakers of languages such as Setswana (29,4%), Xhosa (12,9%) and Zulu (1,4%) also formed part of the sample. Southern Sesotho is the most common African language in the Free State Province, and Setswana closely resembles Southern Sesotho. Thus, people who are conversant in Southern Sesotho are also conversant in Setswana. 40 The sample was obtained through convenience sampling: permission was obtained from all the involved schools and the questionnaire was distributed among and completed by learners during specific school periods. The measuring instrument was a self-compiled biographical questionnaire aimed at collecting data, similar to the data collected by Steyn (1993, 1994) in South Africa and Andrews and Morrison (1997) in the USA. The questionnaire was designed to obtain information regarding household size, family structure, the occupational and educational status of parents, marital status of parents and time spent with parents. In addition, participants also had to indicate whether or not they were taking care of household members with severe disabilities or serious chronic illnesses. RESULTS AND DISCUSSION As the goal of this article is to present demographic information, descriptive statistics were employed. Findings regarding family variables which are considered likely to influence adolescent well-being will subsequently be presented and discussed. Findings in terms of parents’ educational status are considered important variables in terms of adolescent functioning as previous research has established a link between these two factors (Kaltiala-Heino, Rimpelä, Rantanen & Laippala, 2001; Rodgers & Rose, 2002). Table 1 presents findings regarding the educational status of parents. Table 1 Educational status of parents of black adolescents in Mangaung Educational status Father Mother N % N % No schooling 66 9,9 83 7,4 Grade 1-7 87 13,0 134 12,0 Grade 8-12 324 48,3 637 56,9 Some post-school education 193 28,8 266 23,7 Total 670 100 1120 100 41 Table 1 indicates that the majority of parents have some level of high-school education. In terms of the first category of educational status (no schooling), Steyn (1993, 1994) reports figures of 14,7% and 13,9% for males and females respectively. According to Census 2001 (STATSSA, 2001), 12,1% of males and 13% of females in Mangaung have received no formal schooling. The national average in respect of black South Africans older than 20 who have received no schooling is 22,3%. Urban Econ and Andani HR Consulting (2003) report an illiteracy rate of 22% in Mangaung in 2001. In terms of the second category (Grade 1-7), Census 2001 (STATSSA, 2001) cites figures of 30,1% for males and 31,5% for females. Census 2001 (STATSSA, 2001) reports figures of 52,1% for males and 50% for females in terms of the third category, namely grade 8-12. In terms of this category, Steyn (1993, 1994) reported figures of 53,4% and 47,8% for males and females respectively. With regard to the last category (some form of post -school education), Steyn (1993, 1994) reported figures of 4,8% for males and 2,9% for females. Census 2001 (STATSSA, 2001) reported figures of 5,7% and 5,5% for black males and females respectively in Mangaung. Urban Econ and Andani HR Consulting (2003) also found levels of higher education in Mangaung to be below 5%. Compared to these findings, the findings of this study seem particularly high. The reason for this discrepancy lies in the fact that this study took into account any form of post-school education, whereas the other studies referred specifically to tertiary education. When findings of the current study are compared to the other aforementioned results, it appears that parents of respondents in this study have obtained higher levels of education than most other blacks both in Mangaung and nationally. A likely explanation lies in the fact that better-educated parents tend to place a higher premium on academic qualifications and are therefore also likely to emphasize their children’s education. In other words, the participants in the present study could to a cer tain extent be regarded as a selected group. Furthermore, if the average age of the present participants is taken into account, their parents are most likely also younger than many adult residents of Mangaung. Consequently, they have probably had more access to education than the older residents who grew up in another political system where they were to a large extent denied the same educational opportunities. 42 Research has established a relationship between the employment status of parents and adolescent well-being (Barling, Zacharatos & Hepburn, 1999; Sund, Larsson & Wichstrøm, 2003). Results in terms of parents’ employment status are presented in Table 2. Table 2 Employment status of parents of black adolescents in Mangaung Employment status Fathers Mothers N % N % Employed 801 84 1081 73,9 Unemployed 153 16 382 26,1 Total 954 100 1463 100 As is clear from Table 2, the majority of parents involved in this project were employed at the time of the survey. The percentages for unemployment seem particularly low when compared to census data: Census 2001 (STATSSA, 2001) reported unemployment rates of 42,7% and 49,4% for black males and females respectively. Urban Econ and Andani Consulting (2003) reported that 40,09% of the economically active population of Mangaung are unemployed. In terms of national unemployment rates, Census 2001 (STATSSA, 2003) reported a rate of 43,3% and 57,8% for black males and females respectively. The Labour Force Survey of September 2001 (STATSSA, 2003) presented much lower figures than those of the census data, namely 32,2% for black men and 39,8% for black women. Better education increases a person’s chances of obtaining employment. Given the findings presented in Table 1, it is thus not surprising that parents in the current study are in a better position in terms of employment than most other black adults. Some studies have linked family structure with psychological well-being in adolescence (Corcoran & Franklin, 2002; Gore, Aseltine & Colton, 1992; Kaltiala- Heino et al., 2001). Table 3 contains data regarding family structure. 43 Table 3 Family structures of black families in Mangaung Family structure N % Mother and children 374 19,5 Father, mother and children (nuclear) 754 39,2 Male -headed extended 363 18,9 Female-headed extended 430 22,4 Total 1921 100 Prominent in Table 3 is the high percentage of single-mother families, which account for almost one-fifth of the total sample. If the female -headed, extended families are added to this number, the finding is even more significant. This result supports Boult and Cunningham’s (1992) viewpoint that the female-headed, single-parent family is the fastest -growing family structure in society. Andrews and Morrison (1997) also reported an increase in single-parent families in South Carolina, USA. Census 1996 (STATSSA, 1998) revealed that 46% of black children under the age of seven lived with their mothers only, as opposed to 2% who lived with their fathers only. One of the main factors contributing towards the increase in single -mother families is the emancipation of women, which has led to emotional and financial independence. As pointed out by Jones (1999), many women prefer to be single, as their husbands are often economic and emotional liabilities instead of offering them support. According to Pelser (2002), the adolescent birth rate among blacks increased by 32,7% from 1991 to 1998, which in turn led to an increase in single motherhood. As the birth of children out of wedlock is sanctioned in modern society, many young mothers opt to stay single. Extended families make up the majority of the sample (41,3%), while nuclear families account for 39,2% of the sample. These findings corroborate Steyn’s (1993, 1994) findings that 39,8% of black South African families are nuclear and 40% are multigenerational. However, caution should be exercised in the interpretation of results as, owing to the fluid nature of the black family, findings merely reflect the position of the family at the time of the survey (Russell, 1994; Siqwana-Ndulo, 1998). Even though the extended family is still the most prevalent family structure 44 among blacks (even if marginally so), it cannot be denied that western influences have changed the traditional black family and that new family structures are emerging. Marital status is an important variable in family functioning; for example, several studies have indicated the negative effects of divorce on children (cf. Amato & Keith, 1991; Hetherington, 2003). Results regarding parents’ marital status are shown in Table 4. Table 4 Marital status of parents of black adolescents in Mangaung Marital status of parents N % Married 1047 61,3 Never married/live together 289 16,9 Divorced 176 10,3 Separated 90 5,3 Widowed 105 6,2 Total 1707 100 Table 4 indicates that the majority of parents in the sample are married. Unfortunately this result does not take account of parents who may have obtained a divorce from a previous spouse and subsequently re-married (i.e. stepparent families). Steyn (1993, 1994) reported divorce rates of 1,8% for black males and 4,6% for black females. According to STATSSA (2004), the modified divorce rate (divorce rate for married couples) was 595 per 100 000 in 2001. It should be borne in mind, however, that STATSSA ni corporates only those divorces obtained from registrars at the Supreme Courts. Although all the above reported results are not directly comparable with each other, they do seem to suggest that the divorce rate among black people, when compared to other population groups, is actually not high. STATSSA (2004) indicates that white people, for example, have a higher divorce rate than black people. In addition, 9 630 African children were affected by divorce during 2001, compared to 14 854 white children dur ing the same period. 45 Caution should be exercised in the interpretation of divorce figures. Firstly, the questionnaire utilized in this study did not establish whether a currently married person had previously been divorced or not. Secondly, many black marriages end in permanent desertion, without a formalized divorce having ever been obtained. Burman (1991) is of the opinion that approximately 50% of black marriages end in divorce. Thirdly, as many black marriages are solemnized as customary marriages and not as civil/legal marriages, official divorces are never obtained in cases where these marriages break up. In addition, divorces may also have been granted by religious or traditional authorities and not legally through the courts (STATSSA, 2004). The fact that customary marriages are now granted full recognition and are officially registered (STATSSA, 1998) should lead to more accurate statistics in terms of black divorce rates. Amoateng (2004) reports that blacks are less likely to marry than white and Asian South Africans. If partners cohabit without getting married officially, the break-up of such relationships will obviously not be reflected in divorce statistics. Literature indicates that household size has been shown to influence children’s well- being (Berk, 2003; Sund, Larsson & Wichstrøm, 2003; Surajnarayan, 1991). Results regarding household size are contained in Table 5. Table 5 Household size of black families in Mangaung Household size N % 1-3 members 370 17,3 4-6 members 1399 65,4 7 or more members 370 17,3 Total 2139 100 Table 5 shows that the majority of households in this sample consisted of four to six members. The average household size was five, with a standard deviance of 1,76. According to Urban Econ and Andani HR Consulting (2003), the national average household size in South Africa is 3,95 and the average household size in Mangaung is 3,48. The majority of households in this sample (82,7%) thus not only have 46 household sizes that are larger than the other reported averages, but the average household size of the sample is also larger. The argument that smaller households are financially more viable and therefore preferred, is not supported in this study. It should also be borne in mind that a large household size does not necessarily imply a large family. In addition, the results do not indicate whether respondents and their parents prefer large households or not. Parental involvement impacts on various aspects of adolescent functioning. Parent absence is linked to social, academic and emotional problems (Cook & Cook, 2005; Flouri, 2005; Updegraff, Mchale, Grouter & Kupanoff, 2001). Findings regarding time spent with parents per week are presented in Table 6. Table 6 Time spent with parents per week Time spent with parents per week Father Mother N % N % Up to 6 hours 545 53,3 402 23,5 More than 6 hours 477 46,7 1312 76,5 Total 1022 100 1714 100 The most prominent finding contained in Table 6 is that adolescents spend much more time with their mothers than with their fathers. In their study, Andrews and Morrison (1997) reported that children spend little meaningful time with their parents, especially their fathers. In South Africa, job opportunities and availability often require fathers to move to other towns and cities without their families. These fathers are therefore not in a position to have as much contact with their children as they may wish to have. Andrews and Morrison (1997) point out the negative consequences of father absence in a child’s life. For example, children are more likely to be physically abused, and there is an increased risk of alcohol and substance abuse, poor scholastic performance and delinquent behaviour. Other negative consequences that have been linked to father absence include lower self -esteem, emotional instability, lower scores in terms 47 of educational aptitude and achievement, early sexual activity and teenage pregnancy, and conduct problems (Ellis et al., 2003; Flouri, 2005; Johnson, 1993; Murray and Sandqvist, 1990). In addition, the factors associated with father absence (e.g. deterioration in parental control) can also impact negatively on the involved adolescents (cf. Ellis et al., 2003) A prominent concern in terms of not only father absence, but also mother absence, is that children of working parents are often unsupervised before and after school (cf. Lerner, 2002). This is an obviously undesirable situation that can lead to social, academic and emotional problems (cf. Cook & Cook, 2005). In an exploration of the historical context of father absence, Grundmann (1996) found that father absence during times of war (when many families were affected) was experienced as relatively normal, and therefore did not influence children as negatively as one would expect. This led him to conclude that an increase in the number of single -parent families may lead to children experiencing this as a normal condition. This, in turn, may decrease the impact of father absence on children. The struggle for freedom in South Africa led to the fragmentation of many black families. Fathers often had to go into hiding or were exiled. In addition, migrant labour also became a normal occurrence in black families. In accordance with Grundmann’s findings, father absence might thus not have been perceived by black families as being unduly disruptive, in contrast to their white counterparts. This perception has probably been carried over from earlier years, and may have lessened the severity of father absence, even up to the present day. Despite this possibility, however, father absence is not an ideal situation for a family. Table 7 provides possible reasons as to why adolescents do not spend sufficient time with parents. 48 Table 7 Reasons why adolescents do not spend enough time with parents Reason Fathers Mothers N % N % Parent passed away 313 15,8 99 5,1 Parent’s work 466 23,5 446 23,0 Parents separated 268 13,5 97 5,0 Parents divorced 162 8,2 49 2,5 Parents never married 161 8,1 55 2,8 Parent too busy at home 192 9,6 182 9,4 Spends enough time with parent 424 21,3 1 012 52,2 Total 1 986 100 1 940 100 According to Table 7 most participants were of the opinion that their parents’ work kept them from spending time with them. The finding that parents’ work impacts on family life is to be expected (Angrist & Johnson, 2000). In the South African context, the long distances that black parents often have to travel between the workplace and home, as well as migratory labour (Richter, in press), make it even more difficult to spend sufficient time with children. Yeung, Sandberg, Davis-Kean and Hofferth (2001) reported similar findings in their study: fathers’ work caused them to spend less time with their children dur ing the week; mothers, on the other hand, remained closely involved with children even though they were working. Only about one-fifth of adolescents feel that they are spending sufficient time with their fathers, whereas slightly more than one -half think they are spending sufficient time with their mothers. As most people are having fewer children nowadays than in former years, one would expect parents to have more time available to spend with their children (Andrew & Morrison, 1997). Yet the opposite appears to be true. Most modern families have dual-earner parents, which leads to less time being spent with children. Although the finding regarding time spent with fathers is prominent, it should not be overlooked that a substantial number of participants also did not perceive the time spent with their mothers as suffic ient. This is an important finding in view of the fact that mothers still appear to carry the most responsibility for 49 parenting (Updegraff et al., 2001; Yeung et al., 2001). The possibility that participants may have a need to spend more time with their parents should be considered. Disability and chronic illness place a strain on all people in the caregiver families, including children (Makola, 1994; Naylor & Prescott, 2004). Findings regarding the taking care of household members with severe disabilities or serious chronic illnesses are presented in Table 8. Table 8 Households taking care of members with severe disabilities or serious chronic illnesses Households taking care of members with severe N % disabilities or serious chronic illnesses Yes 852 41,8 No 1185 58,2 Total 2037 100 According to Table 8, a large percentage of participants have indicated that they take care of members with severe disabilities or serious chronic illnesses in their households. Andrews and Morrison (1997) also found that there was an increase in the number of families who care for severely disabled people in South Carolina in the USA. STATSSA (2003) indicates that 5% of the population suffer from serious disabilities. This figure is not comparable with the findings in the current study, as the latter also included serious chronic illnesses. It should also be borne in mind that STATSSA (2003) defines a disability as “a physical or mental handicap which has lasted for six months or more … which prevents the person from carrying out daily activities independently, or from participating fully in educational, economic or social activities” (p. vii). In the current study the definition of “severe disability” and “serious chronic illness” was left to each participant. This might have led to an increase in the number of participants who answered “yes” to this question. The prevalence of HIV/Aids probably also contributed considerably to the number of participants who answered “yes”. It is nonetheless generally accepted that disability and chronic illness place great strain on the caregiver familie s (Makola, 1994). For 50 example, expensive treatments and medication can place a heavy financial burden on such families, while the psychological impact (e.g. emotions such as sadness, anger and resentment) should not be underestimated (cf. Hartley, Ojwang, Baguwemu, Ddamulira & Chavuta, 2005). These families are also sometimes ostracized by their communities, especially if there is a stigma attached to an illness, such as HIV/AIDS. Little is known about the needs of siblings of children with disabilities, but research has shown that these siblings do have concerns (e.g. about friends, their future, etc.) and needs that should be addressed (cf. Naylor & Prescott, 2004). This probably also applies in the case of the current study. CONCLUSION The findings contained in this article show that black families managed to a large extent to remain intact, despite the pressure exerted by socio-political factors. Positive changes since 1994 also seem to have strengthened the family. The educational and employment status of parents is improving and a commitment to marriage appears to still exist. The multigenerational family remains the predominant family structure, although the influence of modern society can clearly be seen in the fact that other family structures also occur. The large average household size indicates that “ubuntu” (taking care of each other) is still an important value among Africans. The fact that such a large number of households care for disabled or chronically ill people, further illustrates this aspect. A regrettable finding is that adolescents are not spending sufficient time with their parents, particularly their fathers. Single-parent and female-headed families, which are often faced with financial struggles, are also increasing. On the whole, the results reflect changes in the black family, but also indicate that families in this study have remained reasonably stable despite these changes. Moving away from the predominantly patrilinear family structure seems to be an important shift in the black family. This implies new challenges. For example, the female-headed family needs to be stabilized financially (Pelser, 2002). Father absence, an obvious concern for respondents in this study, needs to be further investigated in an attempt to rectify the situation. It is also recommended that 51 families with disabled or chronically ill members should not only receive material support, but should also be empowered to deal with the psycho-social impact associated with this variable. However, the present results should be interpreted with caution. The African family is by nature dynamic and highly fluid. The results thus only reflect the position of the involved families at the time of the study. A major factor that can be expected to impact on the black family is the HIV/AIDS epidemic. The latter taxes families severely in terms of economic, social and emotional resources (Ntozi & Zirimenya, 1999). This is probably why Foster et al. (1997) found that relatives of orphaned children were reluctant to care for them. Changes in family structure can thus be expected and Crewe (2002) suggests that the definition and construction of families should be redefined in the future. Although the sample size of this study is relatively large, it was restricted to a single residential area in South Africa and results cannot merely be generalized to other geographical areas. On the other hand, a need for more localized data in terms of family functioning, has been expressed (National Programme of Action for Children in South Africa, 2002). As the sample consisted of adolescents attending high school, it is possible that it reflects a fairly stable part of the community, which may have skewed the results. The concept “severely disabled/chronically ill” was too vague and this made it very difficult to adequately interpret findings in this regard. Future research should aim to present all participants with a single definition of each of these phenomena, and to measure them separately. The relationship of the participant to the disabled person should also be established. In addition, as pointed out earlier, the needs and concerns of participants with regard to disabled or chronically ill household members need to be investigated and addressed. The design of the questionnaire made it difficult to obtain accurate figures in terms of divorce. It is thus not certain whether the recorded divorce figures accurately reflect the prevalence of divorce in the sample group. Future research should attempt to obtain more specific information regarding, in particular, the marital status of black parents. Research should not only aim at investigating black family structures as such, but should also attempt to identify factors that cause a high prevalence of 52 variables such as female-headed households and absent fathers. This will contribute to a better understanding of the elements impacting on the black family. Research should endeavour to identify stable, functional families and determine the factors that make them stable and functional. These factors should then be applied in assisting dysfunctional families (Osher, 2001). In the final analysis, Billingsley’s (1992) statement regarding the African-American families also rings true for South African families, namely that these families “are both weak and strong, but their strengths are by far more powerful and contain the seeds of their survival and rejuvenation” (p. 17). 53 LIST OF REFERENCES Amato, P.R. & Keith, B. (1991). Parental divorce and the well-being of children: A meta analysis. Psychological Bulletin, 110 (1), 26-46. Amoateng, A.Y. (2004, August). The South African family: Continuity or change? Paper presented at the Human Sciences Research Council Ten Years of Democracy Seminar Series, Pretoria, South Africa. Andrews, A.B. & Morrison, B. (1997). The changing American family: Impact on South Carolina’s children . Columbia: Institute for Families in Society. Angrist, J.D. & Johnson, J.H. (2000). Effects of work-related absence on families: Evidence from the Gulf war. Industrial and Labor Relations Review, 54(1), 41-58. Retrieved March 1, 2005, from Business Source Premier database. Appolis, K.U.C. (1996). From f ragmentation to wholeness: The black South African family under siege. New York: University Press of America. Barling, J., Zacharatos, A. & Hepburn, C.G. (1999). Parents’ job insecurity affects children’s academic performance through cognitive difficulties. Journal of Applied Psychology, 84 (3), 437-444. Berk, L.E. (2003). Child Development (6th ed.). Boston: Pearson Education. Bester, C.W. (1994). The consequences of urbanization and westernization on black family life in South Africa. Social Work, 30(2), 101-106. Billingsley, A. (1992). Climbing Jacob’s ladder: The enduring legacy of African- American children. New York: Simon & Schuster. Boult, B.E. & Cunningham, P.W. (1992). Black teenage pregnancy: An African perspective. International Journal of Adolescence and Youth, 3, 303-309. 54 Burman, S. (1991). Capitalising on African strengths: Women, welfare and the law. South African Journal on Human Rights, 7 (2), 215-221. Burman, S. (1996). Intergenerational family care: Legacy of the past, implications for the future. Journal of Southern African Studies, 22(4), 585-598. Campbell, C. (1994). Township families and youth identity: The family’s influence on the social identity of township youth in a rapidly changing South Africa. Pretoria: Human Sciences Research Council. Carstens, C. (1995). Die invloed van televisie op die verwesteringsproses by die swart adolessent. Ongepubliseerde doktorale proefskrif. Pretoria: Universiteit van Suid-Afrika. Chiwome, E. (1994). The role of the church in the shift of the marriage and the family institutions from tradition to modernity: The case of Zimbabwe. In T. Sono (Ed.), African family and marriage under stress. Pretoria: Centre for Development Analysis. Cook, J.L. & Cook, G. (2005). Child Development: Principles & Perspectives. Boston: Pearson Education. Corcoran, J. & Franklin, C. (2002). Multi-systemic risk factors predicting depression, self-esteem, and stress in low SES and culturally diverse adolescents. Journal of Human Behavior in the Social Environment, 5(2), 61-76. Crewe, M. (2002). AIDS in Southern Africa. Postgraduate Medical Journal, 78, Retrieved February 23, 2005, from http://pmj.bmjjournals.com/cgi/content/full/78/917/127 De Haas, M. (1989). The socio-political roots of the breakdown of marriage amongst African South Africans. The Social Work Practitioner/Researcher, 2(4), 3-11. 55 De Visser, J. & le Roux, T. (1996). The experience of teenage pregnancy in Knoppieslaagte. South African Journal of Sociology, 27(3), 98-105. Ellis, B.J., Bates, J.E., Dodge, K.A., Fergusson, D.M., Horwood, L.J., Pettit, G.S. & Woodward, L. (2003). Does father absence place daughters at special risk for early sexual activity and teenage pregnancy? Child Development, 74(3), 801- 821. Flouri, E. (2005). Father’s involvement and psychological adjustment in Indian and white British secondary school age children. Child and Adolescent Mental Health, 10 (1), 32-39. Foster, G., Makufa, C., Drew, R. & Kralovec, E. (1997). Factors leading to the establishment of child-headed households: The case of Zimbabwe [Electronic version]. Health Transition Review, 7, 155-168. Gilbert, A., van Vlaenderen, H. & Nkwinti, G. (1995). Planting pumpkins: Socialization and the role of local knowledge in rural South Africa. South African Journal of Psychology, 25(4), 229-235. Gore, S., Aseltine, R.H., Colton, M.E. (1992). Social structure, life stress and depressive symptoms in a high school-aged population. Journal of Health and Social Behavior, 33, 97-113. Grundmann, M. (1996). Historical context of father absence: Some consequences for the family formation of German men. International Journal of Behavioral Development, 19(2), 415-431. Hartley, S, Ojwang, P., Baguwemu, A., Ddamulira, M. & Chavuta, A. (2005). ow do carers of disabled children cope? Ugandan perspective. Child: care, health and development, 31 (2), 167-180. Hetherington, E.M. (2003). Social support and the adjustment of children in divorced and remarried families [Electronic version]. Childhood, 10(2), 217-236. 56 Johnson, W.B. (1993). Father uninvolvement: Impact, etiology and potential solutions. Journal of Psychology and Christianity, 12(4), 301-311. Jones, S. (1999). Singlehood for security: Towards a review of the relative economic status of women and children in women-led households. Society in Transition, 30(1), 224-238. Kaltiala -Heino, R., Rimpelä, M., Rantanen, P. & Laippala, P. (2001). Adolescent depression: The role of discontinuities in life course and social support. Journal of Affective Disorders, 64, 155-166. Kanjo, G. (1994). Western influences on African marriage in Malawai. In T. Sono (Ed.), African family and marriage under stress. Pretoria: Centre for Development Analysis. Lerner, R.M. (2002). Adolescence: Development, diversity, context and application. New Jersey: Pearson Education. Mabetoa, M. (1994). Cycles of disadvantage of African families in South Africa. In T. Sono (Ed.), African family and marriage under stress. Pretoria: Centre for Development Analysis. Maforah, N.F. (1987). The changing role of the black family and its adaptation to the present needs. Social Work, 23, 262-268. Makola, D. (1994). The effects of a chronic disorder on the child, his family and environment. South African Paediatric Medicine, 11, 24-30. Møller,V. (1995). Home environment and educational achievement among high- school pupils living in three-generation urban black households. South African Journal of Sociology, 26(3), 87-97. 57 Murray, A. & Sandqvist, K. (1990). Father absence and children’s achievement from age 13 to 21. Scandinavian Journal of Educational Research, 34 (1), 3-28. National Programme of Action for Children in South Africa (2002). Children in 2001: Report on the state of the nation’s children. Pretoria: The Presidency. Naylor, A. & Prescott, P. (2004). Invisible children? The need for support groups for siblings of disabled children. British Journal of Special Education, 31 (4), 199-206. Ntozi, J.P.M. & Zirimenya, S. (1999). Changes in household composition and family structure during the AIDS epidemic in Uganda. The continuing African HIV/AIDS epidemic, 193-209. Retrieved February 23, 2005, from http://htc.anu.edu.au/pdfs/Contin uingHIV/Ntozi_Ziri.pdf Osher, T.W. (2001). Overview and implications of the Surgeon General’s report on mental health: A family perspective. In C.C. Newman, C.J. Liberton, K. Kutash & R.M. Friedman (Eds), A system of care for children’s mental health: Expanding the research base, 13 th Annual Proceedings. Florida: The Research and Training Center for Children’s Mental Health. Pelser, A. (2002). The social landscape of sexual and reproductive health in South Africa with particular reference to HIV/AIDS. In D. van Rensburg et al., Strengthenig local government and civic responses to the HIV/AIDS epidemic in South Africa. Bloemfontein: Centre for Health Systems Research and Development. Preston-Whyte, E. (1993). Women who are not married: Fertilit y, “illegitimacy” and the nature of households and domestic groups among single African women in Durban. South African Journal of Sociology, 24 (3), 63-71. Pretorius, E. (2005). Family life in South Africa. In J.L. Roopnarine & U.P. Gielen (Eds), Families in global perspective. Boston: Pearson Education. 58 Rautenbach, E.A. & Kellerman, A.P.R. (1990). Die swart middelklasgesin in die stad. Maatskaplike Werk, 26, 228-240. Richter, L. (in press). Fathers and children: Necessary or nice? In L. Richter & R. Morrell (Eds), Baba? Men and fatherhood in South Africa. Pretoria: Human Sciences Research Council. Rodgers, K.B. & Rose, H.A. (2002). Risk and resiliency factors among adolescents who experience marital transitions. Journal of Marriage and the Family, 64(4), 1024-1037. Retrieved September 14, 2004, from Academic Search Premier database. Russell, M. (1994). Do blacks live in nuclear family households? An appraisal of Steyn’s work on urban family structure in South Africa. South African Sociological Review, 6 (2), 56-67. Shisana, O. (2002). The current situation of HIV/AIDS in South Africa. Retrieved August 28, 2004, from http://www.hsrcpublishers.ac.za Siqwane-Ndulo, N. (1998). Rural African family structures in the Eastern Cape province , South Africa. Journal of Comparative Family Studies, 29 (2), 407- 417. South Africa c. 1900 . (n.d.). Retrieved December 6, 2001, from http://library.thinkquest.org/27629/themes/society/sa000.html Statistics South Africa (1998). Marriages and divorces. Pretoria: Author. Statistics South Africa (2001). Database: Census 2001. Retrieved August 26, 2004, from http://www.statssa.gov.za/census2001/Census/Dialog/Saveshow.asp Statistics South Africa (2003). Census 2001: Census in brief (2nd ed) [Electronic version]. Pretoria: Author. 59 Statistics South Africa (2004). Marriages and divorces 2001 [Electronic version]. Pretoria: Author. Steyn, A.F. (1993, 1994). Family s tructures in the RSA. Pretoria: Human Sciences Research Council. Steyn, A.F. & Viljoen, S. (1996). The state of family theory and research in South Africa. In M.B. Sussman & R.S. Hanks (Eds), Intercultural variation in family research and theory: Imp lications for cross-national studies. New York: Haworth Press. Sund, A.M., Larsson, B. & Wichstrøm, L. (2003). Psychosocial correlates of depressive symptoms among 12-14 year old Norwegian adolescents. Journal of Child Psychology and Psychiatry, 44, 588-597. Surajnarayan, G. (1991). The influence of the family on the adolescent’s behaviour. Unpublished master’s dissertation. Pretoria: University of South Africa. Thekisho, G.N. (1990). Family disorganization among blacks. Social Work Practice, 1 (90), 6-7. Updegraff, K.A., Mchale, S.G., Grouter, A.C. & Kupanoff, K. (2001). Parents’ involvement in adolescents’ peer relationships: A comparison of mothers’ and fathers’ roles. Journal of Marriage and the Family, 63(3), 655-668. Retrieved March 1, 2005, from Academic Search Premier database. Urban-Econ & Andani HR Consulting (2003). Mangaung economic development program (first draft). Bloemfontein: Mangaung Local Municipality. Van Vuuren, A. (1997). The strengths and problems of black families in informal settlements. Pretoria: Human Sciences Research Council. Viljoen, S. (1994). Strengths and weaknesses in the family life of black South Africans. Pretoria: Human Sciences Research Council. 60 Vorster, L.P. (1993). Substitusie in die inheemse reg. TSAR, 4, 552-560. Yeung, W.J., Sandberg, J.F., Davis-Kean, P.E. & Hofferth, S.L. (2001). Children’s time with fathers in intact families. Journal of Marriage and the Family, 63(1), 136-154. Retrieved March 1, 2005, from Academic Search Premie r database. 61 ARTICLE THREE THE INFLUENCE OF SELECTED FAMILY VARIABLES ON DEPRESSION IN BLACK ADOLESCENTS 62 TABLE OF CONTENTS INTRODUCTION............................................................................ 65 LITERATURE REVIEW ................................................................ 66 METHODOLOGY .......................................................................... 71 RESULTS AND DISCUSSION........................................................ 73 CONCLUSION................................................................................ 81 LIST OF REFERENCES................................................................. 84 LIST OF TABLES Page Table 1 Cronbach’s α -coefficients for the CES-D scale 72 Table 2 Distribution of group in terms of family variables 73 Table 3 Results of ANOVAs with depression as 77 dependent variable Table 4 Results of t-tests with depression as dependent 79 variable Abstract It is generally accepted that families influence the psychological well-being of their adolescent members. This study thus aims to examine the relationship between depression and selected family variables among black families in South Africa. The depression rate for the group was high (26%). No statistically significant relationships were found between depression scores on the one hand and family structure, marital status of parents, household size, educational status of parents, employment status of fathers and time spent with parents, on the other hand. The only two variables that showed statistically significant relationships with depression scores, were the employment status of mothers and the taking care of household members with severe disabilities or serious chronic illnesses. Family variables thus appear to have virtually no significant impact on adolescent depression, which means that there are other factors contributing to the high depression rate in this sample. Further research is necessary to explore the prevalence of depression amongst black adolescents, and its associated factors. INTRODUCTION It is generally accepted that the family is one of the main agents of socialization that influence a person’s functioning in society (de Visser & le Roux, 1996; Fako, 1996; Mboya, 1998). It is therefore understandable that the family is considered essential in raising adolescents who are socially and emotionally well-adapted (Andrews & Morrison, 1997; Lopez & Thurman, 1993). In fact, several studies have linked high familial support with high self-esteem and lower levels of depression among adolescent family members (Cumsille & Epstein, 1994; Marcotte, Marcotte & Bouffard, 2002; McFarlane, Bellissimo & Norman, 1995). The family, however, may also have a detrimental effect on adolescents’ psychologica l well-being. For example, familial conflict such as divorce can lead to anxiety, depression, lowered self-esteem and aggressiveness (Armistead, Forehand, Beach & Brody, 1995; Hetherington, 2003; Louw, van Ede & Ferns, 1998). It is therefore a pity that very little research has been conducted on the influence of familial factors on children and especially adolescents in black South African families. The fact that South Africa’s history of racial discrimination played a major role in this state of affairs should not be used as a rationalization, but should rather serve to encourage research in this regard. This is especially true in view of the fact 65 that the effects of apartheid did not end with the first democratic elections in 1994, but will most probably continue for generations to come. It was thus decided to investigate the relationship between selected family variables and the psychological well-being of black adolescents in Mangaung, a township in Bloemfontein, the capital of the Free State Province. LITERATURE REVIEW Global descriptions of the adolescent years vary greatly. On one end of the scale, adolescence is frequently described as comprising years of turmoil, emotional instability and the questioning of previously accepted values and rules (Dumont & Provost, 1999; Houlihan, Fitzgerald & O’Regan, 1994). On the other hand, authors such as Frey and Röthlisberger (1996) and Connelly, Johnston, Brown, Mackay and Blackstock (1993) express the viewpoint that the turmoil experienced in adolescenc e is often exaggerated and that most adolescents manage to adapt well. As far as South Africa is concerned, it seems that the enormous adjustments that adolescents had to make since 1994 have led to a struggle with various issues such as self-esteem, familial factors, emotional instability and morality (Boulter, 1995). These factors, in turn, have influenced aspects such as relationships with parents and peers, and identity formation. Black South African adolescents are particularly challenged in the process of resolving these issues, as they have to find a synthesis between their traditional culture and Western culture (Thom, 1988). An important research area in this field concerns the epidemiology of psychological problems experienced during adolescence. Southall and Roberts (2002) consider adolescence to be a time of “considerable vulnerability” (p. 563) to psychological problems such as depression. In fact, many authors are of the opinion that the prevalence of depression is on the increase amongst adolescents (Kaltiala-Heino, Rimpelä, Rantanen & Laippala, 2001; Marcotte, Marcotte, et al., 2002). Reported depression rates vary between 2% and 29% (Rushton, 2002; Chan, 1995). These large discrepancies are usually attributed to the use of different definitions of 66 depression and the utilization of diverse measuring instruments (Brage, 1995; Marcotte, Fortin, Potvin & Papillon, 2002). Authors such as Corcoran and Franklin (2002) and Rushton (2002) confirm the existence of a relationship between culture and adolescent depression. In terms of different cultures, Chan (1995) reports a 29% depression rate for Chinese youth, while Kaltiala-Heino et al. (2001) found a 17,2% prevalence among Finnish adolescents. Australian adolescents show a depression rate of 5,2% (Rey, Sawyer, Clark & Baghurst, 2001). In South Africa, a study amongst adolescents in the Western Cape found a depression rate of 6,6% for major depressive disorder and 31,6% for dysthymic disorder (Ensink, Robertson, Zissis & Leger, 1997), while depression rates of 21,1% and 18,3% were reported among Northern Sotho and Venda adolescents respectively (Bach, 2004). Patten et al. (1997), on the other hand, did not find that ethnicity or race significantly influenced the prevalence of adolescent depression. A trend confirmed by most authors is that girls have higher depression rates than boys (Duggal, Carlson, Sroufe & Egeland, 2001; Marcotte, Marcotte, et al., 2002; Patten et al., 1997; Portes & Zady, 2002; Southall & Roberts, 2002). A possible reason for this phenomenon is that girls tend to internalize turmoil and events (e.g. through self-criticism and lowered self -esteem), whereas boys tend to externalize events (e.g. through overt hostility and delinquency) (Pelkonen, Marttunen & Aro, 2003; Rogers & Rose, 2002). Furthermore, Marcotte, Fortin, et al. (2002) found that pubertal maturation is more likely to occur simultaneously with transition to high school in girls than in boys. In addition, the onset of menarche also appear to increase depression in girls (Patton et al., 1996). According to Brage (1995), the majority of depressed teenagers suffer from unipolar depression, and bipolar disorder is not considered a major problem amongst teenagers. Symptoms are similar to those experienced by adults, including loss of interest, loss of energy and a disturbed sleep cycle. Other symptoms identified by Chan (1995) include body image distortions, fatigue, lowered self-esteem, guilt feelings, a sense of failure and crying. He also found that weight loss and loss of libido, common symptoms among adults, are amongst the least frequently reported 67 symptoms among adolescents. This is probably owing to increased hormonal activity during adolescence, which leads to increased libido and often also weight gain. Mash and Wolfe (1999) list symptoms such as irritability, guilt, shame, oversensitivity to criticism, substance abuse, verbal sarcasm and destructive behaviour. According to the relevant literature, cognitive processes may be considered important contributors to adolescent depression. Marcotte, Marcotte, et al. (2002) indicate that the presence and role of distorted thinking styles in adolescent depression have been neglected in research. They found that dysfunctional cognitions are linked to depression rates in both males and females. Depressed adolescents also have more dysfunctional attitudes concerning achievement. In an investigation into the role of perceived social support and dysfunctional attitudes in predicting Taiwanese adolescents’ depressive tendencies, Liu (2002) found a significant, positive correlation between dysfunctional attitudes and depression scores. Southall and Roberts (2002) and Brage (1995) indicate that negative attributional styles, combined with low self-esteem, put teenagers at a greater risk of developing depressive symptoms. In corroboration of the above, Ostrander, Weinfurt and Nay (1998) also found that negative cognitions contribute to depression in adolescence. The relationship between cognitive processes and depression is a complex one; and most researchers indicate that certain factors act as buffers, or moderating factors, in the link between the two factors. For example, Liu (2002) found that peer support had a moderating effect on the relationship between dysfunctional attitudes and depression. Other buffers that have been suggested include social support and self-esteem. Several researchers found a relationship between low self-esteem and increased depression rates (Dori & Overholser, 1999; Marcotte, Fortin, et al., 2002; Pelkonen et al., 2003; Southall & Roberts, 2002). Corcoran and Franklin (2002) found that white adolescents, who are not growing up with their parents, are at higher risk of developing lower self-esteem and higher depression than those in intact families. In terms of African-American youth, however, growing up with both parents was found to be linked to higher depression rates and growing up without parents was linked to higher self -esteem. Brage and Meredith (1994) found both an indirect link between depression and family strengths through self-esteem, as well as a direct link between depression and family strengths. According to them, the latter relationship is 68 probably affected by the developmental tasks of adolescence, such as establishing identity, obtaining autonomy, experiencing competence and acceptance by peers. The aforementioned research suggests the possibility of a complex interactive relationship between self-esteem, depression and familial factors. It thus seems plausible that self-esteem plays a significant role in contributing to depression amongst teenagers. In fact, Dumont and Provost (1999) single out self-esteem as the most important buffer against the impact of daily stressors that can lead to depression. Numerous projects have investigated the role that social and familial factors play in adolescent depression. Many researchers agree that both family support and support from peers are important variables in adolescent depression (Cornwell, 2003; Pelkonen et al., 2003; Rodgers & Rose, 2002; Wong & Wiest, 1999). Chan (1995) and Liu (2002) agree that social support from friends appears to be more significantly linked to depression than social support from the family. The majority of researchers (e.g. Cumsille & Epstein, 1994; Lasko et al., 1996; McFarlane et al., 1995; Sim, 2000; Way & Robinson, 2003), on the other hand, indicate that familial support (particularly from parents) is more significant than social support from peers in buffering teenagers against depression. Marcotte, Marcotte, et al. (2002) also link high levels of familial support with low depression levels. Contrary to these findings, Dumont and Provost (1999) and Vitaro, Pelletier, Gagnon and Baron (1995) found no significant relationships between perceived support from friends and family and depressive symptomatology among adolescents. McFarlane et al. (1995) are of the opinion that most research has succeeded in establishing only a minor link between support factors and adolescent depression, despite clinical experience suggesting that these factors are of major importance. In terms of family structure, neither Cumsille and Epstein (1994) nor Patten et al. (1997) found a significant link between family structure and depression rates amongst adolescents. Corcoran and Franklin (2002) did, however, confirm the existence of a relationship between family structure and depression. As mentioned earlier, though, race appears to influence this relationship. Kaltiala-Heino et al. (2001) confirmed an association between family structure and depression. Several familial factors have been identified as constituting an increased risk for adolescent depression. These 69 factors include an unhappy atmosphere at home (Pelkonen et al., 2003); a lack of support and cohesion and a tendency towards more criticism, hostility and rejection in the family (Duggal et al., 2001); perceived poor communication with parents (Stewart, McKenry, Rudd & Gavazzi, 1994); the timing of parental divorce (Monck, Graham, Richman & Dobbs, 1994; Palosaari & Aro, 1994); a genetic predisposition to depression and parental depression (Duggal et al., 2001; Hawkridge, 1997); parental unemployment (Visser, 1991); low parental education (Lewinsohn, Clarke, Seeley & Rohde, 1994); and parental substance abuse (Kim & Kim, 2001). Whereas Corcoran and Franklin (2002) did not find that socio-economic status influenced adolescent depression, Gore, Aseltine and Colton (1992) identified perceived standard of living as a risk factor in depression. A major problem in researching adolescent depression, is the fact that it appears to be very difficult to establish a causal direction (Chan, 1995; Liu, 2002; Patten et al., 1997; Marcotte, Marcotte, et al., 2002). In other words, it is possible that the factors considered to cause depression, may actually be caused by the depression. Although links have been established between depression and all the aforementioned variables, researchers have not yet been able to establish, beyond doubt, the direction of most of these causal relationships. For example, although research has confirmed that self- esteem contributes to depression, it may be equally true that depression can lead to lowered self -esteem. This factor should be kept in mind when interpreting research results relating to adolescent depression. 70 METHODOLOGY The research was non-experimental in nature and a criterion group design (Huysamen, 1983) was employed. The participants in this study were comprised of 2 505 black male and female adolescents in grades eight to twelve, attending six different schools in Mangaung. The sample was one of convenience. The first step in the procedure was to obtain permission to conduct the research, which was granted by the Free State Department of Education. Information regarding family variables was obtained by means of a self-compiled, biographical questionnaire aimed at collecting similar data to those obtained by Steyn (1993, 1994) in South Africa and Andrews and Morrison (1997) in the USA. The questionnaire was designed to obtain information regarding the household size, family structure, occupational and educational status of parents, marital status of parents and time spent with parents. Participants also indicate d whether, in their households, they were taking care of members with severe disabilities or serious chronic illnesses. Depression was measured by means of the Center for Epidemiologic Studies Depression Scale (CES-D Scale) (Radloff, 1977). This self -report questionnaire was developed for use among the general population and is considered highly user- friendly. It consists of 20 questions aimed at measuring current general depressive symptomatology. A minimum score of 0 and a maximum score of 60 can be obtained. A higher score indicates higher depressive symptomatology. Psychometric properties for the CES-D Scale were found to be good in the USA, and this instrument was also deemed suitable for use among adolescents (Radloff, 1991). The use of this scale among adolescents in South Africa comprises an important contribution to local research. The scale has high internal consistency, along with a well-established concurrent and construct validity (Radloff, 1977). Internal consistency in this study was measured by calculating Cronbach’s α -coefficients with the aid of the SPSS computer programme (SPSS Incorporated, 1983). The results appear in Table 1. 71 Table 1 Cronbach’s α -coefficients for the CES-D scale Questionnaire/Construct Number Minimum Maximum α -coefficients of items Depression 20 0 60 0,7511 The calculated coefficients in Table 1 indicate that the measuring instrument has reasonably high internal consistency. Owing to large differences in frequencies among the categories of the ten independent variables, it was decided to conduct an analysis of variance (ANOVA) in the case of more than two categories, and a t-test for independent groups in the case of two categories. If multivariate analyses of variance (MANOVAs) had been conducted, only respondents with fully completed questionnaires would have been included in all analyses. This means that the maximum number of respondents involved would have been 670 (refer to Table 2), which is a mere 31,3% of the total sample. Although the chosen method may lead to an increase in type 1 error, it was nevertheless followed so as to utilize as much of the information obtained from respondents as possible. In addition to determining whether results were statistically significant, the practical significance of such results was also investigated. To this end, effect sizes were calculated. When ANOVAs were utilized, the following indexes were used to determine effect sizes (Nolan, 2002): R² = SS between SS total The following guiding values were used to determine effect sizes: R² = 0,01 was considered a small effect; R² = 0,06 was considered a medium effect; and R² = 0,14 was considered a large effect. As mentioned, t-tests for independent groups were used for variables with two categories. Given tha t the sizes of the groups differed, it was decided to make use of the pooled variance estimate. According to Howell (2002), this procedure is preferable when sample sizes differ. The effect sizes of the 72 differences between the two sets of averages were calculated by means of determining Cohen’s d (Steyn, 1999). This procedure expresses the difference between the two means in terms of the size of the standard deviation. In other words, the difference has been standardized. Cohen (Nolan, 2002) laid down very general guidelines for what he considered to be small, medium and large effect sizes. He characterized d = 0,20 as a small, but probably meaningful effect size; d = 0,50 as a medium effect (half a standard deviation difference); and d = 0,80 as a large effect. Effect sizes were only calculated when results were found to be significant on either the 1% level or the 5% level. RESULTS AND DISCUSSION In order to provide a context for the results of the statistical analysis, the frequencies in terms of the ten selected independent variables are shown in Table 2. Table 2 Distribution of group in terms of family variables Independent variables N % Family structures: Mother and children 374 19,5 Father, mother and children (nuclear) 754 39,2 Male-headed extended family 363 18,9 Female-headed extended family 430 22,4 Subtotal 1921 Marital status: Married 1047 61,4 Never married/Live together 289 16,9 Divorced/Separated/Widowed 371 21,7 Subtotal 1707 Household size: 1 – 3 370 17,3 4 – 6 1399 65,4 7 or more 370 17,3 Subtotal 2139 73 Table 2 Distribution of group in terms of family variables (continued) Independent variables N % Educational level – father: None 66 9,9 Grade 1-7 87 13,0 Grade 8 –12 324 48,3 Some form of post-school education 193 28,8 Subtotal 670 Educational level – mother: None 83 7,4 Grade 1-7 134 12,0 Grade 8-12 637 56,9 Some form of post-school education 266 23,7 Subtotal 1120 Employment status – father: Employed 801 84,0 Unemployed 153 16,0 Subtotal 954 Employment status – mother: Employed 1081 73,9 Unemployed 382 26,1 Subtotal 1463 Time spent with father per week: Up to 6 hours 545 53,3 More than 6 hours 477 46,7 Subtotal 1022 Time spent with mother per week: Up to 6 hours 402 23,5 More than 6 hours 1312 76,5 Subtotal 1714 Households taking care of members with severe disabilities or serious chronic illnesses Yes 852 41,8 No 1185 58,2 Subtotal 2037 74 Table 2 shows considerable differences in terms of feedback rates for the different variables. The lowest frequency (N=670) was reported for “educational level of father” and the highest frequency (N=2139) was reported for household size. This discrepancy is owing to the fact that many questionnaires were not completed in full. Participants may also not have been aware of facts such as their father’s educational level. In addition, questionnaires were not completed in the participants’ home language, but in English. This might have prevented full understanding of some of the questions. Complete information regarding the depression scale was available for 1 438 adolescents. The average depression score for the group was 22 (21,995) with a standard deviance of 9,12. Radloff (1991) points out that adolescents’ depression scores may be inflated, owing to the inherent nature of adolescence. For example, mood swings, an increased self-awareness coupled with a willingness to report symptoms, conflict with parents and society in general and concerns regarding peer group relationships, can lead to adolescents reporting high depression scores. These symptoms may, however, be short-lived and can even be expected to change from day to day. To allow for this occurrence, Radloff (1991) suggests a very high cut-off score of 28 in determining depression that is considered more than merely “mild”. When this cut-off score was applied in the current study, a depression rate of 26% was calculated. This is considered a high depression rate in comparison to earlier reported rates. Caution should be exercised in the interpretation of these results. It should be borne in mind that the scale used is not an “official” diagnostic tool for clinical depression. In addition, it is a self-report scale, which means that some bias might have been present. Corroborating responses were not obtained from significant others in the lives of the respondents, such as teachers or parents. In this regard it is significant that Monck et al. (1994) indicated that in most studies, there is very low agreement in terms of relations’ perceptions of the adolescent’s depression and the adolescent’s self-report on depression. It should also be taken into account that the CES-D scale measures depressive symptomatology as applicable at the time of completing the questionnaire. As a re-test was not conducted, it was not possible to report on the consistency of the results. 75 There are several possible explanations for the high depression score obtained by Mangaung youth. Contributing factors identified by international studies probably also contribute to depression in black South African adolescents. Such factors include cognitive distortions (Liu, 2002; Marcotte, Marcotte, et al., 2002), social and familial factors (cf. Brage, 1995; Cumsille & Epstein, 1994; Kaltiala -Heino et al., 2001; Sim, 2000) and poor self-esteem (Dori & Overholser, 1999; Southall & Roberts, 2002). However, in addition to the mentioned generally recognized variables, there are also etiological factors that are specifically linked to the South African context. Although the country became a democracy a decade ago, the legacy of apartheid is still very much part of black people’s psyche. Hickson and Kriegler (1991) contend that depression and lowered self-esteem were two of the main psychological effects of apartheid. From an early age, black children learnt to accept that they were somehow inferior to white people. This lowered self -esteem and confidence levels considerably . Further to this, Sonn and Louw (1989) highlight the negative effect of apartheid on the self-concept of black people. As shown in other research (Brage & Meredith, 1994; Chan, 1995; Pelkonen et al., 2003), low self-esteem is linked to higher levels of depression. It thus seems plausible that the legacy of a racist South African society led to depression amongst blacks, both directly, as well as indirectly through lowered self -esteem and a poor self-concept. Another contributing factor to high rates of depression could be poverty. The poverty that was structurally imposed on blacks during the apartheid years limited their access to education and consequently also their ability to obtain higher -paying employment. Recent statistics suggest that most bla ck people still have not managed to escape from the stronghold of poverty. For example, the Labour Force Survey of September 2001 showed an unemployment rate of 35,9% for black people (Statistics South Africa [STATSSA], 2003), and Pelser (2002) indicates that seven out of ten blacks were regarded as poor at the turn of the century. The inability to escape from poverty despite a strong desire to do so can lead to hopelessness. An increase in hopelessness has been linked to a decrease in self-esteem and an increase in depression (Dori & Overholser, 1999). Whether directly, or indirectly through a lack of self-esteem and a sense of hopelessness, poverty could lead to depression. 76 Cleveland, Williams and Herman (2004) identified a correlation between perceive d vulnerability to HIV/AIDS and depression among South Africans. Given the full extent of this epidemic in South Africa, it is highly probable that it could impact negatively on the psychological well-being of adolescents. As mentioned, ANOVAs (for varia bles with more than two categories) and t-tests for independent groups (for variables with two categories) were used to investigate the influence of family variables on depression among respondents. Table 3 summarizes findings regarding family structure, marital status, household size and the educational status of the adolescent’s mother and father. Table 3 Results of ANOVAs with depression as dependent variable Variables F p Family structure 0,33 0,7187 Marital status 0,20 0,8154 Household size 0,63 0,5313 Educational status – father 1,49 0,2161 Educational status – mother 0,86 0,4631 ** p = 0,01 * p = 0,05 As can be seen in Table 3, no significant relationships were found between depression and the five independent variables relating to the family. In terms of family structure, neither Cumsille and Epstein (1994) nor Patten et al. (1997) found this variable to be significantly linked to adolescent depression. According to Lansford, Ceballo, Abbey and Stewart (2001), family processes are of greater importance than actual family structure. Kaltiala-Heino et al. (2001) found that respondents living apart from both parents reported higher levels of depression than those living with both parents. In terms of marital status, Sund, Larsson and Wichstrøm (2003) found higher depression scores in non-intact families. Several researchers indicate that parents’ marital status as such does not have a major impact 77 on children, but that associated factors do have an impact. These associated factors include hostility and conflict between parents, perceptions of poor parental relationships (Amato & Keith, 1991; Armistead et al., 1995; Grossman & Rowat, 1995) and possibly also the age of children at the time of parental divorce (Monck et al., 1994). It seems that the quality of the parental relationship is more important than the actual marital status of parents. Regarding household size, the correlation of this variable with depression was not significant. This could be owing to the fact that household size, and not household density, was determined. More specifically, a review of the relevant literature suggests that overcrowding is more likely to adversely affect children than the actual household size (Mahabeer, 1992; Richter, 1989). Contrary to the finding in this study, Sund et al. (2003) did find a positive relationship between larger families (more than three siblings) and depression. It should be borne in mind that household size, and not family size as such, was measured in this study. It is highly probable that household size and family size may impact on adolescents in different ways. In terms of family size, for example, the literature indicates that children in smaller families receive more attention and also have better opportunities for fur ther education (cf. Rice & Dolgin, 2005). Berk (2003) indicates that the effect of family size on the family’s socio-economic status should be taken into account in determining the effect of family size on children’s well-being. Thus, a large family (or household size) will not necessarily impact negatively on a wealthy family. As far as the educational status of parents is concerned, several research projects established a link between this variable and adolescent depression. Hortacsu (1995) found that mothers’ educational status was particularly important in decreasing child insecurities. Supporting this finding, Rodgers and Rose (2002) found that adolescents whose mothers had low levels of education were more prone to both internalizing and externalizing behaviours. Lewinsohn et al. (1994) also observed a relationship between the onset of depression and low levels of parental education. For independent variables that consisted of two categories, t-tests were conducted. The results appear in Table 4. 78 Table 4 Results of t-tests with depression as dependent variable Variable N X s t p Cohen D Employment status of father: Employed 555 21,41 8,67 Unemployed 108 22,92 9,23 -1,63 0,1036 Employment status of mother: Employed 750 21,30 9,10 Unemployed 248 23,79 9,04 -3,74 0,0002** 0,27 Weekly time with father: Up to 6 hours 385 21,47 8,21 More than 6 hours 329 21,60 22,56 -0,19 0,8530 Weekly time with mother: Up to 6 hours 283 22,23 9,53 More than 6 hours 898 21,63 8,83 0,98 0,3292 Taking care of members with severe disabilities/serious chronic illnesses: Yes 551 23,13 8,77 No 834 21,32 9,34 3,61 0,0003** 0,20 ** p ≤ 0,01 * p = 0,05 Table 4 indicates that significant differences were found in average depression scores for both “employment status of mother” (p = 0,01) and “taking care of household members with severe disabilities/serious chronic illnesses” (p = 0,01). The table indicates that adolescents with unemployed mothers have higher depression scores than those whose mothers are employed. Adolescents in households who have to take care of members with severe disabilities or serious chronic illnesses, also show higher levels of depression. The corresponding effect sizes are small. It is interesting that the employment status of the mother was found to be more significant than the employment status of the father. Monck et al. (1994) found higher depression rates among girls in families where the chief earner was unemployed (whether the mother or the father). Sund et al. (2003) also reported higher depression scores among adolescents with unemployed parents. Liem and Liem (1988) found that paternal unemployment altered the general mood in a household, even in households where the mother was still earning an income. Visser 79 (1991) found more disruption and conflict in unemployed families than in employed families. In addition, unemployed fathers were perceived as less available. Although results in terms of the effect of parental employment are inconclusive, the general trend in previous research suggests that paternal unemployment is of greater consequence. The opposite was found in this study. More respondents (N=1 463) were aware of their mother’s employment status than of their father’s employment status (N=954) (refer to Table 2). As mentioned earlier, fathers may be involved in migratory labour which tends to diminish the effect of paternal employment on the children. The findings concerning disability/chronic illness should be interpreted against the background of two aspects. Firstly, the socio-political landscape in South Africa prior to 1994 led to a situation in which black communities in South Africa did not have equal access to facilities catering for chronically ill or disabled people. They had no option but to care for these people within their communities. Access to facilities is still not sufficient, as more pressing matters, such as housing, receive more attention in the post-apartheid era (Hickson & Kriegler, 1991; Nkabinde & Ngwenya, 1996). Secondly, in traditional black culture, much value was placed on “ubuntu” (taking care of each other). Black families thus have always been characterized by a high level of commitment to each other. Everybody contributed towards caring for people with special needs. It can thus be inferred that respondents are used to having disabled or chronically ill people in their households. This fact, in turn, probably reduces the negative impact that the presence of such household members may have on them. The significant link between taking care of a disabled or chronically ill household member and depressive symptomatology does suggest, however, that the presence of such a member in a household increases the st ress and strain experienced by the family, and specifically the adolescent within the family (Makola, 1994). The exact nature of the disabilities and chronic illnesses was not established in the research and respondents were reliant on their own conception of what constitutes a “severe” disability or a “serious chronic” condition. If more specific information had been obtained, different results may have been obtained. The closeness of the participants’ relationships with the disabled/chr onically ill household members; the type of disability/chronic condition; and the level, amount and nature of the care expected from the participants, would also influence the impact thereof on the participants. 80 CONCLUSION An overview of results contained in this article shows that family variables have virtually no significant impact on adolescent depression. Employment status of mother and taking care of a disabled or chronically ill household member were the only two variables that showed statistically significant links with adolescent depression. Although an overview of relevant literature created an expectation that more significant results would be yielded, several factors should be taken into account in order to understand the lack of significant findings in this study. Most other research projects did not investigate a linear relationship between family variables and adolescent depression, but combined several variables in their studies. For example, Brage and Meredith (1994) found that family strengths influence depression indirectly through self -esteem. The relationship between family variables and adolescent depression is apparently more complex, and involves a greater number of other (unidentified) factors, than one would initially expect. Regardless of this, though, the current research still serves to supplement a very meager database on family variables and black adolescents’ psychological well-being. It is important, however, that the abovementioned results should not be interpreted as indicating that family variables play no role in adolescent depression. In fact, it seems that these results point to other factors that should be brought into the equation. Such factors include, amongst others, the adolescents’ perception of the quality of the parental relationship, the effect of daily hassles, the effect of the adolescents’ perception of the family’s socio - economic status and the role of adolescent self-esteem. An important finding that should not be overlooked, is the high depression rate of 26%. Connelly et al. (1993) indicate the inherent pitfalls in discussions on adolescent depression: at one extreme, depression may be considered a normal part of adolescence, with the subsequent result that those at real risk do not receive the attention they need; at the other extreme, all adolescents may be considered as being prone to psychological distress. Even if family variables do not play a major role in adolescent depression, the high depression rate strongly suggests that there are other 81 factors that may be causing adolescent depression. These factors may or may not be linked to family variables. The high depression rate suggests that respondents in this study are experiencing serious problems with adolescent depression. This warrants further investigation to identify those who are at risk and intervene appropriately. Two main role-players in dealing with adolescent depression are the school and parents. Teachers have rightfully been called the “gatekeepers” in terms of identifying adolescents with problems such as depression (cf. Leane & Shute, 1998), as they spend a great deal of time with them. Unfortunately, Leane & Shute (1998) also found that teachers’ knowledge in respect of aspects such as signs of potential suicide in adolescents is ala rmingly low. Teachers should be trained to identify adolescents who are at risk of developing depression, or who are exhibiting signs thereof. They should also be taught how to refer such children appropriately. Marais (1997) found that the implementation of a parental guidance programme decreased the prevalence of adolescent depression. A combination of group counseling for the adolescents and parental guidance for the parents was even more effective in reducing depressive symptomatology. Parents thus play an important role in the treatment of adolescents with depression, even if they are not directly the cause of the depression. In addition, peers and extended family members should also be involved in intervention. The fact that gender differences were not taken into account is probably the most significant shortcoming of the research. This is unfortunate, as most research has established that boys and girls experience depression, as well as aspects of their family life, differently. If gender had been taken into account, more significant results may have been obtained. The respondents in the study were learners in grade 8 to grade 12, thus falling within the age range of approximately 13 years to 19 years. As research has shown that the experience of adolescence varies according to age (cf. Laugesen, Dugas & Bukowski, 2003; Van Wel, Ter Bogt & Raaijmakers, 2002), the study would have benefited if a specific (narrower) age range had been targeted (for example, ages 13 -–15 or ages 17 – 19). A longitudinal study would assist in determining the development process of depression during adolescence and its possible continuation into adulthood. 82 Despite the shortcomings of the project and the apparent lack of significant findings, it is hoped that this study will stimulate further research in order to augment the meager information available on black adolescents. Adolescents are the future of the country, and in fact the world and everything possible should be done to ensure that they reach their maximum potential. To this end, it is vital to identify adolescents at risk for developing mental health problems such as depression. This will enable relevant persons and institutions to deal with it effectively and on a timely basis. 83 LIST OF REFERENCES Amato, P.R. & Keith, B. (1991). Parental divorce and the well-being of children: A meta analysis. Psychological Bulletin , 110(1), 26-46. Andrews, A.B. & Morrison, B. (1997). The changing American family: Impact on South Carolina’s children. Columbia: Institute for Families in Society. Armistead, L., Forehand, R., Beach, S.R.H. & Brody, G.H. (1995). Predicting interpersonal competence in young adulthood: The roles of family, self, and peer systems during adolescence. Journal of Child and Family Studies, 4(4), 445-460. Bach, J.M. (2004). Depression, posttraumatic stress disorder and exposure to violence among Venda and Northern Sotho adolescents. Unpublished master’s dissertation. Bloemfontein: University of the Free State. Berk, L.E. (2003). Child Development (6th ed.). Boston: Pearson Education. Boulter, S.L. (1995). Factors affecting standard six adjustment to high school. Unpublished master’s dissertation. Bellville: University of the Western Cape. Brage, D.G. (1995). Adolescent depression: A review of the literature. Archives of Psychiatric Nursing, 9(1), 45-55. Brage, D. & Meredith, W. (1994). A causal model of adolescent depression. The Journal of Psychology, 128(4), 455-468. Chan, D.W. (1995). Depressive symptoms and coping strategies among Chinese adolescents in Hong Kong. Journal of Youth and Adolescence , 24(3), 267- 279. 84 Cleveland, J, Williams, D & Herman, A.A. (2004, November). Perceived vulnerability to HIV/AIDS and depression in South Africa. Poster presented at the HIV/AIDS International Health Poster Session, Washington DC, USA. Retrieved September 4, 2004, from http://apha.confex.com/apha/132am/techprogram/paper_92694.htm Connelly, B., Johnston, D, Brown, I.D.R., Mackay, S. & Blackstock, E.G. (1993). The prevalence of depression in a high school population. Adolescence, 28(109), 149-158. Retrieved September 14, 2004, from Academic Search Premier database. Corcoran, J. & Franklin, C. (2002). Multi-systemic risk factors predicting depression, self-esteem, and stress in low SES and culturally diverse adolescents. Journal of Human Behavior in the Social Environment, 5(2), 61- 76. Cornwell, B. (2003). The dynamic properties of social support: Decay, growth and staticity and their effects on adolescent depression. Social Forces, 81(3), 953- 978. Cumsille, P.E. & Epstein, N. (1994). Family cohesion, family adaptability, social support and adolescent depressive symptoms in outpatient clinic families. Journal of Family Psychology, 8(2), 202-214. De Visser, J. & le Roux, T. (1996). The experience of teenage pregnancy in Knoppies-laagte. South African Journal of Sociology, 27(3), 98-105. Dori, G.A. & Overholser, J.C. (1999). Depression, hopelessness, and self-esteem: Accounting for suicidality in adolescent psychiatric inpatients. Suicide and Life-threatening Behavior, 29(4), 309-318. Duggal, S., Carlson, E.A., Sroufe, L.A. & Egeland, B. (2001). Depressive symptomatology in childhood and adolescence. Development and Psychopathology, 13(1), 143-164. 85 Dumont, M. & Provost, M.A. (1999). Resilience in adolescents: Protective role of social support, coping strategies, self-esteem, and social activities on experience of stress and depression. Journal of Youth and Adolescence, 28(3), 343-363. Ensink, K., Robertson, A.B., Zissis, C. & Leger, P. (1997). Posttraumatic stress disorder in children exposed to violence. South African Medical Journal, 11, 1526-1530. Fako, T.T. (1996). The definition and functions of the family: Towards a rationale for teaching family values through schools in Botswana. South African Journal of Sociology, 27(1), 10-16. Frey, C.U. & Röthlisberger, C. (1996). Social support in healthy adolescents. Journal of Youth and Adolescence, 25(1), 17-31. Gore, S., Aseltine, R.H. & Colton, M.E. (1992). Social structure, life stress and depressive symptoms in a high school-aged population. Journal of Health and Social Behavior, 33, 97-113. Grossman, M. & Rowat, K.M. (1995). Parental relationships, coping strategies, received support, and well-being in adolescents of separated or divorced, and married parents. Research in Nursing and Health, 18, 249-261. Hawkridge, S.M. (1997). Assessment and treatment of major depression in adolescents. Specialist Medicine, 19(4), 32-41. Hetherington, E.M. (2003). Social support and the adjustment of children in divorced and remarried families [Electronic version]. Childhood, 10(2), 217-236. Hickson, J. & Kriegler, S. (1991). Childshock: The effects of apartheid on the mental health of South Africa’s children. International Journal for the Advancement of Counselling, 14(2), 141-154. 86 Hortacsu, N. (1995). Parents’ education levels, parents’ beliefs, and child outcomes. Journal of Genetic Psychology, 156 (3), 373-383. Retrieved September 14, 2004 from Academic Search Premier database. Houlihan, B., Fitzgerald, M. & O’Regan, M. (1994). Self -esteem, depression and hostility in Irish adolescents. Journal of Adolescence, 17, 565-577. Howell, D.C. (2002). Statistical methods for psychology (5th ed.). Johannesburg: Duxbury. Huysamen, G.K. (1983). Inferensiële statistiek en navorsingsontwerp: ‘n Inleiding. Kaapstad: Academica. Kaltiala -Heino, R., Rimpelä, M., Rantanen, P. & Laippala, P. (2001). Adolescent depression: The role of discontinuities in life course and social support. Journal of Affective Disorders, 64, 155-166. Kim, O & Kim, K. (2001). Body weight, self-esteem and depression in Korean female adolescents. Adolescence, 36(142), 315-322. Retrieved October 6, 2004, from http://www.findarticles.com/p/articles/mi_m2248/is_142_36/ai_79251801/ Lansford, J.E., Ceballo, R., Abbey, A. & Stewart, A.J. (2001). Does family structure matter? A comparison of adoptive, two-parent biological, single -mother, stepfather, and stepmother households. Journal of Marriage and the Family, 63(3), 840-851. Lasko, D.S., Field, T.M., Gonzalez, K.P., Harding, J., Yando, R. & Bendell, B. (1996). Adolescent depressed mood and parental unhappiness. Adolescence, 31(121), 49-57. Retrieved September 14, 2004, from Academic Search Premier database. 87 Laugesen, N., Dugas, M.J. & Bukowski, W.M. (2003, November). The relationship between perceived social support and worry in adolescence. Poster presented at the 37th Annual Convention for the Association for the Advancement of Behavior Therapy, Boston, MA, USA. Leane, W. & Shute, R. (1998). Youth suicide: The knowledge and attitudes of Australian teachers and clergy. Suicide and Life-Threatening Behavior, 28(2), 165-173. Lewinsohn, P.M., Clarke, G.N., Seeley, J.R. & Rohde, P. (1994). Major depression in community adolescents: Age at onset, episode duration and time to recurrence. Journal of the American Academy of Child and Adolescent Psychiatry, 33(6), 809-818. Liem, R. & Liem, J.H. (1988). Psychological effects of unemployment on workers and their families. Journal of Social Issues, 44(4), 87-105. Liu, Y. (2002). The role of perceived social support and dysfunctional attitudes in predicting Taiwanese adolescents’ depressive tendency. Adolescence, 37(148), 823-834. Lopez, F.G. & Thurman, C.W. (1993). High-trait and low -trait angry college students: A comparison of family environments. Journal of Counselin g and Development, 71, 524-527. Louw, D.A., van Ede, D.M. & Ferns, I. (1998). Middle childhood. In D.A. Louw, D.M. van Ede & A.E. Louw (Eds), Human Development (2nd ed.). Cape Town: Kagiso Tertiary. Mahabeer, M. (1992). The influence of age and pa rents’ marital status on Indian children’s perceptions of self, family, teachers and school. South African Journal of Sociology, 23(2), 60-65. 88 Makola, D. (1994). The effects of a chronic disorder on the child, his family and environment. South African Paediatric Medicine, 11 , 24-30. Marais, J.L. (1997). Die opvoedkundige begeleiding van die depressiewe adolessent binne gesins- en skoolverband. Tydskrif vir Geesteswetenskappe, 37 (3), 191- 202. Marcotte, D., Fortin, L., Potvin, P. & Papillon, M. (2002). Gender differences in depressive symptoms during adolescence: Role of gender-typed characteristics, self-esteem, body image, stressful life events, and pubertal status. Journal of Emotional and Behavioral Disorders, 10(1), 29-42. Retrieved October 6, 2004, from http://www.findarticles.com/p/articles/mi_m0FCB/is_1_10/ai_83790824/ Marcotte, G., Marcotte, D. & Bouffard, T. (2002). The influence of familial support and dysfunctional attitudes on depression and delinquency in an adolescent population. European Journal of Psychology of Education, 17(4), 363-376. Mash, E.J. & Wolfe, D.A. (1999). Abnormal child psychology. London: Wadsworth Publishing Company. Mboya, M.M. (1998). Family relations and the self-concept of African adolescents: Gender-related differences. Journal of Comparative Family Studies, 29(1), 201-213. McFarlane, A.H., Bellissimo, A. & Norman, G.R. (1995). The role of family and peers in social self-efficacy: Links to depression in adolescence. American Journal of Orthopsychiatry, 65(3), 402-410. Monck, E., Graham, P., Richman, N. & Dobbs, R. (1994). Adolescent girls I: Self- reported mood disturbance in a community population. British Journal of Psychiatry, 165, 760-769. 89 Nkabinde, Z.P. & Ngwenya, T.P. (1996). Interviews with African parents of children with disabilities in South Africa. Paidonomia, 19(1), 34-42. Nolan, S. (2002). Effect size. Retrieved November 30, 2004, from http://members.aol.com/johnp71/javastat.html Ostrander, R., Weinfurt, K.P. & Nay, W.R. (1998). The role of age, family support and negative cognitions in the prediction of depressive symptoms. School Psychology Review, 27 (1), 121-137. Retrieved September 14, 2004, from Academic Search Premier database. Palosaari, U. & Aro, H. (1994). Effect of timing of parental divorce on the vulnerability of children to depression in young adulthood. Adolescence, 29(115), 681-690. Retrieved September 14, 2004, from MEDLINE database. Patten, C.A., Gillin, J.C., Farkas, A.J., Gilpin, E.A., Berry, C.C. & Pierce, J.P. (1997). Depressive symptoms in California adolescents: Family structure and parental support. Journal of Adolescent Health, 20(4), 271-278. Patton, G.C., Hibbert, M.E., Carlin, J., Shao, Q. Rosier, M., Caust, J. & Bowes, G. (1996). Menarche and the onset of depression and anxiety in Victoria, Australia. Journal of Epidemiology and Community Health, 50 , 661-666. Pelkonen, M., Marttunen, M. & Aro, H. (2003). Risk for depression: A 6-year follow-up of Finnish adolescents. Journal of Affective Disorders, 77, 41-51. Pelser, A. (2002). The social landscape of sexual and reproductive health in South Africa with particular reference to HIV/AIDS. In D. van Rensburg et al. , Strengthening local government and civic responses to the HIV/AIDS epidemic in South Africa. Bloemfontein: Centre for Health Systems Research and Development. 90 Portes, P.R. & Zady, M.F. (2002). Self-esteem in the adaptation of Spanish-speaking adolescents: The role of immigration, family conflict and depression. Hispanic Journal of Behavioral Sciences, 24(3), 296-318. Radloff, L.S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385-401. Radloff, L.S. (1991). The use of the Center for Epidemiologic Studies Depression Scale in adolescents and young adults. Journal of Youth and Adolescence, 20(2), 149-166. Rey, J.M., Sawyer, M.G., Clark, J.J. & Baghurst, P.A. (2001). Depression among Australian adolescents. Medical Journal of Australia, 175 (1), 19-23. Rice, F.P. & Dolgin, K.G. (2005). The adolescent: Development, relationships and culture (11th ed.). Boston: Pearson Education. Richter, L.M. (1989). Household density, family size and the growth and development of black children: A cross-sectional study from infancy to middle childhood. South African Journal of Psychology, 19(4), 191-198. Rodgers, K.B. & Rose, H.A. (2002). Risk and resiliency factors among adolescents who experience marital transitions. Journal of Marriage and the Family, 64(4), 1024-1037. Retrieved September 14, 2004, from Academic Search Premier database. Rushton, J. (2002). Depressive disorders in adolescents: Challenges in diagnosis. Psychiatric Times, 19(9). Retrieved September 15, 2004, from http://www.psychiatrictimes.com/p020921.html Sim, H. (2000). Relationship of daily hassles and social support to depression and antisocial behavior among early adolescents. Journal of Youth and Adolescence, 29 (6), 647-659. 91 Sonn, F.C.T. & Louw, D.A. (1989). Sosiokulturele perspektiewe van abnormale gedrag. In D.A. Louw (Red.), Suid -Afrikaanse Handboek van Abnormale Gedrag. Kaapstad: Southern Boekuitgewers. Southall, D. & Roberts, J.E. (2002). Attributional style and self-esteem in vulnerability to adolescent depressive symptoms following life stress: A 14- week prospective study. Cognitive Therapy and Research, 26(5), 563-579. SPSS Incorporated (1983). SPSS user’s guide. New York: Author. Statistics South Africa (2003). Census 2001: Census in brief (2nd ed.) [Electronic version]. Pretoria: Author. Stewart, E.R., McKenry, P.C., Rudd, N.M. & Gavazzi, S.M. (1994). Family processes as mediators of depressive symptomatology among rural adolescents. Family Relations, 43(1), 38-45. Retrieved September 14, 2004, from Academic Search Premier database. Steyn, A.F. (1993, 1994). Family structures in the RSA. Pretoria: Human Sciences Research Council. Steyn, H.S. (1999). Praktiese beduidendheid: Die gebruik van effekgroottes. Potchefstroom: Publikasiebeheerkomitee, PU vir CHO. Sund, A.M., Larsson, B. & Wichstrøm, L. (2003). Journal of Child Psychology and Psychiatry, 44, 588-597. Thom, D.P. (1988). ‘n Psigologiese ontleding en vergelyking van die blanke en swart adolessent ten opsigte van ouer-k ind verhoudings, identiteitsontwikkeling en die vorming van ‘n beroepsidentiteit. Ongepubliseerde doktorale proefskrif. Pretoria: Universiteit van Suid-Afrika. 92 Van Wel, F., ter Bogt, T. & Raaijmakers, Q. (2002). Changes in the parental bond and the well-being of adolescents and young adults. Adolescence, 37(146), 317-333. Retrieved September 14, 2004, from Academic Search Premier database. Visser, M. (1991). The psychological effects of parents’ unemployment on children. Welfare Focus, 26 (2), 38-41. Vitaro, F., Pelletier, D., Gagnon, C. & Baron, P. (1995). Correlates of depressive symptoms in early adolescence. Journal of Emotional and Behavioral Disorders, 3(4), 241-251. Retrieved September 14, 2004, from Academic Search Premier database. Way, N. & Robinson, M.G. (2003). A longitudinal study of the effects of family, friendsand school experiences on the psychological adjustment of ethnic minority, low -SES adolescents. Journal of Adolescent Research, 18 (4), 324- 346. Wong, E.H. & Wiest, D.J. (1999). Adolescent depression: Links to academic coping and perceived autonomy support. Education, 119(4), 668-673. Retrieved September 14, 2004, from Academic Search Premier database. 93 ARTICLE FOUR THE INFLUENCE OF SELECTED FAMILY VARIABLES ON PERCEPTIONS OF SOCIAL SUPPORT AMONG BLACK ADOLESCENTS 94 TABLE OF CONTENTS INTRODUCTION............................................................................ 99 LITERATURE REVIEW ...............................................................100 METHODOLOGY .........................................................................104 RESULTS AND DISCUSSION.......................................................107 Perceived support from friends ...........................................................................107 Perceived support from family............................................................................120 CONCLUSION...............................................................................131 LIST OF REFERENCES................................................................133 LIST OF TABLES Page Table 1 Distribution of group in terms of family variables 105 Table 2 Items contained in Perceived Social Support – Friends 108 (PPS-Fr) Table 3 Family structures and adolescents’ responses (yes/no) 109 items relating to perceived support from friends Table 4 Parents’ marital status and adolescents’ responses 110 (yes/no) to items relating to perceived support from friends Table 5 Household size and adolescents’ responses (yes/no) to 111 items relating to perceived support from friends Table 6 Fathers’ educational status and adolescents’ responses 112 (yes/no) to items relating to perceived support from friends Table 7 Mothers’ educational status and adolescents’ responses 113 (yes/no) to items relating to perceived support from friends Table 8 Fathers’ employment status and adolescents’ responses 114 (yes/no) to items relating to perceived support from friends Table 9 Mothers’ employment status and adolescents’ responses 115 (yes/no) to items relating to perceived support from friends Table 10 Time spent with father per week and adole scents’ responses 116 (yes/no) to items relating to perceived support from friends Table 11 Time spent with mother per week and adolescents’ responses 117 (yes/no) to items relating to perceived support from friends Table 12 Households taking care of members with severe disabilities 118 or serious chronic illnesses and adolescents’ responses (yes/no) to items relating to perceived support from friends Table 13 Items contained in Perceived Social Support-Family (PSS-Fa) 120 Table 14 Family structures and adolescents’ responses (yes/no) to items 121 relating to perceived support from family Table 15 Parents’ marital status and adolescents’ responses (yes/no) to 122 items relating to perceived support from family Table 16 Household size and adole scents’ responses (yes/no) to items 123 relating to perceived support from family Table 17 Fathers’ educational status and adolescents’ responses (yes/no) 124 to items relating to perceived support from family Table 18 Mothers’ educational status and adolescents’ responses (yes/no) 125 to items relating to perceived support from family Table 19 Fathers’ employment status and adolescents’ responses (yes/no) 126 to items relating to perceived support from family Table 20 Mothers’ employment status and adolescents’ responses (yes/no) 127 to items relating to perceived support from family Table 21 Time spent with father per week and adolescents’ responses 128 (yes/no) to items relating to perceived support from family 97 Table 22 Time spent with mother per week and adolescents’ responses 129 (yes/no) to items relating to perceived support from family Table 23 Households taking care of members with severe disabilities or 130 serious chronic illnesses and adolescents’ responses (yes/no) to items relating to perceived support from family 98 Abstract Social support from family and peers is considered important during adolescence, as it can assist adolescents in making a smooth transition from childhood to adulthood. The relationship between social suppo rt and selected family variables was explored amongst a group of black adolescents in South Africa. Family support appears to be of greater importance than peer support. Results suggest that family size, mothers’ employment status and both parents’ educational status impact on adolescents’ perceptions of peer support. It seems that adolescents experience less perceived support in female-headed extended families. Intact families, employed fathers, and more time spent with both mothers and fathers appear to be linked to higher levels of perceived familial support. Identified trends appear to suggest that what is generally considered to be a stable family environment (e.g. intact family, parental employment and sufficient family time) has a positive impact on adolescents’ well-being. INTRODUCTION Adolescence is generally considered to be a time of transition from childhood to adulthood that involves significant changes in social and emotional development, behaviour and cognitions. As interpersonal relationships influence an individual’s well-being, stable and secure relationships with family and peers can assist adolescents in making a smooth transition (Cornwell, 2003). Wong and Wiest (1999) point out that the adolescent forms part of several social units (for example the family, the peer group and the school) and that these units influence the development and well-being of the adolescent. In addition to internal sources (e.g. self-esteem), external sources (e.g. social support) buffer the adolescent against the effects of negative life events (Cornwell, 2003; Dumont & Provost, 1999). Several authors indicate the importance of social support in promoting psychological well-being and healthy development during adolescence (Liu, 2002; Way & Robinson, 2003). As pointed out by Wan, Jaccard and Ramey (1996), support from both peers and parents is important, because a different kind of support will be offered in each case. Emotional support from parents, for example, is likely to differ in form and content from emotional support from friends. Ostrander, Weinfurt and Nay (1998) consider the family to be essential, both in terms of offering support within the family and within the community. In fact, according to Adendorff (1998), the family forms the basis from which the adolescent explores other social relationships. While 99 Bullock (2002) emphasizes the importance of siblings in the adjustment of children, most researchers emphasize the important role that parental support plays in the overall development of adolescents (Grossman & Rowat, 1995; Lerner, 2002; Lopez & Thurman, 1993; Shek, 2002). Frey and Röthlisberger (1996) highlight the importance of friendships during adolescence in helping the adolescent towards social maturity. More specifically, the peer group can be considered as providing an extended opportunity for the adolescent to develop social skills originally learnt within the family (McFarlane, Bellissimo & Norman, 1995; Rice & Dolgin, 2005). Although existing research clearly indicates the importance of social support during adolescence, hardly any such research has been conducted among black adolescents in South Africa. It was thus decided to investigate the influence of various family variables on perceived social support amongst a group of black adolescents. LITERATURE REVIEW Procidano and Heller (1983) define social support as the “extent to which an individual believes that his/her needs for support, information and feedback are fulfilled” (p. 2). Two prominent aspects of social support are its sources and its functions. Mullis, Hill and Readdick (1999) identify three sources of support, namely family, formal sources (e.g. teachers at school) and informal sources (friends and other adults). Support is offered on four levels (Dumont & Provost, 1999; Wan et al., 1996): emotional (e.g. unconditional acceptance), informational (e.g. the giving of advice), companionship (e.g. a sense of belongingness) and tangible/instrumental (e.g. financial support). The majority of available research on adolescents and social support focused on the role of perceived support from family (especially parents) and peers in the psychological well-being of adolescents. Several researchers point out the importance of having both familial and peer support during adolescence (Kaltiala -Heino, Rimpelä, Rantanen & Laippala, 2001; Liu, 2002; Mullis et al., 1999). According to Way and Robinson (2003), Frey and Röthlisberger (1996) and McFarlane et al. (1995), familial support is more important for the psychological well-being of adolescents than peer support. Researchers have linked 100 low levels of familial support to depression, delinquency, maladaptive behaviour and low levels of life satisfaction (Dunn & Tucker, 1993; Marcotte, Marcotte & Bouffard, 2002; Young, Miller, Norton & Hill, 1995). Parental support appears to be particularly necessary. For example, Patten et al. (1997) found that adolescent depression increases when parents are unsupportive. Murberg & Bru (2004) also found that poor parental support links significantly with emotional problems among adolescents. In addition, Frey and Röthlisberger (1996) and Ohanessian, Lerner, von Eye and Lerner (1996) showed that strong parental support impacts positively on adolescent self -esteem. Sim (2000) reported that high perceived parental support is linked to lower levels of maladjustment and Cornwell (2003) goes as far as to claim that decreases in parental support are somewhat traumatic for the adolescent. In contrast with the above findings, Dumont and Provost (1999) did not find that social support shielded adolescents from negative life events. The researchers speculated that other variables in their study (such as daily trials) were probably more influential than social support. Harris (2000) contends that parental influence plays a minimal, and possibly not even a lasting role , in determining child outcomes, and that existing parental influence is not necessarily carried over to different contexts. In Grossman and Rowat’s (1995) research, friends were ranked as the primary source of support, with mothers second and fathers last. This indicates that peer support does play a role in adolescent adjustment. In fact, Field, Diego and Sanders (2002) found peer attachment to be of greater importance than attachment to parents. A review of available research indicates that family (or parental) support is interwoven with peer support, with the former forming the foundation for the latter. Mullis et al. (1999) suggest that perceived attachment to parents paves the way for attachment to friends. Gore, Aseltine and Colton (1992) illustrate the link between family support and peer support with their finding that adolescents in stepparent families experience their families as less supportive. Moreover, these adolescents also experience more relationship problems with friends. Further to these results, Dekovic and Janssens (1992) determined that popular children have more supportive parents than rejected children. Rejected children have parents who tend to be critical and who lack warmth, and these children have more negative interactions with their peers. McFarlane et al. (1995) consider family support as having a direct buffering effect against depression, whereas peers merely have an indirect buffering effect through 101 other variables such as social self-efficacy. According to Frey and Röthlisberger (1996), the family (especially mothers) provides psychological and instrumental support on a daily basis, as well as during crises, whereas peer support mainly serves a social integration function. Contrary to the above research, neither Way and Robinson (2003) nor Field, Lang, Yando and Bendell (1995) found a significant relationship between the peer group and psychological variables such as self -esteem or depression. Harris (2000) is of the opinion that socialization agents such as peers and parents each wield their influence over children only within their specific contexts. This means that behaviour learnt at home is not necessarily carried over into peer group interaction, and vice versa. In terms of this view parental influence and peer influence stand apart from each other, with no mutual influencing. A further review of research indicates gender differences in terms of adolescents’ experiences of social support. Frey and Röthlisberger (1996), Cumsille and Epstein (1994) and Gore et al. (1992) found that girls receive higher levels of support from their peers than boys do. Liu (2002), on the other hand, indicated that mostly the peer group, but also the family, protected adolescent girls from depression, whereas in the case of boys the peer group appeared to be the only shield. Where the peer group seems to be of greater importance to girls (Colarossi, 2001), family support appears to be more important for boys. Cumsille and Epstein (1994) found a negative correlation between family support and depression in boys and Frey and Röthlisberger (1996) showed that boys are more attached to their families than girls are. In addition, the correlation demons trated by Ohanessian et al. (1996) between perceived maternal acceptance, physical appearance, depression and social competence was also stronger for boys than for girls. Girls probably perceive peer relationships as more important because girls tend to be more relationship -focused, invest more in the nurturing and development of relationships and define themselves in terms of their relationships (Field et al., 1995; Thomas & Daubman, 2001). The fact that boys find cross-gender relationships more rewarding than same-gender relationships (Thomas & Daubman, 2001) is probably owing to the fact that girls invest more in relationships. In terms of parental support most researchers found maternal support to be the most important variable influencing aspects such as depression, self -esteem and even attachment to friends and relatives (Colarossi, 2001; Field et al., 1995; Frey & 102 Röthlisberger, 1996; Harris & Marmer, 1996; Mullis et al., 1999). Despite these findings, the importance of fathers should not be underestimated. Patten et al. (1997) indicated that the adolescent depression rate is higher when there is only maternal support than in the case of both maternal and paternal support. Frey and Röthlisberger (1996) found that fathers are mostly approached for practical help (as opposed to the emotional availability of mothers). Harris and Marmer (1996) illustrated that emotional support from fathers has a protective effect against delinquency in poor families. Shek (2002) found a significant relationship between adolescent psychological well-being and paternal parenthood qualities. He attributes this finding to the nature of the Chinese family, in which the father is perceived as the dominant parenting figure. As far as the age of adolescents is concerned, interesting trends have been observed. In their review of literature, Ostrander et al. (1998) concluded that family support is more important during the earlier stages of adolescence and that its importance lessens through the adolescent years. In support of this finding, Adendorff (1998) found greater levels of support from parents, extended family and friends among younger adolescents. Laugesen, Dugas and Bukowski (2003) consider family support a more important shield against worry and anxiety and depression in early adolescence than peer support. As suggested by Ostrander et al. (1998), a decrease in the importance of familial support probably coincides with an increase in the importance of peer relationships. Van Wel, ter Bogt and Raaijmakers (2002) concluded that the parental bond is less positive during middle adolescence, after which it improves for girls during late adolescence and for boys over a longer period of time. Despite research pointing to the importance of familial support in early adolescence, both Cornwell (2003) and Ostrander et al. (1998) emphasize that family support should increase during adolescence to prevent an increase in depression. According to Cornwell (2003), this also holds true for peer support. Researchers have also investigated the effect of culture and ethnicity on social support. Field et al. (1995) indicated that white and Hispanic adolescents reported greater intimacy with their fathers and friends than black adolescents did. Adendorff (1998) also found less parental support among black adolescents in South Africa. This finding is probably owing to the fact that black families are still affected by 103 migrant labour, where parents work in towns and visit their children over weekends and/or holidays only. Furthermore, many black parents travel long distances between their homes and places of employment, thereby leaving less time for interaction with children. Her research further showed that both black and white South African adolescents considered their relationships with their parents as the greatest source of stress in their lives. Whereas white adolescents placed friends second and school/teachers third on their list, black adolescents placed finances second and friends third on their list. Dunn and Tucke r (1993) found that black children who had a father figure who was present in their lives showed more adaptive functioning. As mentioned earlier, Shek (2002) found the paternal parenting figure to be of greater importance than the maternal parenting figure and attributes this to the dominance of the father figure in Chinese culture. METHODOLOGY The research was non-experimental in nature and a criterion group design (Huysamen, 1983) was employed. The participants in the study were 2 505 black adolescents in grades eight to twelve attending six different high schools in Mangaung, a township near Bloemfontein in the Free State province. Convenience sampling was employed. Permission to conduct the research was granted by the Free State Department of Education. Participants were required to complete a biographical questionnaire that related to various family variables, namely household size, family structure, the employment and educational status of parents, the marital status of parents, and time spent wit h parents. The questionnaire was aimed at collecting similar information to that which was obtained by Steyn (1993, 1994) in South Africa and Andrews and Morrison (1997) in the USA. Participants also indicated whether, in their households, they were taking care of members with severe disabilities or serious chronic illnesses. Participants’ perception of social support was measured by means of Perceived Social Support-Family (PSS-Fa) and Perceived Social Support-Friends (PSS -Fr) (Procidano & Heller, 1983). These two questionnaires, consisting of twenty items each, evaluate 104 the extent to which an individual perceives his family and friends as meeting his needs for feedback, support and information (Marcotte et al., 2002). A higher score indicates higher levels of perceived support. Procidano and Heller (1983) report well- established construct validity for both questionnaires. In order to put the statistical analysis and subsequent results into perspective, the frequencies in terms of the ten selected independent variables are shown in Table 1. Table 1 Distribution of group in terms of family variables Independent variables N % Family structure: Mother and children 374 19,5 Father, mother and children (nuclear) 754 39,2 Male-headed extended family 363 18,9 Female-headed extended family 430 22,4 Subtotal 1921 Marital status: Married 1047 61,4 Never married/Live together 289 16,9 Divorced/Separated/Widowed 371 21,7 Subtotal 1707 Household size: 1 – 3 370 17,3 4 – 6 1399 65,4 7 or more 370 17,3 Subtotal 2139 Educational level – father: None 66 9,9 Grade 1 – 7 87 13,0 Grade 8 – 12 324 48,3 Some form of post-school education 193 28,8 Subtotal 670 105 Table 1 Distribution of group in terms of family variables (continued) Independent variables N % Educational level – mother: None 83 7,4 Grade 1 – 7 134 12,0 Grade 8 – 12 637 56,9 Some form of post-school education 266 23,7 Subtotal 1120 Employment status – father: Employed 801 84,0 Unemployed 153 16,0 Subtotal 954 Employment status – mother: Employed 1081 73,9 Unemployed 382 26,1 Subtotal 1463 Time spent with father per week: Up to 6 hours 545 53,3 More than 6 hours 477 46,7 Subtotal 1022 Time spent with mother per week: Up to 6 hours 402 23,5 More than 6 hours 1312 76,5 Subtotal 1714 Households taking care of members with severe disabilities or serious chronic illnesses: Yes 852 41,8 No 1185 58,2 Subtotal 2037 To measure the relationship between variables the χ²-test for independence was utilized. A significant χ ² -value merely indicates whether two variables are not independent and therefore, the degree of the relationship of variables was also measured (Howell, 2002). To this end Cramér’s phi coefficient was calculated. This calculation supplies a coefficient that lies between 1 and 0. The closer the coefficient is to 1, the stronger the relationship between the two variables. 106 In addition to determining whether results were statistically significant, the practical significance of such results was also investigated. To this end effect sizes were calculated. Effect sizes were determined by means of the following formula (Steyn, 1999): w = √χ² / N The following guidelines were applied in the interpretation of effect sizes: w = 0,1 was considered a small effect, w = 0,3 was considered a medium effect and w = 0,5 was considered a large effect. Effect sizes were only calculated in instances where statistically significant results had been found on either the 1% level or the 5% level. RESULTS AND DISCUSSION Results regarding perceived support from friends will be presented and discussed first, followed by the results regarding perceived support from family. Perceived support from friends The relationship between the ten selected family variables (see Table 1) and perceived social support from friends was investigated via the χ ²-test for independence. In order to facilitate the interpretation of results the twenty items contained in PPS-Fr are shown in Table 2. 107 Table 2 Items contained in Perceived Social Support-Friends (PPS-Fr) Item no. Statement to which yes/no response was given 1 My friends give me the moral support I need 2 Most other people are closer to their friends than I am 3 My friends enjoy hearing about what I think 4 Certain friends come to me when they have problems or need advice 5 I rely on my friends for emotional support 6 If I felt that one or more of my friends were upset with me, I’d just keep it to myself 7 I feel that I am on the fringe (edge) of my circle of friends 8 There is a friend I can go to if I was just feeling down, without feeling funny about it later 9 My friends and I are very open about what we think about things 10 My friends are sensitive about my personal needs 11 My friends come to me for emotional support 12 My friends are good at helping me solve problems 13 I have a deep sharing relationship with a number of friends 14 My friends get good ideas about how to do things or how to make things from me 15 When I confide in friends it makes me feel uncomfortable 16 My friends seek me out for companionship 17 I think that my friends feel that I am good at helping them solve problems 18 I don’t have a relationship with a frien d that is as intimate as other people’s relationships with friends 19 I’ve recently got a good idea about how to do something for a friend 20 I wish my friends were much different The results for family structure and perceived support from friends are indicated in Table 3. 108 Table 3 Family structures and adolescents’ responses (yes/no) to items relating to perceived support from friends Item Mother and children Nuclear family Male- headed extended Female- headed χ² φ w number extended Yes No Yes No Yes No Yes No Item 1 277 (74,9) 93 (25,1) 579 (77,6) 167 (22,4) 283 (78,6) 77 (21,4) 333 (78,2) 93 (21,8) 1,824 0,031 Item 2 185 (49,9) 186 (50,1) 322 (43,4) 420 (56,6) 151 (42,7) 203 (57,3) 187 (44,7) 231 (55,3) 5,098 0,052 Item 3 285 (77,2) 84 (22,8) 613 (82,2) 133 (17,8) 288 (79,8) 73 (20,2) 334 (78,4) 92 (21,6) 4,627 0,049 Item 4 300 (81,1) 70 (18,9) 588 (78,5) 161 (21,5) 284 (78,4) 78 (21,6) 350 (82,0) 77 (18,0) 2,804 0,038 Item 5 195 (53,4) 170 (46,6) 430 (58,2) 309 (41,8) 215 (60,0) 143 (40, 0) 228 (53,9) 195 (46,1) 5,250 0,053 Item 6 121 (33,3) 242 (66,7) 240 (32,2) 506 (67,8) 114 (31,9) 243 (68,1) 146 (34,5) 277 (65,5) 0,861 0,021 Item 7 121 (34,1) 234 (65,9) 226 (31,8) 485 (68,2) 121 (34,6) 229 (65,4) 144 (35,6) 260 (64,4) 1,989 0,033 Item 8 284 (77,4 83 (22,6) 581 (78,4) 160 (21,6) 273 (76,5) 84 (23,5) 320 (76,9) 96 (23,1) 0,649 0,019 Item 9 290 (78,2) 81 (21,8) 612 (81,8) 136 (18,2) 305 (84,7) 55 (15,3) 357 (84,4) 66 (15,6) 7,167 0,061 Item 10 217 (59,1) 150 (40,9) 457 (61,7) 284 (38,3) 224 (62,9) 132 (37,1) 256 (61,2) 162 (38,8) 1,168 0,025 Item 11 273 (74,4) 94 (25,6) 534 (71,9) 209 (28,1) 259 (73,2) 95 (26,8) 313 (74,4) 108 (25,6) 1,212 0,025 Item 12 255 (68,9) 115 (31,1) 547 (73,1) 201 (26,9) 258 (72,5) 98 (27,5) 296 (70,3) 125 (29,7) 2,641 0,037 Item 13 200 (55,3) 162 (44,7) 435 (58,2) 312 (41,8) 199 (55,9) 157 (44,1) 258 (61,3) 163 (38,7) 3,639 0,044 Item 14 262 (71,4) 105 (28,6) 533 (71,5) 212 (28,5) 261 (73,1) 96 (26,9) 303 (72,5) 115 (27,5) 0,412 0,015 Item 15 123 (33,9) 240 (66,1) 280 (38,2) 452 (61,8) 119 (33,2) 239 (66,8) 156 (37,6) 259 (62,4) 3,870 0,045 Item 16 184 (51,4) 174 (48,6) 353 (48,8) 371 (51,2) 177 (50,9) 171 (49,1) 196 (48,5) 208 (51,5) 1,080 0,024 Item 17 287 (78,6) 78 (21,4) 578 (78,6) 157 (21,4) 282 (79,0) 75 (21,0) 316 (76,5) 97 (23,5) 0,943 0,022 Item 18 165 (45,1) 201 (54,9) 371 (50,6) 362 (49,4) 173 (49,9) 174 (50,1) 211 (50,8) 204 (49,2) 3,536 0,044 Item 19 283 (77,1) 84 (22,9) 567 (76,1) 178 (23,9) 266 (74,1) 93 (25,9) 322 (76,9) 97 (23,1) 1,129 0,024 Item 20 221 (59,7) 149 (40,3) 354 (47,3) 394 (52,7) 185 (51,4) 175 (48,6) 206 (48,9) 215 (51,1) 16,05** 0,092 0,09 ** p ≤ 0,01 * p ≤ 0,05 Table 3 indicates a statistically significant relationship (p ≤ 0,01) between family structure and item 20. The corresponding effect size is small. A more significant relationship between family structure and perceived peer support was expected. Gore et al. (1992), for example, found that children in single-parent families generally experience less supportive relationships. As there are often fewer financial resources available in single-parent families (Statistics South Africa [STATSSA], 1998) and less time available to spend with children, one might expect to find that adolescents in such families show an increased need for peer support. The results of the χ ²-test concerning parents’ marital status and perceived support from friends are shown in Table 4. 109 Table 4 Parents’ marital status and adolescents’ responses (yes/no) to items relating to perceived support from friends Item Married Never married/Live Divorced/Separated/ χ² φ together Widowed Yes No Yes No Yes No Item 1 807 (77,5) 234 (22,5) 224 (78,3) 62 (21,7) 289 (79,0) 77 (21,0) 0,352 0,014 Item 2 444 (43,0) 589 (57,0) 130 (45,8) 154 (54,2) 174 (47,8) 190 (52,2) 2,758 0,040 Item 3 834 (80,4) 204 (19,6) 225 (78,4) 62 (21,6) 298 (81,6) 67 (18,4) 1,075 0,025 Item 4 829 (79,6) 213 (20,4) 233 (81,2) 54 (18,8) 292 (79,4) 76 (20,6) 0,425 0,016 Item 5 599 (58,1) 432 (41,9) 161 (56,7) 123 (43,3) 197 (54,0) 168 (46,0) 1,883 0,033 Item 6 325 (31,5) 707 (68,5) 84 (29,7) 199 (70,3) 131 (36,3) 230 (63,7) 3,821 0,048 Item 7 324 (32,7) 668 (67,3) 84 (31,1) 186 (68,9) 126 (35,6) 228 (64,4) 1,561 0,031 Item 8 808 (78,4) 223 (21,6) 206 (73,6) 74 (26,4) 294 (80,8) 70 (19,2) 4,916 0,054 Item 9 867 (83,3) 174 (16,7) 223 (78,0) 63 (22,0) 298 (81,2) 69 (18,8) 4,452 0,051 Item 10 649 (62,9) 383 (37,1) 166 (59,3) 114 (40,7) 222 (61,2) 141 (38,8) 1,323 0,028 Item 11 746 (72,5) 283 (27,5) 210 (73,7) 75 (26,3) 274 (75,1) 91 (24,9) 0,941 0,024 Item 12 759 (73,1) 280 (26,9) 199 (69,8) 86 (30,2) 258 (70,3) 109 (29,7) 1,754 0,032 Item 13 575 (55,8) 456 (44,2) 170 (59,9) 114 (40,1) 228 (62,5) 137 (37,5) 5,486 0,057 Item 14 751 (72,3) 288 (27,7) 206 (72,8) 77 (27,2) 257 (71,2) 104 (28,8) 0,232 0,012 Item 15 380 (37,1) 645 (62,9) 94 (33,7) 185 (66,3) 120 (33,8) 235 (66,2) 1,878 0,034 Item 16 495 (49,2) 510 (50,8) 148 (54,0) 126 (46,0) 168 (47,5) 186 (52,5) 2,832 0,042 Item 17 800 (78,1) 225 (21,9) 223 (80,2) 55 (19,8) 277 (76,5) 85 (23,5) 1,257 0,027 Item 18 501 (49,1) 520 (50,9) 143 (51,1) 137 (48,9) 168 (46,8) 191 (53,2) 1,176 0,027 Item 19 778 (75,3) 255 (24,7) 215 (75,4) 70 (24,6) 291 (79,7) 74 (20,3) 3,040 0,043 Item 20 526 (50,6) 513 (49,4) 143 (50,5) 140 (49,5) 203 (55,3) 164 (44,7) 2,551 0,039 ** p ≤ 0,01 * p ≤ 0,05 Table 4 shows that there is no statistically significant relationship between the marital status of parents and perceived support from friends. The intactness or non-intactness of the family does not influence the adolescent’s perceived support from friends. This lack of a significant finding suggests that there may be other factors influencing perceptions of peer support among the adolescents in this study. Factors such as the quality of the parental relationship (e.g. the degree of spousal conflict) and parenting style may be more important in determining the quality of friendships than the actual status of the family (Arnett, 2001; Dekovic & Janssens, 1992; Grossman & Row at, 1995). The results of the χ ²-test for household size and perceived support from friends are shown in Table 5. 110 Table 5 Household size and adolescents’ responses (yes/no) to items relating to perceived support from friends Item 1 – 3 4 – 6 7 or more χ² φ w Yes No Yes No Yes No Item 1 276 (75,2) 91 (24,8) 1074 (77,6) 310 (22,4) 283 (77,1) 84 (22,9) 0,944 0,021 Item 2 181 (49,7) 183 (50,3) 608 (44,4) 761 (55,6) 157 (43,0) 208 (57,0) 4,049 0,044 Item 3 287 (78,4) 79 (21,6) 1111 (80,5) 269 (19, 5) 287 (77,8) 82 (22,2) 1,769 0,029 Item 4 307 (83,7) 60 (16,3) 1101 (79,5) 284 (20,5) 292 (78,9) 78 (21,1) 3,547 0,041 Item 5 192 (52,9) 171 (47,1) 772 (56,4) 598 (43,6) 226 (61,8) 140 (38,2) 6,011* 0,054 0,05 Item 6 123 (33,8) 241 (66,2) 453 (33,0) 918 (67,0) 128 (34,9) 239 (65,1) 0,456 0,015 Item 7 129 (36,1) 228 (63,9) 437 (33,2) 877 (66,8) 122 (34,4) 233 (65,6) 1,068 0,023 Item 8 284 (78,7) 77 (21,3) 1061 (77,1) 315 (22,9) 271 (75,3) 89 (24,7) 1,178 0,024 Item 9 292 (79,6) 75 (20,4) 1139 (82,4) 244 (17,6) 307 (83,4) 61 (16,6) 2,101 0,031 Item 10 221 (61,1) 141 (38,9) 838 (61,1) 534 (38,9) 228 (63,3) 132 (36,7) 0,643 0,018 Item 11 257 (71,2) 104 (28,8) 995 (72,6) 376 (27,4) 273 (74,6) 93 (25,4) 1,083 0,023 Item 12 265 (72,4) 101 (27,6) 986 (71,4) 396 (28,6) 258 (71,5) 103 (28,5) 0,161 0,009 Item 13 207 (56,4) 160 (43,6) 796 (58,2) 572 (41,8) 208 (57,1) 156 (42,9) 0,432 0,014 Item 14 254 (69,8) 110 (30,2) 1007 (73,5) 364 (26,5) 256 (70,1) 109 (29,9) 2,904 0,037 Item 15 137 (37,7) 226 (62,3) 507 (37,4) 847 (62,6) 126 (34,7) 237 (65,3) 1,016 0,022 Item 16 180 (50,6) 176 (49,4) 653 (48,8) 685 (51,2) 178 (50,6) 174 (49,4) 0,574 0,017 Item 17 278 (77,4) 81 (22,6) 1064 (78,1) 299 (21,9) 286 (79,7) 73 (20,3) 0,589 0,017 Item 18 178 (49,2) 184 (50,8) 661 (48,8) 693 (51,2) 184 (51,7) 172 (48,3) 0,934 0,021 Item 19 289 (79,6) 74 (20,4) 1036 (75,5) 336 (24,5) 279 (76,4) 86 (23,6) 2,681 0,036 Item 20 197 (54,1) 167 (45,9) 698 (50,6) 682 (49,4) 195 (53,3) 171 (46,7) 1,912 0,030 ** p ≤ 0,01 * p ≤ 0,05 Table 5 indicates a statistically significant relationship (p ≤ 0,05) between household size and item 5 of PSS-Fr. The corresponding effect size is small. The results suggest a slight possibility that adolescents in larger families experience less emotional support from their families. Parents in particular may be required to divide their attention among several household members in the case of larger households. They may also feel overwhelmed by the responsibility of caring for a large household thereby providing less support to their children. Adolescents thus possibly turn to their friends for support. Peer support probably mediates the effect of insufficient familial and/or parental support (Cumsille & Epstein, 1994; Rodgers & Rose, 2002). The results of the respective χ²-tests for fathers’ educational status (Table 6) and mothers’ educational status (Table 7), and perceived support from friends, will be presented and discussed together. 111 Table 6 Fathers’ educational status an d adolescents’ responses (yes/no) to items relating to perceived support from friends Item None Grade 1-7 Grade 8 – 12 Post-school χ² φ w Yes No Yes No Yes No Yes No Item 1 47 (71,2) 19 (28,8) 69 (80,2) 17 (19,8) 249 (77,3) 73 (22,7) 156 (81,7) 35 (18,3) 3,605 0,073 Item 2 37 (56,1) 29 (43,9) 44 (52,4) 40 (47,6) 134 (42,0) 185 (58,0) 81 (42,4) 110 (57,6) 6,783 0,101 Item 3 52 (78,8) 14 (21,2) 61 (71,8) 24 (28,2) 272 (85,0) 48 (15,0) 152 (78,8) 41 (21,2) 8,827* 0,115 0,12 Item 4 55 (84,6) 10 (15,4) 67 (77,0) 20 (23,0) 249 (77,1) 74 (22,9) 148 (77,1) 44 (22,9) 1,931 0,054 Item 5 40 (60,6) 26 (39,4) 47 (55,3) 38 (44,7) 191 (59,7) 129 (40,3) 121 (63,4) 70 (36,6) 1,684 0,050 Item 6 19 (29,2) 46 (70,8) 34 (39,5) 52 (60,5) 94 (29,6) 223 (70,4) 64 (33,5) 127 (66,5) 3,474 0,073 Item 7 22 (33,3) 44 (66,7) 32 (39,5) 49 (60,5) 107 (34,6) 202 (65,4) 58 (31,0) 129 (69,0) 1,901 0,054 Item 8 57 (89,1) 7 (10,9) 63 (74,1) 22 (25,9) 252 (79,0) 67 (21,0) 144 (75,8) 46 (24,2) 6,053 0,095 Item 9 52 (78,7) 14 (21,2) 67 (77,0) 20 (23,0) 275 (85,7) 46 (14,3) 161 (83,9) 31 (16,1) 4,784 0,084 Item 10 42 (63,6) 24 (36,4) 56 (64,4) 31 (35,6) 209 (65,5) 110 (34,5) 108 (57,1) 81 (42,9) 3,708 0,075 Item 11 44 (67,7) 21 (32,3) 60 (71,4) 24 (28,6) 229 (72,5) 87 (27,5) 141 (73,1) 52 (26,9) 0,753 0,034 Item 12 51 (77,3) 15 (22,7) 63 (75,0) 21 (25,0) 239 (74,5) 82 (25.,5) 148 (76,7) 45 (23,3) 0,458 0,026 Item 13 27 (40,9) 39 (59,1) 42 (48,8) 44 (51,2) 184 (57,7) 135 (42,3) 119 (62,0) 73 (38,0) 11,045* 0,128 0,13 Item 14 53 (81,5) 12 (18,5) 63 (74,1) 22 (25,9) 238 (74,6) 81 (25,4) 130 (67,4) 63 (32,6) 6,006 0,095 Item 15 22 (33,3) 44 (66,7) 32 (38,1) 52 (61,9) 118 (37,8) 194 (62,2) 58 (30,4) 133 (69,6) 3,304 0,071 Item 16 28 (42,4) 38 (57,6) 37 (44,6) 46 (55,4) 152 (48,9) 159 (51,1) 94 (49,7) 95 (50,3) 1,533 0,049 Item 17 54 (83,1) 11 (16,9) 70 (82,4) 15 (17,6) 241 (77,5) 70 (22,5) 151 (79,5) 39 (20,5) 1,668 0,051 Item 18 32 (49,2) 33 (50,8) 43 (51,2) 41 (48,8) 150 (47,9) 163 (52,1) 95 (49,7) 96 (50,3) 0,350 0,023 Item 19 48 (72,7) 18 (27,3) 64 (77,1) 19 (22,9) 255 (79,7) 65 (20,3) 144 (74,6) 49 (25,4) 2,609 0,063 Item 20 36 (54,6) 30 (45,4) 56 (65,1) 30 (34,9) 154 (48,1) 166 (51,9) 89 (46,3) 103 (53,7) 9,823 0,121 ** p ≤ 0,01 * p ≤ 0,05 Table 6 indicates a significant relationship (p ≤ 0,05) between the educational status of fathers and items 3 and 13 of PSS-Fr. The corresponding effect sizes are small. 112 Table 7 Mothers’ educational status and adolescents’ responses (yes/no) to items relating to perceived support from friends Item None Grade 1 – 7 Grade 8 – 12 Post-school χ² φ W Yes No Yes No Yes No Yes No Item 1 60 (72,3) 23 (27,7) 97 (73,5) 35 (26,5) 503 (79,8) 127 (20,2) 203 (76,6) 62 (23,4) 4,612 0,064 Item 2 50 (61,7) 31 (38,3) 64 (48,8) 67 (51,2) 267 (42,7) 358 (57,3) 109 (41,4) 154 (58,6) 12,533** 0,107 0,11 Item 3 64 (79,0) 17 (21,0) 96 (72,2) 37 (27,8) 527 (83,5) 104 (16,5) 205 (77,4) 60 (22,6) 11,232* 0,101 0,10 Item 4 70 (85,4) 12 (14,6) 104 (78,8) 28 (21,2) 508 (80,0) 127 (20,0) 201 (75,9) 64 (24,1) 3,965 0,060 Item 5 51 (61,5) 32 (38,5) 68 (52,3) 62 (47,7) 379 (60,8) 244 (39,2) 153 (58,0) 111 (42,0) 3,588 0,057 Item 6 29 (35,8) 52 (64,2) 48 (36,9) 82 (63,1) 189 (30,3) 434 (69,7) 89 (33,8) 174 (66,2) 3,103 0,053 Item 7 32 (39,0) 50 (61,0) 57 (44,2) 72 (55,8) 192 (32,3) 403 (67,7) 79 (30,9) 177 (69,1) 8,815* 0,091 0,09 Item 8 66 (81,5) 15 (18,5) 95 (73,6) 34 (26,4) 490 (78,3) 136 (21,7) 205 (77,4) 60 (22,6) 2,032 0,043 Item 9 63 (75,9) 20 (24,1) 99 (74,4) 34 (25,6) 535 (84,5) 98 (15,5) 217 (82,2) 47 (17,8) 10,013* 0,094 0,10 Item 10 49 (59,0) 34 (41,0) 89 (67,4) 43 (32,6) 413 (66,0) 213 (34,0) 147 (56,1) 115 (43,9) 9,377* 0,092 0,10 Item 11 62 (74,7) 21 (25,3) 98 (74,2) 34 (25,8) 466 (74,8) 157 (25,2) 183 (69,3) 81 (30,7) 2,998 0,052 Item 12 63 (75,9) 20 (24,1) 98 (74,2) 34 (25,8) 470 (74,7) 159 (25,3) 182 (68,7) 83 (31,3) 3,889 0,059 Item 13 39 (47,0) 44 (53,0) 70 (53,4) 61 (46,6) 355 (56,8) 270 (43,2) 167 (63,0) 98 (37,0) 7,997* 0,085 0,09 Item 14 58 (69,9) 25 (30,1) 97 (73,5) 35 (26,5) 467 (74,7) 158 (25,3) 180 (68,2) 84 (31,8) 4,367 0,063 Item 15 26 (31,7) 56 (68,3) 49 (38,3) 79 (61,7) 209 (33,7) 411 (66,3) 89 (34,0) 173 (66,0) 1,246 0,034 Item 16 41 (50,6) 40 (49,4) 59 (47,6) 65 (52,4) 307 (50,4) 302 (49,6) 137 (52,5) 124 (47,5) 0,834 0,028 Item 17 68 (82,9) 14 (17,1) 107 (83,0) 22 (17,0) 483 (78,2) 135 (21,8) 197 (75,5) 64 (24,5) 3,908 0,060 Item 18 42 (52,5) 38 (47,5) 76 (57,6) 56 (42,4) 289 (46,6) 331 (53,4) 118 (45,4) 142 (54,6) 6,685 0,078 Item 19 63 (76,8) 19 (23,2) 99 (76,7) 30 (23,3) 487 (77,6) 141 (22,4) 193 (73,1) 71 (26,9) 2,058 0,043 Item 20 55 (66,3) 28 (33,7) 87 (65,9) 45 (34,1) 311 (49,3) 320 (50,7) 129 (49,0) 134 (51,0) 19,691** 0,133 0,13 ** p ≤ 0,01 * p ≤ 0,05 Table 7 indicates a significant relationship between the educational status of adolescents’ mothers and perceived support from friends for items 2 and 20 (p ≤ 0,01) and for items 3, 7, 9, 10 and 13 (p ≤ 0,05). For all the above seven items the calculated effect sizes are small. Results contained in Table 6 and Table 7 seem to suggest that the educational status of both fathers and mothers impacts on perceived social support from friends. Hortacsu (1995) argues that the educational status of parents influences the way in which parents relate to their children and that this in turn will influence the way in which children relate to their friends. In his study, however, mothers appeared to carry more weight than fathers. Gore et al. (1992), on the other hand, found no relation between parental education and measures of support. A parent with a good education may be a source of self-esteem for an adolescent, which in turn may enhance the quality of his friendships. This may be particularly true in African 113 culture, where the father is often still viewed as the dominant parent and where the importance of education has been emphasized, especially during the last few decades. Table 8 indicates the results of the χ²-test for fathers’ employment status and perceived support from friends. Table 8 Fathers’ employment status and adolescents’ responses (yes/no) to items relating to perceived support from friends Item Employed Unemployed χ² φ Yes No Yes No Item 1 625 (78,8) 168 (21,2) 120 (79,0) 32 (21,0) 0,001 -0,001 Item 2 340 (43,0) 450 (57,0) 66 (44,0) 84 (56,0) 0,048 -0,007 Item 3 645 (81,1) 150 (18,9) 130 (86,7) 20 (13,3) 2,620 -0,053 Item 4 619 (77,6) 179 (22,4) 122 (80,8) 29 (19,2) 0,772 -0,029 Item 5 463 (58,8) 325 (41,2) 87 (57,2) 65 (42,8) 0,121 0,011 Item 6 248 (31,3) 544 (68,7) 49 (33,1) 99 (66,9) 0,186 -0,014 Item 7 235 (30,8) 527 (69,2) 53 (36,8) 91 (63,2) 1,988 -0,047 Item 8 610 (77,2) 180 (22,8) 126 (84,0) 24 (16,0) 3,415 -0,060 Item 9 665 (83,5) 131 (16,5) 125 (82,8) 26 (17,2) 0,053 0,007 Item 10 491 (62,3) 297 (37,7) 98 (65,3) 52 (34,7) 0,493 -0,023 Item 11 574 (72,8) 214 (27,2) 104 (69,8) 45 (30,2) 0,581 0,025 Item 12 584 (73,6) 210 (26,4) 118 (78,2) 33 (21,8) 1,402 -0,039 Item 13 467 (59,0) 325 (41,0) 80 (53,7) 69 (46,3) 1,433 0,039 Item 14 578 (72,7) 217 (27,3) 115 (77,2) 34 (22,8) 1,288 -0,037 Item 15 289 (36,9) 494 (63,1) 51 (34,9) 95 (65,1) 0,207 0,015 Item 16 380 (49,2) 393 (50,8) 73 (50,3) 72 (49,7) 0,069 -0,009 Item 17 622 (79,5) 160 (20,5) 119 (81,0) 28 (19,0) 0,153 -0,013 Item 18 395 (50,7) 384 (49,3) 74 (49,7) 75 (50,3) 0,054 0,008 Item 19 610 (76,7) 185 (23,3) 116 (77,3) 34 (22,7) 0,026 -0,005 Item 20 386 (48,6) 408 (51,4) 75 (49,7) 76 (50,3) 0,056 -0,008 ** p ≤ 0,01 * p ≤ 0,05 The results contained in Table 8 indicate that there is no significant relationship, at least on the 5% level, between fathers’ employment status and adolescents’ perception of support from friends. Contrary to these findings, Visser (1991) reported that South African children with unemployed fathers viewed their fathers as less available and tended to withdraw more from their social group. Visser’s findings suggest the possibility of a relationship between paternal unemployment and perceived support from friends. A lack of significant findings in the present study suggests that other factors are in operation. It is possible, for example, that the availability and support 114 provided by mothers compensate for the loss of paternal support, as mothers have been repeatedly mentioned in the relevant literature as the dominant source of support (Colarossi, 2001; Frey & Röthlisberger, 1996; Harris & Marmer, 1996). Results of the χ²-test for mothers’ employment status and perceived support from friends are indicated in Table 9. Table 9 Mothers’ employment status and adolescents’ responses (yes/no) to items relating to perceived support from friends Item Employed Unemployed χ² φ w Yes No Yes No Item 1 839 (78,3) 232 (21,7) 295 (78,0) 83 (22,0) 0,014 0,003 Item 2 471 (44,1) 596 (55,9) 160 (43,2) 210 (56,8) 0,090 0,008 Item 3 849 (79,2) 223 (20,8) 310 (81,8) 69 (18,2) 1,174 -0,028 Item 4 849 (78,9) 227 (21,1) 309 (81,8) 69 (18,2) 1,394 -0,031 Item 5 618 (58,1) 445 (41,9) 217 (57,6) 160 (42,4) 0,038 0,005 Item 6 341 (32,1) 720 (67,9) 122 (32,7) 251 (67,3) 0,041 -0,005 Item 7 342 (33,1) 690 (66,9) 111 (31,6) 240 (68,4) 0,273 0,014 Item 8 818 (77,2) 242 (22,8) 306 (81,8) 68 (18,2) 3,525 -0,050 Item 9 883 (82,4) 189 (17,6) 304 (80,0) 76 (20,0) 1,056 0,027 Item 10 676 (63,7) 385 (36,3) 214 (57,7) 157 (42,3) 4,251* 0,054 0,06 Item 11 797 (74,8) 268 (25,2) 255 (67,8) 121 (32,2) 6,942** 0,069 0,07 Item 12 772 (72,0) 301 (28,0) 264 (70,6) 110 (29,4) 0,252 0,013 Item 13 637 (59,8) 428 (40,2) 202 (54,0) 172 (46,0) 3,832* 0,052 0,05 Item 14 755 (70,7) 313 (29,3) 284 (76,3) 88 (23,7) 4,385* -0,055 0,06 Item 15 372 (35,3) 682 (64,7) 142 (38,4) 228 (61,6) 1,129 -0,028 Item 16 523 (50,4) 514 (49,6) 172 (47,6) 189 (53,4) 0,833 0,024 Item 17 822 (77,8) 234 (22,2) 294 (79,5) 76 (20,5) 0,422 -0,017 Item 18 515 (48,9) 538 (51,1) 183 (49,6) 186 (50,4) 0,051 -0,006 Item 19 821 (76,9) 246 (23,1) 279 (74,4) 96 (25,6) 0,993 0,026 Item 20 539 (50,3) 532 (49,7) 213 (56,4) 165 (43,6) 4,059* -0,053 0,05 ** p ≤ 0,01 * p ≤ 0,05 From Table 9 it is evident that there are significant relationships between the employment status of mothers and perceived support from friends for item 11 (p ≤ 0,01) and for items 10, 13, 14 and 20 (p ≤ 0,05). Corresponding effect sizes are small. The finding suggests an inclination towards a positive correlation between employed mothers and perceived support from friends. It may be that children of employed mothers have more self-confidence, thereby enhancing overall friendship quality. 115 The results of the χ²-test for time spent with fathers and perceptions of support from friends are shown in table 10. Table 10 Time spent with father per week and adolescents’ responses (yes/no) to items relating to perceived support from friends Item Up to 6 hours More than 6 hours χ² φ W Yes No Yes No Item1 409 (75,5) 133 (24,5) 370 (78,2) 103 (21,8) 1,080 -0,033 Item 2 241 (44,7) 298 (55,3) 207 (44,2) 261 (55,8) 0,024 0,005 Item 3 442 (82,2) 96 (17,8) 377 (79,5) 97 (20,5) 1,121 0,033 Item 4 436 (80,2) 108 (19,8) 377 (79,5) 97 (20,5) 0,059 0,008 Item 5 300 (55,8) 238 (44,2) 283 (60,6) 184 (39,4) 2,402 -0,049 Item 6 164 (30,4) 376 (69,6) 144 (30,9) 322 (69,1) 0,033 -0,006 Item 7 152 (29,4) 365 (70,6) 169 (37,2) 285 (62,8) 6,687** -0,083 0,08 Item 8 409 (76,2) 128 (23,8) 365 (77,7) 105 (22,3) 0,315 -0,018 Item 9 443 (81,7) 99 (18,3) 406 (85,5) 69 (14,5) 2,567 -0,050 Item 10 332 (61,9) 204 (38,1) 294 (62,6) 176 (37,4) 0,040 -0,006 Item 11 391 (73,0) 145 (27,0) 328 (69,9) 141 (30,1) 1,114 0,033 Item 12 394 (72,8) 147 (27,2) 352 (74,3) 122 (25,7) 0,267 -0,016 Item 13 307 (57,0) 232 (43,0) 268 (57,0) 202 (43,0) 0,000 -0,001 Item 14 398 (73,3) 145 (26,7) 335 (71,1) 136 (28,9) 0,594 0,024 Item 15 172 (32,1) 363 (67,9) 191 (40,9) 276 (59,1) 8,263** -0,091 0,09 Item 16 263 (49,3) 270 (60,7) 230 (50,4) 226 (49,6) 0,118 -0,011 Item 17 419 (78,2) 117 (21,8) 367 (79,3) 96 (20,7) 0,177 -0,013 Item 18 270 (50,8) 262 (49,2) 230 (49,9) 231 (50,1) 0,073 0,009 Item 19 397 (73,5) 143 (26,5) 364 (77,5) 106 (22,5) 2,088 -0,045 Item 20 260 (48,1) 280 (51,9) 221 (46,6) 253 (53,4) 0,235 0,015 ** p ≤ 0,01 * p ≤ 0,05 According to Table 10 there is a significant relationship (p ≤ 0,01) between time spent with fathers and perceived support from friends for items 7 and 15. Calculated effect sizes are small. A surprising finding is that more time spent with fathers is linked to less perceived support from friends. A hypothetical explanation is that adolescents have less need for friendships and peer support when fathers are more available. It should also be borne in mind that black culture is traditionally collectivistic in nature, as opposed to typically Westernized individualistic culture. Familial and parental interaction may therefore to a large extent satisfy the adolescents’ need for support. 116 Table 11 indicates the χ²-test results for time spent with mothers and perceived support from friends. Table 11 Time spent with mother per week and adolescents’ responses (yes/no) to items relating to perceived support from friends Item Up to 6 hours More than 6 hours χ² φ Yes No Yes No Item 1 302 (75,5) 98 (24,5) 1020 (78,5) 280 (21,5) 1,552 -0,030 Item 2 182 (45,7) 216 (54,3) 577 (44,8) 711 (55,2) 0,106 0,008 Item 3 317 (79,5) 82 (20,5) 1047 (80,5) 253 (19,5) 0,229 -0,012 Item 4 317 (78,9) 85 (21,1) 1056 (81,1) 246 (18,9) 0,994 -0,024 Item 5 220 (55,1) 179 (44,9) 746 (58,0) 540 (42,0) 1,026 -0,025 Item 6 130 (32,8) 266 (67,2) 402 (31,1) 889 (68,9) 0,401 0,015 Item 7 125 (32,2) 263 (67,8) 424 (34,3) 813 (65,7) 0,560 -0,019 Item 8 307 (77,1) 91 (22,9) 1009 (78,2) 281 (21,8) 0,207 -0,011 Item 9 319 (80,0) 80 (20,0) 1080 (82,8) 224 (17,2) 1,719 -0,032 Item 10 251 (63,4) 145 (36,6) 795 (61,8) 491 (38,2) 0,315 0,014 Item 11 275 (69,3) 122 (30,7) 946 (73,4) 343 (26,6) 2,580 -0,039 Item 12 289 (72,4) 110 (27,6) 933 (72,2) 360 (27,8) 0,011 0,003 Item 13 224 (56,3) 174 (43,7) 760 (59,1) 527 (409) 0,961 -0,024 Item 14 299 (74,9) 100 (25,1) 933 (72,3) 357 (27,7) 1,053 0,025 Item 15 125 (32,0) 265 (68,0) 478 (37,3) 802 (62,7) 3,629 -0,047 Item 16 183 (46,4) 211 (53,6) 636 (50,6) 620 (49,4) 2,107 -0,036 Item 17 301 (75,4) 98 (24,6) 1009 (79,0) 268 (21,0) 2,276 -0,037 Item 18 192 (49,1) 199 (50,9) 616 (48,4) 657 (51,6) 0,061 0,006 Item 19 304 (77,4) 89 (22,6) 989 (76,3) 308 (23,7) 0,203 0,011 Item 20 201 (50,4) 198 (49,6) 657 (50,7) 639 (49,3) 0,012 -0,003 ** p ≤ 0,01 * p ≤ 0,05 As is clearly indicated by Table 11, no significant relationship was found between time spent with mothers per week and adolescents’ perceptions in respect of support from friends. As the relevant literature has indicated mothers as the major source of support (Colarossi, 2001; Harris & Marmer, 1996), one would have expected time spent with mothers to influence perceived support from friends. Furthermore, Updegraff, Mchale, Grouter and Kupanoff (2001) found that mothers are more well- informed about their adolescents’ peer relationships. However, their findings revealed differences in mothers’ involvement in the peer relationships of boys and girls respectively. The lack of a significant finding in terms of time spent with parents in this study, may therefore be owing to the fact that gender was not taken into account. 117 The last family variable that was investigated for a significant relationship with perceived support from friends, was that of taking care of household members with severe disabilities or serious chronic illnesses. Table 12 summarizes the χ²-test results for these variables. Table 12 Households taking care of members with severe disabilities or serious chronic illnesses and adolescents’ responses (yes/no) to items relating to perceived support from friends Item Taking care of Not taking care of χ² φ W disabled/chronically ill disabled/chronically ill household members household members Yes No Yes No Item 1 645 (76,2) 202 (23,8) 912 (77,8) 261 (22,2) 0,711 -0,019 Item 2 392 (47,3) 436 (52,7) 504 (43,0) 668 (57,0) 3,695 0,043 Item 3 663 (78,7) 179 (21,3) 947 (80,7) 227 (19,3) 1,128 -0,024 Item 4 677 (80,1) 168 (19,9) 942 (80,0) 236 (20,0) 0,007 0,002 Item 5 493 (58,8) 346 (41,2) 639 (54,9) 526 (45,1) 3,035 0,039 Item 6 290 (34,9) 541 (65,1) 381 (32,4) 793 (67,6) 1,306 0,026 Item 7 296 (36,4) 516 (63,6) 363 (32,3) 760 (67,7) 3,578 0,043 Item 8 642 (77,0) 192 (23,0) 909 (77,8) 260 (22,2) 0,170 -0,009 Item 9 683 (81,1) 159 (18,9) 970 (82,5) 206 (17,5) 0,619 -0,018 Item 10 541 (64,8) 294 (35,2) 692 (59,6) 469 (40,4) 5,533* 0,053 0,05 Item 11 618 (74,1) 216 (25,9) 838 (71,9) 328 (28,1) 1,222 0,025 Item 12 612 (73,0) 226 (27,0) 830 (70,8) 343 (29,2) 1,244 0,025 Item 13 488 (58,7) 344 (41,3) 675 (57,7) 495 (42,3) 0,185 0,010 Item 14 622 (74,0) 219 (26,0) 822 (70,6) 342 (29,4) 2,705 0,037 Item 15 342 (41,6) 480 (58,4) 395 (33,9) 769 (66,1) 12,148** 0,078 0,08 Item 16 385 (47,2) 430 (52,8) 575 (50,4) 565 (49,6) 1,946 -0,032 Item 17 666 (80,2) 164 (19,8) 885 (76,5) 272 (23,5) 3,968* 0,045 0,04 Item 18 432 (52,6) 389 (47,4) 544 (47,1) 610 (52,9) 5,760* 0,054 0,05 Item 19 655 (78,4) 181 (21,6) 876 (75,1) 291 (24,9) 2,918 0,038 Item 20 488 (58,0) 353 (42,0) 542 (46,2) 630 (53,8) 27,196** 0,116 0,12 ** p ≤ 0,01 * p ≤ 0,05 Table 12 shows significant relationships between perceived support from friends and taking care of disabled or chronically ill household members for items 15 and 20 (p ≤ 0,01) and for items 10, 17 and 18 (p ≤ 0,05). Corresponding effect sizes are small. The results do not illustrate a clear trend towards either a greater amount, or a smaller amount, of perceived peer support. Black people in South Africa have been marginalized in terms of access to special services for people with chronic illnesses 118 and disabilities. The burden of care rested on the community and mostly the family. Black children are thus probably used to the presence of disabled or chronically ill household members (cf. Senel & Akkök, 1996). In interpreting the effects of disability and chronic illness on black adolescents, two prominent aspects need to be incorporated. Firstly, the understanding and impact of disabilit y and chronic illness can be expected to differ across cultures (cf. Harry, 2002). It is therefore possible that issues such as stigmatization, for example, are not a problem among black adolescents in this study and their friends. Secondly, aspects such as the amount and type of care required from the adolescent will also determine how the adolescent’s peer group relationships are affected. 119 Perceived support from family The relationship between the ten family variables (see Table 1) and perceived support from family was subsequently investigated. The results will now be presented and discussed. To facilitate the interpretation of results the items of PSS-Fa to which respondents had to give a “yes” or a “no” response are presented in Table 13. Table 13 Items contained in Perceived Social Support-Family (PSS-Fa) Item Statements to which respondents gave a yes or no response nr 1 My family gives me the moral support I need 2 I get good ideas about how to do things or make things from my family 3 Most other people are closer to their family than I am 4 When I confide in members of my family who are closest to me, I get the idea that it makes them uncomfortable 5 My family enjoys hearing about what I think 6 Members of my family share many of my interests 7 Certain members of my family come to me when they have problems or need advice 8 I rely on my family for emotional support 9 There is a member of my family I could go to if I was just feeling down, without feeling funny about it later 10 My family and I are very open about what we think about things 11 My family is sensitive about my personal needs 12 Members of my family come to me for emotional support 13 Members of my family are good at helping me solve problems 14 I have a deep sharing relationship with a number of members of my family 15 Members of my family get good ideas about how to do things or make things from me 16 When I confide in members of my family it makes me feel uncomfortable 17 Members of my family seek me out for companionship 18 I think that my family feels that I am good at helping them solve problems 19 I don’t have a relationship with a member of my family that is as close as other people’s relationships with family members 20 I wish my family were much different 120 Table 14 indicates the results of the χ²-test for family structures and perceived support from family. Table 14 Family structures and adolescents’ responses (yes/no) to items relating to perceived support from family Item Mother and children Nuclear family Male- headed, extended Female- headed, χ² φ w extended Yes No Yes No Yes No Yes No Item 1 332 (89,5) 39 (10,5) 670 (89,5) 79 (10,5) 309 (85,8) 51 (14,2) 360 (85,1) 63 (14,9) 7,066 0,061 Item 2 330 (89,9) 37 (10,1) 676 (91,0) 67 (9,0) 305 (84,7) 55 (15,3) 361 (85,1) 63 (14,9) 14,636** 0,088 0,09 Item 3 121 (32,8) 248 (67,2) 203 (27,4) 538 (72,6) 118 (32,9) 241 (67,1) 149 (35,2) 274 (64,8) 9,083* 0,069 0,07 Item 4 119 (33,0) 242 (67,0) 252 (34,8) 471 (65,2) 128 (36,3) 225 (63,7) 146 (35,6) 264 (64,4) 0,974 0,023 Item 5 294 (79,9) 74 (20,1) 593 (79,5) 153 (20,5) 273 (76,5) 84 (23,5) 316 (75,1) 105 (24,9) 4,343 0,048 Item 6 266 (72,9) 99 (27,1) 513 (69,4) 226 (30,6) 248 (69,7) 108 (30,3) 268 (63,8) 152 (36,2) 7,960* 0,065 0,06 Item 7 189 (51,5) 178 (48,5) 383 (51,6) 360 (48,4) 196 (55,1) 160 (44,9) 193 (46,2) 225 (53,8) 6,368 0,058 Item 8 282 (78,8) 76 (21,2) 585 (80,7) 140 (19,3) 273 (76,5) 84 (23,5) 304 (74,0) 107 (26,0) 7,568 0,064 Item 9 260 (71,2) 105 (28,8) 531 (72,0) 207 (28,0) 255 (71,4) 102 (28,6) 277 (66,3) 141 (33,7) 4,582 0,049 Item 10 269 (72,9) 100 (27,1) 537 (72,2) 207 (27,8) 242 (68,0) 114 (32,0) 268 (63,8) 152 (36,2) 11,344** 0,077 0,08 Item 11 283 (77,1) 84 (22,9) 552 (74,6) 188 (25,4) 259 (73,0) 96 (27,0) 287 (68,5) 132 (31,5) 8,330* 0,067 0,07 Item 12 197 (53,5) 171 (46,5) 394 (53,8) 339 (46,2) 199 (55,6) 159 (44,4) 183 (43,6) 237 (56,4) 15,006** 0,089 0,09 Item 13 278 (75,8) 89 (24,2) 577 (78,1) 162 (21,9) 265 (74,2) 92 (25,8) 303 (71,8) 119 (28,2) 6,100 0,057 Item 14 260 (71,6) 103 (28,4) 545 (74,1) 191 (25,9) 258 (72,3) 99 (27,7) 281 (67,4) 136 (32,6) 5,888 0,056 Item 15 258 (71,5) 103 (28,5) 531 (72,2) 204 (27,8) 251 (70,5) 105 (29,5) 273 (65,5) 144 (34,5) 6,231 0,058 Item 16 120 (32,9) 245 (67,1) 273 (37,3) 458 (62,7) 143 (41,1) 205 (58,9) 161 (39,1) 251 (60,9) 5,684 0,055 Item 17 186 (53,3) 163 (46,7) 365 (50,2) 362 (49,8) 180 (53,3) 158 (46,7) 208 (51,5) 196 (48,5) 1,338 0,027 Item 18 218 (60,2) 144 (39,8) 448 (61,3) 283 (38,7) 209 (59,5) 142 (40,5) 234 (56,3) 182 (43,7) 2,860 0,039 Item 19 130 (35,7) 234 (64,3) 261 (35,1) 483 (64,9) 126 (35,6) 228 (64,4) 184 (44,5) 229 (55,5) 11,670** 0,079 0,08 Item 20 166 (45,1) 202 (54,9) 313 (42,0) 432 (58,0) 178 (50,1) 177 (49,9) 214 (51,1) 205 (48,9) 11,653** 0,079 0,08 ** p ≤ 0,01 * p ≤ 0,05 According to Table 14, significant relationships were found in respect of items 2, 10, 12, 19 and 20 (p ≤ 0,01) and items 3, 6 and 11 (p ≤ 0,05). Corresponding effect sizes are small. The results indicate a trend in terms of which female -headed, extended families, in particular, seem to be related to lower levels of perceived family support. The concept of “female -headed, extended family” implies two aspects that could possibly impact on adolescents in such families. Firstly, “extended” suggests the possibility of a big household. A large household size hampers parent-child relationships (Surajnarayan, 1991) and probably also other relationships among family members. Berk (2003) points out that large families le ad to less parent-child interaction, fewer available resources per member, and parents who are more 121 impatient and more punitive. All these factors, in turn, influence perceived family support. Secondly, “female-headed” suggests that these families are single-parent families. Gore et al. (1992) established that adolescents in single -parent families experience less supportive relationships. Contrary to this finding, Lansford, Ceballo, Abbey and Stewart (2001) found that children in single-mother families did not differ from children in two-parent families in terms of either relationships or well-being. The results of the χ²-test for parents’ marital status and perceived support from family are shown in Table 15. Table 15 Parents’ marital status and adolescents’ responses (yes/no) to items relating to perceived support from family Item Married Never married/Live Divorced/Separated/ χ² φ w together Widowed Yes No Yes No Yes No Item 1 926 (89,0) 115 (11,0) 258 (90,2) 28 (9,8) 321 (87,7) 45 (12,3) 1,029 0,025 Item 2 930 (89,9) 104 (10,1) 264 (92,3) 22 (7,7) 309 (84,9) 55 (15,1) 10,516** 0,079 0,08 Item 3 293 (28,4) 739 (71,6) 91 (32,0) 193 (68,0) 124 (33,9) 242 (66,1) 4,409 0,051 Item 4 363 (35,9) 647 (64,1) 86 (30,6) 195 (69,4) 128 (35,6) 231 (64,4) 2,847 0,042 Item 5 820 (79,2) 215 (20,8) 236 (82,5) 50 (17,5) 283 (77,5) 82 (22,5) 2,496 0,038 Item 6 713 (69,6) 312 (30,4) 207 (72,9) 77 (27,1) 253 (69,9) 109 (30,1) 1,197 0,027 Item 7 539 (52,1) 495 (47,9) 140 (49,6) 142 (50,4) 179 (49,0) 186 (51,0) 1,293 0,028 Item 8 817 (80,5) 198 (19,5) 218 (78,1) 61 (21,9) 265 (74,2) 92 (25,8) 6,262* 0,062 0,06 Item 9 724 (70,8) 299 (29,2) 195 (70,1) 83 (29,9) 259 (70,8) 107 (29,2) 0,044 0,005 Item 10 734 (71,1) 299 (28,9) 198 (70,0) 85 (30,0) 265 (72,2) 102 (27,8) 0,397 0,015 Item 11 768 (74,7) 260 (25,3) 206 (74,1) 72 (25,9) 273 (75,2) 90 (24,8) 0,102 0,008 Item 12 565 (55,4) 455 (44,6) 133 (47,0) 150 (53,0) 181 (49,3) 186 (50,7) 8,338* 0,070 0,07 Item 13 799 (77,5) 232 (22,5) 215 (76,2) 67 (23,8) 272 (74,9) 91 (25,1) 1,036 0,025 Item 14 743 (72,5) 282 (27,5) 207 (74,2) 72 (25,8) 253 (70,1) 108 (29,9) 1,400 0,029 Item 15 740 (72,3) 284 (27,7) 193 (69,4) 85 (30,6) 252 (70,2) 107 (29,8) 1,158 0,026 Item 16 382 (37,7) 632 (62,3) 91 (32,5) 189 (67, 5) 126 (35,4) 230 (64,6) 2,701 0,040 Item 17 508 (50,9) 491 (49,1) 150 (54,6) 125 (45,4) 172 (49,9) 173 (50,1) 1,528 0,031 Item 18 616 (60,8) 397 (39,2) 165 (59,4) 113 (40,6) 215 (59,6) 146 (40,4) 0,297 0,013 Item 19 345 (33,5) 684 (66,5) 120 (42,9) 160 (57,1) 137 (38,2) 222 (61,8) 9,156** 0,074 0,07 Item 20 460 (44,5) 573 (55,5) 131 (46,1) 153 (53,9) 174 (47,7) 191 (52,3) 1,130 0,026 ** p ≤ 0,01 * p ≤ 0,05 Table 15 shows significant links between parents’ marital status and perceived family support for items 2 and 19 (p ≤ 0,01) and for items 8 and 12 (p ≤ 0,05). Calculated effect sizes are small. An analysis of the items suggests a trend towards a higher level 122 of perceived family support amongst adolescents whose parents are married. It is probable that the overall quality of family relationships and family processes is more influential in respect of adolescents’ well-being than the actual intactness or non- intactness of the family (Arnett, 2001; Lansford et al., 2001). Table 16 indicates the results for the χ ²-test for household size and perceived support from the family. Table 16 Household size and adolescents’ responses (yes/no) to items relating to perceived support from family Item 1 – 3 members 4 – 6 members 7 or more members χ² φ W Yes No Yes No Yes No Item 1 319 (89,1) 39 (10,9) 1212 (87,4) 175 (12,6) 324 (87,8) 45 (12,2) 0,787 0,019 Item 2 322 (90,7) 33 (9,3) 1212 (87,9) 167 (12,1) 318 (86,4) 50 (13,6) 3,355 0,040 Item 3 113 (31,7) 243 (68,3) 420 (30,6) 951 (69,4) 133 (36,2) 234 (63,8) 4,194 0,045 Item 4 119 (34,0) 231 (66,0) 475 (35,4) 865 (64,6) 134 (37,5) 223 (62,5) 0,987 0,022 Item 5 279 (78,2) 78 (21,8) 1076 (78,4) 297 (21,6) 279 (75,8) 89 (24,2) 1,116 0,023 Item 6 250 (69,8) 108 (30,2) 946 (69,4) 417 (30,6) 242 (66,3) 123 (33,7) 1,458 0,026 Item 7 172 (48,4) 183 (51,6) 715 (52,2) 656 (47,8) 187 (51,2) 178 (48,8) 1,549 0,027 Item 8 267 (75,9) 85 (24,1) 1045 (77,6) 301 (22,4) 280 (78,0) 79 (22,0) 0,598 0,017 Item 9 248 (69,5) 109 (30,5) 964 (70,9) 395 (29,1) 250 (68,5) 115 (31,5) 0,948 0,021 Item 10 259 (72,6) 98 (27,4) 967 (70,5) 404 (29,5) 246 (66,9) 122 (33,1) 2,992 0,038 Item 11 282 (79,0) 75 (21,0) 993 (72,9) 369 (27,1) 261 (71,5) 104 (28,5) 6,505* 0,056 0,06 Item 12 178 (49,9) 179 (50,1) 716 (52,6) 646 (47,4) 192 (52,8) 172 (47,2) 0,898 0,021 Item 13 268 (74,9) 90 (25,1) 314 (22,9) 262 (71,8) 103 (28,2) 4,537 0,047 Item 14 257 (72,8) 96 (27,2) 964 (70,9) 396 (29,1) 267 (73,0) 99 (27,0) 0,923 0,021 Item 15 226 (64,0) 127 (36,0) 979 (72,0) 381 (28,0) 252 (69,4) 111 (30,6) 8,613* 0,064 0,06 Item 16 123 (35,0) 228 (65,0) 511 (37,8) 839 (62,2) 141 (39,4) 217 (60,6) 1,499 0,027 Item 17 182 (53,1) 161 (46,9) 673 (50,9) 649 (49,1) 190 (53,8) 163 (46,2) 1,219 0,025 Item 18 213 (60,9) 137 (39,1) 804 (59,7) 543 (40,3) 214 (59,3) 147 (40,7) 0,210 0,010 Item 19 135 (38,3) 217 (61,7) 510 (37,3) 856 (62,7) 138 (38,3) 222 (61,7) 0,202 0,010 Item 20 158 (44,9) 194 (55,1) 617 (44,9) 757 (55,1) 194 (53,3) 170 (46,7) 8,520* 0,064 0,06 ** p ≤ 0,01 * p ≤ 0,05 The results in Table 16 indicate significant relationships (p ≤ 0,05) between household size and perceived family support for items 11, 15 and 20. Corresponding effect sizes are small. The results suggest a tendency for adolescents in smaller households to have more positive perceptions about their families. Larger families imply less attention to individual members and sharing of resources, which can lead 123 adolescents to perceive support from family members as being of a low level (Rice & Dolgin, 2005; Surajnarayan, 1991). The results of the χ ²-test concerning the educational level of fathers and perceived family support are indicated in Table 17. Table 17 Fathers’ educational status and adolescents’ responses (yes/no) to items relating to perceived support from family Item None Grade 1 - 7 Grade 8 – 12 Post-school χ² φ w Yes No Yes No Yes No Yes No Item 1 62 (93,9) 4 (6,1) 78 (89,7) 9 (10,3) 278 (86,1) 45 (13,9) 180 (94,2) 11 (5,8) 10,118* 0,123 0,13 Item 2 62 (93,9) 4 (6,1) 73 (84,9) 13 (15,1) 284 (89,0) 35 (11,0) 176 (92,6) 14 (7,4) 5,443 0,091 Item 3 20 (30,3) 46 (69,7) 27 (32,1) 57 (67,9) 92 (28,7) 228 (71,3) 53 (28,0) 136 (72,0) 0,544 0,029 Item 4 21 (32,3) 44 (67,8) 29 (36,7) 50 (63,3) 114 (35,8) 204 (64,2) 63 (34,6) 119 (65,4) 0,402 0,025 Item 5 58 (87,9) 8 (12,1) 66 (78,6) 18 (21,4) 254 (78,6) 69 (21,4) 153 (80,5) 37 (19,5) 3,082 0,068 Item 6 46 (69,7) 20 (30,3) 54 (62,8) 32 (37,2) 222 (69,6) 97 (30,4) 133 (70,7) 55 (29,3) 1,886 0,053 Item 7 37 (56,1) 29 (43,9) 46 (52,9) 41 (47,1) 171 (53,1) 151 (46,9) 106 (56,7) 81 (43,3) 0,766 0,034 Item 8 56 (84,9) 10 (15,1) 66 (80,5) 16 (19,5) 245 (77,0) 73 (23,0) 149 (80,5) 36 (19,5) 2,464 0,062 Item 9 44 (67,7) 21 (32,3) 52 (60,5) 34 (39,5) 216 (67,9) 102 (32,1) 139 (75,1) 46 (24,9) 6,403 0,099 Item 10 50 (76,9) 15 (23,1) 62 (72,1) 24 (27,9) 219 (68,0) 103 (32,0) 134 (71,7) 53 (28,3) 2,470 0,061 Item 11 51 (77,3) 15 (22,7) 64 (74,4) 22 (25,6) 244 (76,5) 75 (23,5) 135 (71,8) 53 (28,2) 1,594 0,049 Item 12 30 (46,1) 35 (53,9) 53 (62,4) 32 (37,6) 180 (56,3) 140 (43,7) 105 (56,5) 81 (43,5) 3,973 0,078 Item 13 56 (84,9) 10 (15,1) 61 (70,9) 25 (29,1) 244 (76,5) 75 (23,5) 150 (80,7) 36 (19,3) 5,433 0,091 Item 14 47 (72,3) 18 (27,7) 61 (70,9) 25 (29,1) 230 (72,1) 89 (27,9) 134 (72,0) 52 (28,0) 0,053 0,009 Item 15 46 (69,7) 20 (30,3) 61 (72,6) 23 (27,4) 230 (72,8) 86 (27,2) 139 (75,1) 46 (24,9) 0,804 0,035 Item 16 25 (38,5) 40 (61,5) 33 (39,8) 50 (60,2) 121 (38,4) 194 (61,6) 63 (34,2) 121 (65,8) 1,152 0,042 Item 17 34 (54,8) 28 (45,2) 38 (46,3) 44 (53,7) 170 (55,0) 139 (45,0) 85 (46,4) 98 (53,6) 4,547 0,085 Item 18 47 (71,2) 19 (28,8) 60 (70,6) 25 (29,4) 191 (61,0) 122 (39,0) 114 (63,0) 67 (37,0) 4,366 0,082 Item 19 27 (41,5) 38 (58,5) 39 (45,9) 46 (54,1) 110 (34,5) 209 (65,5) 66 (35,5) 120 (64,5) 4,504 0,083 Item 20 37 (56,1) 29 (43,9) 48 (57,1) 36 (42,9) 147 (45,5) 176 (54,5) 65 (34,8) 122 (65,2) 16,224** 0,157 0,16 ** p ≤ 0,01 * p ≤ 0,05 Table 17 shows significant relations between fathers’ educational status and perceived family support for item 1 (p ≤ 0,05) and item 20 (p ≤ 0,01). Effect sizes are small. Although the results in terms of item 1 may seem unusual, the explanation probably lies in the content of the item and more specifically, the term “moral support”. It is well known that children often exhibit loyalty towards their families (whether deserved or not). This may lead to positive responses to items even if these responses are not necessarily true. Adolescents’ responses to item 20 suggest that families in which fathers have a higher educational status are perceived in a more positive light. 124 This may be because fathers with a higher educational status are probably also in a position to obtain better employment. This in turn increases financial resources and leads to an overall higher standard of living, which is likely to impact positively on adolescents’ psychological well-being. Harris and Marmer’s (1996) finding that fathers in non-poor families are more emotionally and physically available for their children than fathers in poor families, lends further support to the research finding. The χ ²-test results for mothers’ educational status and perceived family support are shown in Table 18. Table 18 Mothers’ educational status and adolescents’ responses (yes/no) to items relating to perceived support from family Item None Grade 1 – 7 Grade 8 – 12 Post-school χ² φ W Yes No Yes No Yes No Yes No Item 1 77 (92,8) 6 (7,2) 117 (88,0) 16 (12,0) 564 (89,1) 69 (10,9) 232 (88,6) 30 (11,4) 1,409 0,036 Item 2 74 (89,2) 9 (10,8) 116 (87,9) 16 (12,1) 574 (91,6) 53 (8,4) 223 (85,4) 38 (14,6) 7,741 0,084 Item 3 26 (31,3) 57 (68,7) 38 (29,0) 93 (71,0) 171 (27,2) 457 (72,8) 88 (33,7) 173 (66,3) 3,928 0,060 Item 4 30 (37,0) 51 (63,0) 50 (40,0) 75 (60,0) 212 (34,4) 405 (65,6) 92 (36,4) 161 (63,6) 1,601 0,039 Item 5 66 (79,5) 17 (20,5) 109 (82,6) 23 (17,4) 506 (80,6) 122 (19,4) 198 (76,2) 62 (23,8) 2,992 0,052 Item 6 56 (68,3) 26 (31,7) 95 (71,4) 38 (28,6) 436 (69,5) 191 (30,5) 178 (69,3) 79 (3 0,7) 0,292 0,016 Item 7 48 (57,8) 35 (42,2) 74 (56,1) 58 (43,9) 340 (54,1) 288 (45,9) 115 (44,6) 143 (55,4) 8,789* 0,089 0,09 Item 8 67 (83,8) 13 (16,2) 101 (80,2) 25 (19,8) 475 (77,4) 139 (22,6) 208 (81,3) 48 (18,7) 2,969 0,053 Item 9 54 (65,9) 28 (3 4,1) 88 (67,2) 43 (32,8) 435 (69,9) 187 (30,1) 181 (70,7) 75 (29,3) 1,078 0,031 Item 10 56 (68,3) 26 (31,7) 91 (68,4) 42 (31,6) 455 (72,5) 173 (27,5) 163 (63,4) 94 (36,6) 7,192 0,081 Item 11 61 (74,4) 21 (25,6) 92 (70,8) 38 (29,2) 483 (77,2) 143 (22,8) 186 (72,9) 69 (27,1) 3,380 0,056 Item 12 41 (50,6) 40 (49,4) 83 (62,9) 49 (37,1) 345 (55,2) 280 (44,8) 115 (44,7) 142 (55,3) 13,560** 0,111 0,11 Item 13 64 (77,1) 19 (22,9) 97 (74,1) 34 (25,9) 483 (77,4) 141 (22,6) 194 (75,5) 63 (24,5) 0,887 0,028 Item 14 61 (73,5) 22 (26,5) 93 (71,5) 37 (28,5) 470 (75,6) 152 (24,4) 166 (65,1) 89 (34,9) 10,030* 0,096 0,10 Item 15 60 (74,1) 21 (25,9) 95 (74,2) 33 (25,8) 456 (73,7) 163 (26,3) 173 (67,6) 83 (32,4) 3,792 0,059 Item 16 27 (33,7) 53 (66,3) 58 (44,6) 72 (55,4) 211 (34,5) 401 (65,5) 98 (38,6) 156 (61,4) 5,497 0,071 Item 17 43 (54,4) 36 (45,6) 66 (53,2) 58 (46,8) 313 (52,3) 286 (47,7) 114 (45,1) 139 (54,9) 4,551 0,066 Item 18 51 (61,5) 32 (38,5) 89 (67,9) 42 (32,1) 391 (63,4) 226 (36,6) 141 (56,2) 110 (43,8) 6,080 0,075 Item 19 31 (37,3) 52 (62,7) 51 (39,2) 79 (60,8) 224 (36,0) 398 (64,0) 104 (40,6) 152 (59,4) 1,816 0,041 Item 20 43 (51,8) 40 (48,2) 66 (50,8) 64 (49,2) 283 (44,8) 348 (55,2) 112 (43,6) 145 (56,4) 3,240 0,054 ** p ≤ 0,01 * p ≤ 0,05 Table 18 shows significant relations between mothers’ educational status and perceived family support for item 12 (p ≤ 0,01) and for items 7 and 14 (p ≤ 0,05). The calculated effect sizes are small. The results suggest that better-educated mothers are positively linked to perceptions of less family support. A possible reason for this finding is that mothers with higher educational levels have better employment 125 opportunities that may in turn require them to work longer hours. As it has repeatedly been shown that the mother is usually the more supportive parent (Colarossi, 2001; Field et al., 1995; Harris & Marmer, 1996), seeing less of their mothers might have been perceived as a loss of family support by the adolescents in this study. The results of the χ²-test for fathers’ employment status and perceived family support are shown in Table 19. Table 19 Fathers’ employment status and adolescents’ responses (yes/no) to items relating to perceived support from family Item Employed Unemployed χ² φ W Yes No Yes No Item 1 708 (88,9) 88 (11,1) 135 (89,4) 16 (10,6) 0,027 -0,005 Item 2 716 (90,5) 75 (9,5) 133 (88,1) 18 (11,9) 0,848 0,030 Item 3 219 (27,8) 570 (72,2) 44 (29,3) 106 (70,7) 0,155 -0,013 Item 4 269 (34,7) 506 (65,3) 52 (35,4) 95 (64,6) 0,024 -0,005 Item 5 625 (78,6) 170 (21,4) 124 (82,1) 27 (17,9) 0,944 -0,032 Item 6 551 (70,0) 236 (30,0) 100 (68,5) 46 (31,5) 0,135 0,012 Item 7 404 (51,0) 388 (49,0) 78 (52,0) 72 (48,0) 0,049 -0,007 Item 8 623 (80,2) 154 (19,8) 116 (78,9) 31 (2 1,1) 0,124 0,012 Item 9 566 (71,9) 221 (28,1) 100 (67,6) 48 (32,4) 1,151 0,035 Item 10 564 (71,2) 228 (28,8) 100 (66,7) 50 (33,3) 1,253 0,036 Item 11 593 (75,5) 192 (24,5) 112 (75,2) 37 (24,8) 0,009 0,003 Item 12 423 (54,2) 358 (45,8) 74 (49,0) 77 (51,0) 1,351 0,038 Item 13 621 (78,9) 166 (21,1) 107 (71,3) 43 (28,7) 4,170* 0,067 0,07 Item 14 581 (74,1) 203 (25,9) 98 (65,3) 52 (34,7) 4,883* 0,072 0,07 Item 15 574 (73,2) 210 (26,8) 104 (69,8) 45 (30,2) 0,735 0,028 Item 16 287 (36,9) 490 (63,1) 59 (39,9) 89 (60,1) 0,455 -0,022 Item 17 395 (51,6) 371 (48,4) 75 (51,4) 71 (48,6) 0,002 0,001 Item 18 475 (61,5) 298 (38,5) 89 (60,5) 58 (39,5) 0,043 0,007 Item 19 271 (34,4) 517 (65,6) 56 (37,8) 92 (62,2) 0,651 -0,026 Item 20 333 (42,1) 457 (57,9) 76 (50,3) 75 (49,7) 3,451 -0,061 ** p ≤ 0,01 * p ≤ 0,05 Table 19 shows statistically significant relations (p ≤ 0,05) between fathers’ employment status and perceived family support for items 13 and 14. Corresponding effect sizes are small. Participants’ responses to the three items show a slight trend towards a positive relationship between employed fathers and perceived family support. Visser (1991) found that there is more disintegration in families where 126 fathers are unemployed. Children in such families perceive their fathers as less available and this can lead to a perception of decreased family support. Table 20 indicates the results of the χ²-test for mothers’ employment status and perceived family support. Table 20 Mothers’ employment status and adolescents’ responses (yes/no) to items relating to perceived support from family Item Employed Unemployed χ² φ Yes No Yes No Item 1 956 (89,2) 116 (10,8) 339 (89,2) 41 (10,8) 0,000 -0,000 Item 2 949 (89,4) 113 (10,6) 343 (90,0) 38 (10,0) 0,133 -0,010 Item 3 315 (29,6) 749 (70,4) 105 (27,8) 273 (72,2) 0,451 0,018 Item 4 363 (34,7) 684 (65,3) 134 (36,2) 236 (63,8) 0,287 -0,014 Item 5 850 (79,7) 217 (20,3) 306 (81,0) 72 (19,0) 0,290 -0,014 Item 6 738 (69,9) 323 (30,4) 262 (70,1) 112 (29,9) 0,032 -0,005 Item 7 535 (50,3) 529 (49,7) 202 (53,4) 176 (46,6) 1,113 -0,028 Item 8 826 (79,5) 213 (20,5) 299 (80,0) 75 (20,0) 0,034 -0,005 Item 9 752 (71,3) 303 (28,7) 277 (74,1) 97 (25,9) 1,062 -0,027 Item 10 742 (69,8) 321 (30,2) 267 (70,6) 111 (29,4) 0,092 -0,008 Item 11 791 (74,8) 266 (25,2) 279 (74,2) 97 (25,8) 0,059 0,006 Item 12 549 (51,8) 510 (48,2) 207 (55,4) 167 (44,6) 1,363 -0,031 Item 13 816 (76,9) 245 (23,1) 276 (73,8) 98 (26,2) 1,472 0,032 Item 14 750 (71,6) 298 (28,4) 285 (75,6) 92 (24,4) 2,267 -0,040 Item 15 741 (70,8) 305 (29,2) 275 (73,1) 101 (26,9) 0,715 -0,022 Item 16 386 (36,8) 662 (63,2) 128 (34,7) 241 (65,3) 0,543 0,020 Item 17 525 (51,5) 495 (48,5) 191 (52,3) 174 (47,7) 0,079 -0,008 Item 18 625 (59,8) 421 (40,2) 235 (63,5) 135 (36,5) 1,622 -0,034 Item 19 396 (37,6) 656 (62,4) 130 (34,4) 248 (65,6) 1,264 0,030 Item 20 468 (44,0) 595 (56,0) 181 (47,8) 198 (52,2) 1,571 -0,033 ** p ≤ 0,01 * p ≤ 0,05 Table 20 shows that there were no significant results for the relationship between the two mentioned variables. In view of the earlier finding of a significant relationship between mothers’ educational status and perceived family support, this is an unusual finding. Yeung, Sandberg, Davis-Kean and Hofferth (2001) found that mothers do most of the parenting, even in the case of working mothers, with fathers being involved mostly over weekends. One might therefore have expected adolescents to be more sensitive in terms of any perceived changes in maternal support caused by mothers’ employment status. It is possible, though, that mediating factors such as 127 support from other household members (e.g. a grandmother) compensate for a perceived loss of maternal support. The χ²-test results for time spent with father and perceived family support are summarized in Table 21. Table 21 Time spent with father per week and adolescents’ responses (yes/no) to items relating to perceived support from family Item Up to 6 hours More than 6 hours χ² φ w Yes No Yes No Item 1 463 (85,9) 76 (14,1) 440 (92,8) 34 (7,2) 12,503** -0,111 0,11 Item 2 470 (87,7) 66 (12,3) 430 (91,7) 39 (8,3) 4,273* -0,065 0,07 Item 3 160 (29,9) 375 (70,1) 115 (24,7) 351 (75,3) 3,417 0,058 Item 4 188 (36,1) 333 (63,9) 152 (33,0) 308 (67,0) 0,998 0,032 Item 5 417 (77,9) 118 (22,1) 383 (81,0) 90 (19,0) 1,406 -0,037 Item 6 362 (67,8) 172 (32,2) 344 (74,3) 119 (25,7) 5,081* -0,071 0,07 Item 7 272 (50,8) 263 (49,2) 258 (54,8) 213 (45,2) 1,557 -0,039 Item 8 399 (75,9) 127 (24,1) 375 (81,7) 84 (18,3) 4,973* -0,071 0,07 Item 9 359 (67,5) 173 (32,5) 359 (77,5) 104 (22,5) 12,463** -0,112 0,11 Item 10 361 (67,6) 173 (32,4) 354 (75,3) 116 (24,7) 7,261** -0,085 0,09 Item 11 385 (72,2) 148 (27,8) 360 (77,1) 107 (22,9) 3,089 -0,056 Item 12 287 (54,2) 243 (45,8) 247 (53,4) 216 (46,6) 0,064 0,008 Item 13 394 (73,8) 140 (26,2) 383 (81,8) 85 (18,2) 9,293** -0,096 0,10 Item 14 379 (72,2) 146 (27,8) 354 (75,3) 116 (24,7) 1,251 -0,035 Item 15 365 (69,3) 162 (30,7) 346 (74,6) 118 (25,4) 3,431 -0,059 Item 16 206 (39,5) 316 (60,5) 169 (36,7) 292 (63,3) 0,816 0,029 Item 17 258 (50,1) 257 (49,9) 239 (52,5) 216 (47,5) 0,571 -0,024 Item 18 295 (56,5) 227 (43,5) 303 (66,5) 153 (33,5) 10,110** -0,102 0,10 Item 19 195 (36,8) 335 (63,2) 163 (35,0) 302 (65,0) 0,325 0,018 Item 20 238 (44,6) 296 (55,4) 186 (39,9) 280 (60,1) 2,208 0,047 ** p ≤ 0,01 * p ≤ 0,05 Results in Table 21 show significant relations between time spent with fathers and perceived support from family for items 1, 9, 10 and 18 (p ≤ 0,01) and for items 2, 6 and 8 (p ≤ 0,05). Corresponding effect sizes are small. The general trend observed was that adolescents who spent more time with their fathers reported more positive perceptions of support. Despite mothers being indicated as a primary source of familial support, it is clear that fathers and support from fathers are also important in terms of the well-being of adolescents (cf. Berk, 2003; Lansford et al., 2001). 128 The results for the χ²-test for time spent with mothers and perceived family support are supplied in Table 22. Table 22 Time spent with mother per week and adolescents’ responses (yes/no) to items relating to perceived support from family Item Up to 6 hours More than 6 hours χ² φ W Yes No Yes No Item 1 325 (81,7) 73 (18,3) 1203 (92,5) 97 (7,5) 40,037** -0,154 0,16 Item 2 331 (83,6) 65 (16,4) 1190 (92,0) 103 (8,0) 24,155** -0,120 0,12 Item 3 137 (34,8) 257 (65,2) 363 (28,1) 927 (71,9) 6,359* 0,061 0,06 Item 4 141 (36,6) 244 (63,4) 413 (32,6) 852 (67,4) 2,091 0,036 Item 5 287 (72,5) 109 (27,5) 1065 (82,2) 230 (17,8) 18,040** -0,103 0,10 Item 6 242 (61,4) 152 (38,6) 940 (73,3) 343 (26,7) 20,327** -0,110 0,11 Item 7 175 (44,3) 220 (55,7) 694 (53,9) 594 (46,1) 11,105** -0,081 0,08 Item 8 280 (72,0) 109 (28,0) 1036 (81,6) 234 (18,4) 16,718** -0,100 0,10 Item 9 256 (65,0) 138 (35,0) 937 (73,4) 340 (26,6) 10,404** -0,079 0,08 Item 10 250 (63,1) 146 (36,9) 967 (74,8) 326 (25,2) 20,455** -0,110 0,11 Item 11 278 (70,4) 117 (29,6) 981 (76,5) 301 (23,5) 6,086* -0,060 0,06 Item 12 183 (46,4) 211 (53,6) 699 (54,7) 580 (45,3) 8,136** -0,070 0,07 Item 13 268 (68,0) 126 (32,0) 1038 (80,7) 249 (19,3) 27,774** -0,129 0,13 Item 14 264 (67,9) 125 (32,1) 976 (76,3) 304 (23,7) 10,980** -0,081 0,08 Item 15 255 (65,6) 134 (34,4) 935 (73,0) 346 (27,0) 8,058** -0,069 0,07 Item 16 162 (42,1) 223 (57,9) 443 (34,8) 828 (65,2) 6,650** 0,063 0,06 Item 17 187 (49,9) 188 (50,1) 676 (54,2) 572 (45,8) 2,141 -0,036 Item 18 210 (54,1) 178 (45,9) 796 (62,9) 470 (37,1) 9,545** -0,076 0,08 Item 19 164 (41,9) 227 (58,1) 451 (35,2) 830 (64,8) 5,847* 0,059 0,06 Item 20 183 (46,4) 211 (53,6) 563 (43,5) 730 (56,5) 1,033 0,025 ** p ≤ 0,01 * p ≤ 0,05 Statistically significant relationships were found between time spent with mothers and perceived family support for items 1, 2, 5, 6, 7, 8, 9, 10, 12, 13, 14, 15, 16 and 18 (p ≤ 0,01) and for items 3, 11 and 19 (p ≤ 0,05). Calculated effect sizes are small. The findings show a strong trend towards positive relationships between more time spent with mothers and perceived familial support. This is an expected finding, given the emphasis placed by researchers on the primary importance of mothers as a source of emotional support for their children (Colarossi, 2001; Field et al., 1995; Young et al., 1995). 129 The χ²-test result for households taking care of household members with severe disabilities or serious chronic illnesses is shown in Table 23. Table 23 Households taking care of members with severe disabilities or serious chronic illnesses and adolescents’ responses (yes/no) to items relating to perceived support from family Item Taking care of Not taking care of χ² φ w disabled/chronically ill disabled/chronically ill household members household members Yes No Yes No Item 1 739 (87,3) 108 (12,7) 1030 (88,2) 138 (11,8) 0,401 -0,014 Item 2 754 (89,3) 90 (10,7) 1008 (87,0) 151 (13,0) 2,581 0,036 Item 3 295 (35,3) 540 (64,7) 341 (29,4) 818 (70,6) 7,797** 0,063 0,06 Item 4 328 (40,2) 488 (59,8) 372 (32,7) 766 (67,3) 11,650** 0,077 0,08 Item 5 656 (78,3) 182 (21,7) 892 (76,8) 270 (23,2) 0,641 0,018 Item 6 580 (69,3) 257 (30,7) 784 (68,0) 369 (32,0) 0,379 0,014 Item 7 462 (55,5) 371 (44,5) 554 (47,7) 607 (52,3) 11,641** 0,076 0,08 Item 8 659 (80,0) 165 (20,0) 862 (75,6) 279 (24,4) 5,364* 0,052 0,05 Item 9 571 (69,0) 256 (31,0) 825 (71,1) 336 (28,9) 0,937 -0,022 Item 10 605 (72,3) 232 (27,7) 793 (68,2) 370 (31,8) 3,882* 0,044 0,04 Item 11 639 (76,3) 198 (23,7) 829 (72,0) 323 (28,0) 4,816* 0,049 0,05 Item 12 481 (57,5) 356 (42,5) 547 (47,5) 604 (52,5) 19,187** 0,098 0,10 Item 13 659 (78,7) 178 (21,3) 853 (73,5) 307 (26,5) 7,147** 0,060 0,06 Item 14 601 (72,5) 228 (27,5) 812 (70,4) 342 (29,6) 1,072 0,023 Item 15 594 (71,7) 234 (28,3) 789 (68,4) 365 (31,6) 2,594 0,036 Item 16 326 (39,8) 493 (60,2) 418 (36,4) 729 (63,6) 2,296 0,034 Item 17 398 (50,0) 398 (50,0) 597 (52,8) 534 (47,2) 1,451 -0,027 Item 18 531 (64,3) 295 (35,7) 642 (56,4) 497 (43,6) 12,484** 0,080 0,08 Item 19 337 (40,8) 489 (59,2) 417 (36,0) 741 (64,0) 4,692* 0,049 0,05 Item 20 448 (53,5) 390 (46,5) 482 (41,7) 674 (58,3) 27,021** 0,116 0,12 ** p ≤ 0,01 * p ≤ 0,05 Table 23 shows significant relations between taking care of a disabled or chronically ill household member and perceived family support for items 3, 4, 7, 12, 13, 18 and 20 (p ≤ 0,01) and for items 8, 10, 11 and 19 (p ≤ 0,05). The corresponding effect sizes are small. Participants’ responses to the mentioned items did not indicate a clear trend towards either a higher or lower level of perceived familial support in the case of households who have to take care of disabled or chronically ill members. As noted earlier, black people in South Africa always had either poor access or no access at all to health services (including mental health services) (Hickson & Kriegler, 1991). Senel and Akkök’s (1996) observation with regard to Turkish adolescents is probably 130 also true of the adolescents in this study, namely that they are used to living with disabled or chronically ill household members. In addition, the nature of the adolescents’ relationship with the disabled/chronically ill household member is also of importance. For example, if the disabled/chronically ill member is a sibling, the impact on the adolescent will be different from the impact that would result in the case of an adult household member. The amount and type of caregiving that the adolescent is required to provide to the involved household member will also influence the effect thereof on the adolescent. In general it thus appears that this variable does not impact significantly on adolescents’ perception of family support. Yet it is clear that other factors should be taken into account in order to fully understand the effect of the presence of a chronically ill/disabled household member on the adolescent. CONCLUSION An important finding of the current study is that fathers’ educational status is correlated with perceived support from friends and perceived support from family. Several other trends have been observed. Overall, family support appears to be of greater importance to the adolescents in this study than peer support. It seems that adolescents experience less perceived support in larger families, and particularly in female-headed extended families. Intact families, employed fathers, and more time spent with parents, appear to be associated with positive perceptions in respect of familial support. Employment among mothers appears to link up with higher levels of perceived support from peers. Although more significant results were expected, the general findings of this study indicate that selected family variables do impact on adolescents’ perceptions of support. Moreover, the identified trends suggest that what is generally considered to be a stable family environment (e.g. an intact family, parental employment, good parental education and sufficient family time) has a positive impact on adolescents’ psychological well-being. South Africa’s socio-political history caused major disruption in black families. For example, divorce rates increased; migratory labor created an undesirable situation in which parents and children were forced to live apart for long periods of time; and 131 parental authority declined. In post-apartheid South Africa the black family needs assistance to overcome these negative effects and to find a synthesis between traditional family values and the demands of modern society. If the black family can be stabilized, this will impact positively on all family members, including adolescents. The results of this study indicate that the family is still very important to the black adolescent. It is thus important that intervention strategies should be considered a matter of urgency. As the family is still largely considered a pillar of the community, positive interventions in the family will also have a positive impact on communities in general. A shortcoming of this project is that perceptions of social support were not measured separately for boys and girls. Measuring parental support instead of global family support might also have yielded more significant results. In addition, measuring perceived support from mothers and fathers separately could also lead to more significant findings. The concepts “severe disability” and “serious chronic illness” are too broad and vague. This complicated the interpretation of results relating to this variable. Future studies should aim to focus on either disability or chronic illness, and to define these terms beforehand. The relationship of respondents to disabled or chronically ill household members should also be determined during the collection of data. What comes to light in this study is that there may be mediating factors in the relationship between family variables and adolescents’ perceptions of social support which have not been identified in this study. Future research should aim to identify these factors, and to examine their impact on adolescent well-being. This should be done specifically within the South African context, as it is likely that there are certain variables that are unique to blacks in this country. The lack of relevant, local data on the state of children and families in South Africa complicates the planning and execution of intervention strategies (National Programme of Action for Children in South Africa, 2002). The current study forms a solid foundation from which further research can be conducted, and will hopefully contribute to correcting the current state of affairs, thereby enhancing targeted intervention. 132 LIST OF REFERENCES Adendorff, E .P. (1998). Stress, sosiale bronne en coping by verskillende groepe adolessente. Ongepubliseerde meestersverhandeling. Potchefstroom: Potchefstroomse Universiteit vir Christelike Hoër Onderwys. Andrews, A.B. & Morrison, B. (1997). The changing American family: Impact on South Carolina’s children. Columbia: Institute for Families in Society. Arnett, J.J. (2001). Adolescence and emerging adulthood: A cultural approach. New Jersey: Prentice Hall. Berk, L.E. (2003). Child Development (6th ed.). Boston: Pearson Education. Bullock, B.M. (2002). Sibling collusion and problem behaviour in early adolescence: toward a process model for family mutuality. Journal of Abnormal Child Psychology. Retrieved October 6, 2004, from http://www/findarticles.com/p/articles/mi_m0902/is_2_30/ai_85500248/ Colarossi, L.G. (2001). Adolescent gender differences in social support: Structure, function, and provider type. Social Work Research, 25(4), 233-241. Cornwell, B. (2003). The dynamic properties of social support: Decay, growth and staticity, and their effects on adolescent depression. Social Forces, 81 (3), 953- 978. Cumsille, P. & Epstein, N. (1994). Family cohesion, family adaptability, social support, and adolescent depressive symptoms in outpatient clinic families. Journal of Family Psychology, 8(2), 202-214. Dekovic, M. & Janssens, J.M.A.M. (1992). Parents’ child-rearing style and child’s sociometric status. Developmental Psychology, 28 (5), 925-943. 133 Dumont, M. & Provost, M.A. (1999). Resilience in adolescents: Protective role of social support, coping strategies, self-esteem, and social activities on experience of stress and depression. Journal of Youth and Adolescence, 28(3), 343-363. Dunn, C.W. & Tucker, C.M. (1993). Black children’s adaptive functioning and maladaptive behaviour associated with quality of family support. Journal of Multicultural Counseling and Development, 21, 79-87. Field, T., Diego, M. & Sanders, C. (2002). Adolescents’ parent and peer relationships. Adolescence, 37 (145), 121-130. Retrieved March 1, 2005, from Academic Search Premier database. Field, T., Lang, H.C., Yando, R. & Bendell, D. (1995). Adolescents’ intimacy with parents and friends. Adolescence, 30(117), 133-140. Retrieved September 14, 2004, from Academic Search Premier database. Frey, C.U. & Röthlisberger, C. (1996). Social support in healthy adolescents. Journal of Youth and Adolescence, 25 (1), 17-31. Gore, S., Aseltine, R.H, & Colton, M.E. (1992). Social structure, life stress and depressive symptoms in a high school-aged population. Journal of Health and Social Behavior, 33, 97-113. Grossman, M. & Rowat, K.M. (1995). Parental relationships, coping strategies, received support, and well-being in adolescents of separated or divorced and married parents. Research in Nursing and Health, 18, 249-261. Harris, J.R. (2000). Socialization, personality development, and the child’s environments: Comment on Vandell (2000). Developmental Psychology, 36(6), 699-710. Retrieved February 28, 2000, from http://home.att.net/~xchar/tna/devpayjh.htm 134 Harris, K.M. & Marmer, J.K. (1996). Poverty, paternal involvement and adolescent well-being. Journal of Family Issues, 17, 614-640. Harry, B. (2002). Trends and issues in serving culturally diverse families of children with disabilities. Journal of Special Education, 36 (3), 131-138. Retrieved March 1, 2005, from Academic Search Premier database. Hickson, J. & Kriegler, S. (1991). Childshock: The effects of apartheid on the mental health of South Africa’s children. International Journal for the Advancement of Counseling, 14(2), 141-154. Hortacsu, N. (1995). Parents’ education levels, parents’ beliefs, and child outcomes. Journal of Genetic Psychology, 156 (3), 373-383. Howell, D.C. (2002). Statistical methods for psychology (5th ed.) Johannesburg: Duxbury. Huysamen, G.K. (1983). Inferensiële statistiek en navorsingsontwerp: ‘n Inleiding. Kaapstad: Academica. Kaltiala -Heino, R., Rimpelä, M., Rantanen, P & Laippala, P. (2001). Adolescent depression: The role of discontinuities in life course and social support. Journal of Affective Disorders, 64, 155-166. Lansford, J.E., Ceballo, R., Abbey, A. & Stewart, A.J. (2001). Does family structure matter? A comparison of adoptive, two-parent biological, single-mother, stepfather, and stepmother households. Journal of Marriage and the Family, 63(3), 840-851. Laugesen, N., Dugas, M.J. & Bukowski, W.M. (2003, November). The relationship between perceived social support and worry in adolescence. Poster presented at the 37th Annual Convention for the Association for the Advancement of Behavior Therapy, Boston, MA, USA. 135 Lerner, R.M. (2002). Adolescence: Development, diversity, context and application. New Jersey: Pearson Education. Liu, Y. (2002). The role of perceived social support and dysfunctional attitudes in predicting Taiwanese adolescents’ depressive tendency. Adolescence, 37(148), 823-834. Lopez, F.G. & Thurman, C.W. (1993). High-trait and low -trait angry college students: A comparison of family environments. Journal of Counseling and Development, 71, 524-527. Marcotte, G., Marcotte, D. & Bouffard, T. (2002). The influence of familial support and dysfunctional attitudes on depression and delinquency in an adolescent population. European Journal of Psycholo gy of Education, 17 (4), 363-376. McFarlane, A.H., Bellissimo, A. & Norman, G.R. (1995). The role of family and peers in social self-efficacy: Links to depression in adolescence. American Journal of Orthopsychiatry, 65 (3), 402-410. Mullis, R.L., Hill, E.W. & Readdick, C.A. (1999). Attachment and social support among adolescents. Journal of Genetic Psychology, 160(4), 500-502. Retrieved March 1, 2005, from Academic Search Premier database. Murberg, T.A. & Bru, E. (2004). Social support, negative life events and emotional problems among Norwegian adolescents. School Psychology International, 25(4), 387-403. National Programme of Action for Children in South Africa (2002). Children in 2001: Report on the state of the nation’s children. Pretoria: The Presidency. Ohanessian, C.M., Lerner, R.M., von Eye., A & Lerner, J.V. (1996). Direct and indirect relations between perceived parental acceptance, perceptions of the self, and emotional adjustment during early adolescence. Family and Consumer Sciences Research Journal, 25(2), 159-183. 136 Ostrander, R., Weinfurt, K.P. & Nay, W.R. (1998). The role of age, family support, and negative cognitions in the prediction of depressive symptoms. School Psychology Review, 27 (1), 121-137. Retrieved September 14, 2004, from Academic Search Premier database. Patten, C.A., Gillin, J.C., Farkas, A.J., Gilpin, E.A., Berry, C.C. & Pierce, J.P. (1997). Depressive symptoms in California adolescents: Family structure and parental support. Journal of Adolescent Health , 20(4), 271-278. Procidano, M.E. & Heller, K. (1983). Measures of perceived social support from friends and from family: Three validation studies. American Journal of Community Psychology, 11(1), 1-24. Rice, F.P. & Dolgin, K.G. (2005). The adolescent: Development, relationships and culture. Boston: Pearson Education. Rodgers, K.B. & Rose, H.A. (2002). Risk and resiliency factors among adolescents who experience marital transitions. Journal of Marriage and the Family, 64(4)1024-1037. Retrieved September 14, 2004, from Academic Search Premier database. Senel, H.G. & Akkök, F. (1996). Stress levels and attitudes of normal siblings of children with disabilities. International Journal for the Advancement of Counselling, 18, 61-68. Shek, D.T.L. (2002). The relation of parental qualities to psychological well-being, school adjustment and problem behavior in Chinese adolescents with economic disadvantage. American Journal of Family Therapy, 30(3), 215- 230. Sim, H. (2000). Relationship of daily hassles and social support to depression and antisocial behavior among early adolescents. Journal of Youth and Adolescence, 29 (6), 647-659. 137 Statistics South Africa (1998). Marriages and Divorces. Pretoria: Author. Steyn, A.F. (1993, 1994). Family structures in the RSA. Pretoria: Human Sciences Research Council. Steyn, H.S. (1999). Praktiese beduidendheid: Die gebruik van effekgroottes. Potchefstroom: Publikasiebeheerkomitee, PU vir CHO. Surajnarayan, G. (1991). The influence of the family on the adolescent’s behaviour. Unpublished master’s dissertation. Pretoria: University of South Africa. Thomas, T.J. & Daubman, K.A. (2001). The relationship between friendship quality and self -esteem in adolescent girls and boys. Sex roles: A Journal of Research. Retrieved October 6, 2004, from http://www.findarticles.com/p/articles/mi_m2294/is_2001_July/ai_81478074/ Updegraff, K.A., Mchale, S.G., Grouter, A.C. & Kupanoff, K. (2001). Parents’ involvement in adolescents’ peer relationships: A comparison of mothers’ and fathers’ roles. Journal of Marriage and the Family, 63(3), 655-668. Van Wel, F., ter Bogt, T. & Raaijmakers, Q. (2002). Changes in the parental bond and the well-being of adolescents and young adults. Adolescence, 37(146), 317-333. Retrieved September 14, 2004, from Academic Search Premier database. Visser, M. (1991). The psychological effect of parents’ unemployment on children. Welfare Focus, 26 (2), 38-41. Wan, C.K., Jaccard, J. & Ramey, S.L. (1996). The relationship between social support and life satisfaction as a function of family structure. Journal of Marriage and the Family, 58, 502-513. 138 Way, N. & Robinson, M.G. (2003). A longitudinal study of the effects of family, friends, and school experiences on the psychological adjustment of ethnic minority, low -SES adolescents. Journal of Adolescent Research, 18 (4), 324- 346. Wong, E.H. & Wiest, D.J. (1999). Adolescent depression: Links to academic coping and perceived autonomy support. Education, 119(4), 668-673. Retrieved September 14, 2004, from Academic Search Premier database. Yeung, W.J., Sandberg, J.F., Davis-Kean, P.E. & Hofferth, S.L. (2001). Children’s time with fathers in intact families. Journal of Marriage and the Family, 63(1), 136-154. Retrieved March 1, 2005, from Academic Search Premier database. Young, M.H., Miller, B.C., Norton, M.C. & Hill, E.J. (1995). The effect of parental supportive behaviors on life satisfaction of adolescent offspring. Journal of Marriage and the Family, 57(3), 813-822. 139 ARTICLE FIVE THE INFLUENCE OF SELECTED FAMILY VARIABLES ON SELF-ESTEEM AND LIFE SATISFACTION AMONG BLACK ADOLESCENTS 140 TABLE OF CONTENTS INTRODUCTION...........................................................................143 LITERATURE REVIEW ...............................................................144 Self-esteem.............................................................................................................144 Life satisfaction.....................................................................................................147 METHODOLOGY .........................................................................150 RESULTS AND DISCUSSION.......................................................152 CONCLUSION...............................................................................160 LIST OF REFERENCES................................................................163 LIST OF TABLES Page Table 1 Cronbach’s α -coefficient for the Rosenberg Self-esteem Scale 150 Table 2 Cronbach’s α -coefficient for the Satisfaction with Life Scale 151 Table 3 Distribution of group in terms of family variables 153 Table 4 Results of ANOVAs with self-esteem as dependent variable 155 Table 5 Results of t-tests with self-esteem as dependent variable 156 Table 6 Results of ANOVAs with life satisfaction as dependent variable 157 Table 7 Scheffé results regarding household size 158 Table 8 Results of t-tests with life satisfaction as dependent variable 159 Abstract Earlier research has shown the importance of self-esteem in adolescent well- being. In addition, life satisfaction is also considered a key element of subjective well-being. This study aimed to supplement the scarce existing data on self-esteem, and life satisfaction in particular, among black South African adolescents. Self-esteem scores among respondents were acceptable. In terms of life satisfaction, respondents indicated that they were slightly satisfied with their lives. No statistically significant relationships were found between self-esteem on the one hand and family structure, parents’ marital status, parents’ educational status, parents’ empl oyment status, household size or time spent with parents on the other hand. The only variable that showed a statistically significant relationship with self -esteem was that of taking care of household members with severe disabilities or serious chronic illnesses. Participants with such members in their households reported significantly lower self-esteem. No statistically significant relationships were found between family structure, parents’ marital status, parents’ educational level, employment status of mothers and taking care of household members with severe disabilities or serious chronic illnesses on the one hand, and life satisfaction on the other hand. Results showed that adolescents in larger households experienced significantly higher levels of life satisfaction than those in smaller households. In addition, employed fathers and more time spent with parents were linked to higher levels of life satisfaction. INTRODUCTION Given the relatively vulnerable nature of adolescence, it is not surprising that adolescent self-esteem has received considerable research attention. High self -esteem has indeed been found to be the most important buffer against daily stressors in adolescence (Dumont & Provost, 1999). Numerous variables that may impact on adolescent self-esteem have been investigated, among them several that relate to family factors such as the family’s socio -economic status and the marital status of parents. For example, in their review of literature, Chubb, Fertman and Ross (1997) found positive correlations between self-esteem and perceived intimacy with parents, higher family income, life satisfaction, academic achievement and a positive body image. Researchers are focusing increasingly on the subjective well-being of individuals as opposed to psychopathology, which has received most of the research attention in the past (Gilligan & Huebner, 2002; Gilman, 2001; Huebner, Suldo, Smith & McKnight, 2004). Life satisfaction is considered a key element of subjective well-being (Diener & Diener, 1995) and has in fact been included as a predictor of people’s well-being 143 (Gilligan & Huebner, 2002). Unfortunately most research on life satisfaction has focused on adults (Gilman, 2001; Suldo & Huebner, 2004; Valois, Zullig, Huebner & Drane, 2004) . As is the case with African-American adolescents (Chapman & Mullis, 2000), black South African adolescents do not only struggle with the normal strains of adolescence, but often also with complications such as poverty and racism. South Africa’s history in respect of the racist political system of apartheid, has led to problems such as low self-esteem among blacks (Boulter, 1995; Hickson & Kriegler, 1991). It is therefore understandable that, during the apartheid years, black South Africans achieved significantly lower scores than their white counterparts in terms of their perception of general well-being (Møller, 1988). As there is a dearth of information on self -esteem and particularly life satisfaction among black South African adolescents, this study has attempted to expand the existing database. More specifically, the main goal was to determine the relationship between self-esteem and life satisfaction, on the one hand, and selected family variables on the other, among black adolescents. LITERATURE REVIEW An overview of prominent findings regarding self-esteem and life satisfaction in adolescence will be provided next. Self-esteem Numerous variables have been linked with self-esteem, i.e. one’s opinion or evaluation of oneself. Much research has revolved around the relationship between family relationships and self-esteem. Van Wyk (1998) found a positive relationship between family functioning and self-esteem and Mahabeer (1993) linked higher expressiveness in families with higher self -esteem. In Brage and Meredith’s (1994) research, family strengths were found to have the strongest impact of all variables on self-esteem. Taylor et al. (1997) established that adolescents who have higher 144 perceived levels of family responsibility also reported higher self -esteem. This is probably owing to the fact that family responsibility creates a sense of importance and of being valued by the family. Regarding parental relationships specifically, interesting results have been obtained. Ferreira and Monyemorathwe (1993) indicate that a child needs parental love to develop a positive self-concept. Portes and Zady (2002) found parent -child conflict to be a predictor of self-esteem among Spanish-speaking Americans, and Pawlak and Klein (1997) established a link between inter-parental conflict and low self -esteem. Williams (1999) demonstrated a significant positive correlation between parental involvement and self-esteem. Nillsen and Metha (1994) replicated these findings by establishing a positive relationship between parental support and self-esteem in non- clinical adolescents. Mboya (1998) established a similar link between parental support and self-concept among black South African adolescents. A correlation between high perceived parental abuse and low self-esteem (Blake & Slate, 1993) lends further support to the hypothesis that parent-adolescent relationship factors impact strongly on adolescent self-esteem. Pawlak and Klein (1997) and Ohanessian, Lerner, von Eye and Lerner (1996) found both paternal and maternal nurturing and acceptance respectively to impact positively on self-esteem. Field, Lang, Yando and Bendell’s (1995) study indicated that, although intimacy with both mothers and fathers is linked to higher self-esteem, intimacy with mother s carries the most weight. Ohanessian et al. (1996) found that the adolescent-parent relationship is more important in the development of self-esteem among girls than among boys. A similar finding was reported by Heinonen, Räikkonen and Keltikangas-Järvinen (2003), namely that hostile maternal attitudes in childhood are more strongly linked to low self-esteem in adolescence among girls than among boys. Contrary to these findings, Mboya (1998) found that parental behaviours towards boys were more important than parental behaviours towards girls in terms of the development of a self-concept. Other family variables, such as socio-economic status, have also been investigated. Van Heerden (1995) found that adolescents in bigger families have a more positive self-concept. In her study, the educational level of parents was found to have no impact on the development of self-concept among adolescents. Holland and Andre (1994) found no significant relationship between parents’ marital status, parents’ 145 occupational status and parents’ educational status, on the one hand, and adolescent self-esteem, on the other. They propose that personal characteristics (e.g. participating in extracurricular activities) are more important in the development of self-esteem than environmental traits (e.g. parents’ socio-economic status). Among participants in their study, Corcoran and Franklin (2002) found that self-esteem was highest among adolescents who lived in intact families, followed by adolescents in single -parent families. Although the majority of research projects investigated family relationship variables, peer support also received attention. Although both family support and peer support correlate positively with adolescent self-esteem (Dumont and Provost, 1999), family support appears to be the more important variable of the two (Field et al., 1995; van Heerden, 1995; van Wyk, 1998). On the other hand, McFarlane, Bellissimo and Norman (1995) found that peer support impacted more on social self-efficacy (an element of self-esteem) than did family support. Several studies report an inverse relationship between depression and self-esteem in adolescence. Brage and Meredith (1994) and Southall and Roberts (2002) demonstrated that self -esteem impacts on adolescent depression, whereas Portes and Zady (2002) found depression to be a key predictor in respect of self-esteem. Harter (1993) also found a link between self-esteem and depression, but cautions that the directionality of this link is not clear. DeSimone, Murray and Lester (1994), Marcotte, Fortin, Potvin and Papillon (2002) and Dori and Overholser (1999) also demonstrated an inverse relationship between depression and self-esteem. Houlihan, Fitzgerald and O’Regan (1994) consider self-esteem to be an important buffer against adolescent depression. Many research projects show that self-esteem tends to be higher in adolescent boys than adolescent girls (Brage & Meredith, 1994; Chubb et al., 1997; Marcotte et al., 2002). Nillsen and Metha (1994) found boys to have higher self -esteem than girls in one domain only, namely competence, while no gender differences were found in the other domains investigated in their research. Fernandez and Castro (2003) and Holland and Andre (1994) did not find that gender significantly impacted on self- esteem. 146 Life satisfaction Given the small amount of specific research available in terms of adolescent life satisfaction, prominent findings of studies conducted among adults are also mentioned briefly. Concerning predictors of life satisfaction, Wilson, Henry and Peterson (1997) found subjective experience of life circumstances to be a stronger predictor of life satisfaction than actual circumstances. Schimmack, Diener and Oishi’s (2002) literature review led them to a similar conclusion. They established that people tend to judge their life satisfaction according to relevant aspects of their lives. Moreover, people use different sources to rate global and monthly life satisfaction. What may be considered a source of life satisfaction by one person (e.g., monetary rewards) may not necessarily be considered in the same light by another person. Gilman (2001) established that both personal and environmental factors impact on life satisfaction. Westaway, Maritz and Golele (2003) found objective indicators of life satisfaction, such as demographic variables, to be of less importance than subjective indicators such as personality traits. It has been noted that life satisfaction appears to remain stable over time (Mokgatlhe & Schoe man, 1998; Schimmack et al., 2002; Suldo & Huebner, 2004). Several projects established a link between family factors and life satisfaction among adolescents. Phinney and Ong (2002) found discordancy between adolescents and parents to significantly lower life satisfaction among both Vietnamese Americans and European Americans. Further to this finding, Huebner, Suldo, Smith, et al. (2004), in their review of literature, also found a link between children’s home environments and levels of life satisfaction. Family factors pinpointed by these authors as positively correlating with life satisfaction in children are authoritative parenting, parental support, intactness of the parental marriage and a loving parental relationship. Young, Miller, Norton and Hill (1995) reported a positive link between perceptions of parental support and life satisfaction among adolescents. This link was stronger in the case of perceived maternal support than in that of perceived paternal support, and 147 mothers were generally viewed as the more supportive parents. A significant positive correlation between life satisfaction and the relationship with parents was also reported by Leung and Leung (1992). Shek (2002a) found family functioning to correlate positively with life satisfaction, although the correlation is stronger in poor than in non-poor families. He suggests that this may be owing to the fact that poor families have fewer resources to rely upon, with the result that the importance of family as a resource is enhanced. In a further study, Shek (2002b) established a link between parenting qualities and mental health (which included self -esteem and life satisfaction) in economically disadvantaged Chinese adolescents. This link was stronger for paternal than for maternal parenting qualities. Shek (2002b) ascribes this to the dominant role of the father figure in Chinese families. Grossman and Rowat (1995) demonstrated a link between perceived poor parental relationships and low levels of life satisfaction. It appears tha t adolescent-parent relationships are more important than peer relationships in determining life satisfaction (Huebner, Suldo, Smith, et al., 2004; Leung & Leung, 1992). Most studies indicated that gender does not influence life satisfaction significantly (Diener & Diener, 1995; Huebner, Suldo, Valois, Drane & Zullig, 2004; Leung & Leung, 1992; Phinney & Ong, 2002). Socio-economic status (cf. Huebner, Suldo, Smith, et al., 2004; Phinney & Ong, 2002) was not found to link up significantly with life satisfaction. However, among black, white, coloured and Indian university students, Møller (1996) established a correlation between life satisfaction and income. As race and income are inextricably linked in South Africa, this finding also divided the participants in the study along racial lines. White students with a higher income demonstrated higher life satisfaction than black students with low incomes. Westaway et al. (2003) established a significant positive relationship between life satisfaction on the one hand, and educational and employment status, on the other, among white and black South African adults. In terms of cultural differences Huebner, Suldo, Valois, et al. (2004) reported that Caucasian students experienced higher life satisfaction than African-American adolescents; yet this difference was very small and not of practical significance. 148 Similarly, Phinney and Ong (2002) found no difference in life satisfaction levels among Vietnamese Americans and European Americans. As mentioned ear lier, Møller (1996) did, in fact, find differences in life satisfaction among black and white South African students, although these differences are probably attributable to financial status rather than to actual cultural differences. Schimmack et al. (2002) suggest that culture influences the available sources that people use as criteria in making life satisfaction judgements. For example, in collectivistic cultures, harmonious relationships may be a strong predictor of life satisfaction, whereas individual achievements may be more important in individualistic cultures. Thus, although global life satisfaction may not necessarily vary, the sources of life satisfaction may indeed vary across cultures. The effect of racism on life satisfaction should also be taken into account, especially in the South African context. In their literature review, Utsey, Ponteretto, Reynolds and Cancelli (2000) refer to the consistently negative link that was found between life satisfaction and racism. Yet, in their study, Mokgatlhe and Schoeman (1998) found black students to have reasonably high levels of life satisfaction. It can be assumed that the new political climate in South Africa since 1994 is likely to have increased life satisfaction among black people. Their general expectations are more positive, as they have now been granted equal opportunities. Racism, although not yet absent from South African society, can be expected to play a smaller role in determining the global life satisfaction of blacks in post-apartheid South Africa. Several researchers have indicated a positive relationship between life satisfaction and self-esteem (Diener & Diener, 1995; Kang, Shaver, Sue, Min & Jing, 2003; Wilson et al., 1997). Mokgatlhe and Schoeman (1998) found high scores for collective self- esteem among black South African university students, and indicated that this predicted life satisfaction. The results of the current study will subsequently be presented and discussed. The hypothesis postulates that there are significant differences in terms of life satisfaction and self -esteem scores for the population of black adolescents in terms of a number of independent family variables. 149 METHODOLOGY The research was non-experimental and employed a criterion group design (Huysamen, 1983). Participants in the study were comprised of 2 505 black high school students in grades eight to twelve, attending six different schools in Mangaung, a township near Bloemfontein in the Free State Province. Convenience sampling was employed. Permission to conduct the study was obtained from the Free State Department of Education. A self-compiled biographical questionnaire was aimed at collecting data similar to those obtained by Steyn (1993, 1994) in South Africa and Andrews and Morrison (1997) in the USA. Family variables that were investigated included household size, family structure, the occupational and educational status of parents, the marital status of parents, time spent with parents per week and taking care of household members with severe disabilities or serious chronic illnesses. The widely-used Rosenberg Self-esteem Scale (RSE) (Rosenberg, 1965) was employed to measure self-esteem. Higher scores indicate higher self-esteem. The psychometric properties of this scale were found to be satisfactory in the USA (cf. Brage & Meredith, 2002; Rosenberg, 1965). Table 1 shows measured internal consistency in the current study. Table 1 Cronbach’s α -coefficient for the Rosenberg Self-esteem Scale Construct Number Minimum Maximum α -coefficient of items Self -esteem 10 10 40 0,5984 Unfortunately, the measured internal consistency for this study was not high. Cultural and language difference may have influenced the internal consistency in terms of black South African adolescents. 150 Life satisfaction was measured with the aid of the Satisfaction with Life Scale (SWLS) (Diener, Emmons, Larsen & Griffin, 1985). Psychometric properties of this scale have been well established in the USA (cf. Phinney & Ong, 2002). Table 2 Cronbach’s α -coefficient for the Satisfaction with Life Scale Construct Number Minimum Maximum α -coefficient of items Life satisfaction 5 5 35 0,5931 Table 2 indicates that once again internal consistency was not high in this study. As with the self-esteem scale, cultural and language differences may have resulted in lower measured internal consistency. Owing to big differences in frequencies among the categories of the ten independent variables, it was decided to carry out an analysis of variance (ANOVA) in the case of more than two categories, and a t-test for independent groups in the case of two categories. If multivariate analyses of variance (MANOVAs) had been done, only respondents with fully completed questionnaires would have been included in all analyses. In other words, the maximum number of respondents involved would have been 670 (refer to Table 3), amounting to a mere 31,3% of the total sample. Although the chosen method may lead to an increase in type 1 error, it was nevertheless followed so as to utilize as much of the information obtained from respondents as possible. 151 In addition to determining whether results were statistically significant, the practical significance of such results was also investigated. To this end, effect sizes were calcula ted. When ANOVAs were utilized, the following indexes were used to determine effect sizes (Nolan, 2002): R² = SS between SS total The following guiding values were used: R² = 0,01 was considered a small effect, R² = 0,06 was considered a medium effect, and R² = 0,14 was considered a large effect. As mentioned, t-tests for independent groups were used for variables with two categories. Given that the sizes of the groups differed, it was decided to make use of the pooled variance estimate. According to Howell (2002), this procedure is preferable when sample sizes differ. The effect sizes of the differences between the two sets of averages were calculated by means of determining Cohen’s d (Steyn, 1999). This procedure expresses the difference between the two means in terms of the size of the standard deviation. In other words, the difference has been standardized. Cohen (Nolan, 2002) laid down very general guidelines for what he considered to be small, medium and large effect sizes. He characterized d = 0,20 as a small, but probably meaningful effect size, d = 0,50 as a medium effect (half a standard deviation difference) and d = 0,80 as a large effect. Effect sizes were only calculated when results were found to be significant on either the 1% leve l or the 5% level. RESULTS AND DISCUSSION In order to place the results of the statistical analysis in the right context, the frequencies in terms of the ten selected independent variables are shown in Table 3. 152 Table 3 Distribution of group in terms of family variables Independent variables N % Family structures: Mother and children 374 19,5 Father, mother and children (nuclear) 754 39,2 Male-headed extended family 363 18,9 Female-headed extended family 430 22,4 Subtotal 1921 Marital status: Married 1047 61,4 Never married/Live together 289 16,9 Divorced/Separated/Widowed 371 21,7 Subtotal 1707 Household size: 1 – 3 370 17,3 4 – 6 1399 65,4 7 or more 370 17,3 Subtotal 2139 Educational level – father: None 66 9,9 Grade 1-7 87 13,0 Grade 8 –12 324 48,3 Some form of post-school education 193 28,8 Subtotal 670 Educational level – mother: None 83 7,4 Grade 1-7 134 12,0 Grade 8-12 637 56,9 Some form of post-school education 266 23,7 Subtotal 1120 Employment status – father: Employed 801 84,0 Unemployed 153 16,0 Subtotal 954 153 Table 3 Distribution of group in terms of family variables (continued) Independent variables N % Employment status – mother: Employed 1081 73,9 Unemployed 382 26,1 Subtotal 1463 Time spent with father per week: Up to 6 hours 545 53,3 More than 6 hours 477 46,7 Subtotal 1022 Time spent with mother per week: Up to 6 hours 402 23,5 More than 6 hours 1312 76,5 Subtotal 1714 Households taking care of members with severe disabilities or serious chronic illnesses: Yes 852 41,8 No 1185 58,2 Subtotal 2037 Notable in Table 3 is that, despite the large sample size (2 505 adolescents), the responses in each category mostly do not add up to this total. The reason for this discrepancy probably lies in the fact that in many cases, questionnaires were not completed in full. This may be owing to carelessness; or possibly some of the participants misunderstood certain questions. In this regard it should be borne in mind that the questionnaires were not completed in their home language, but in English. In addition, it is also possible that some participants did not know the answers to certain questions. The average self-esteem score for the group is 29,92, with a standard deviance of 4,32. This score is indicative of acceptable levels of self -esteem among the respondents (Fernandez & Castro, 2003). 154 As mentioned, ANOVAs (for variables with more than two categories) and t-tests for independent groups (for variables with two categories) were used to investigate the influence of family variables on self-esteem among respondents. Table 4 summarizes findings regarding family structure, the marital status of parents, household size and the educational status of mothers and fathers. Table 4 Results of ANOVAs with self-esteem as dependent variable Family variable F P Family structure 0,70 0,5527 Marital status of parents 0,64 0,5250 Household size 0,14 0,8722 Educational level – father 0,66 0,5757 Educational level – mother 1,90 0,1274 ** p = 0,01 * p = 0,05 As can be seen from Table 4, none of the independent variables showed a significant relationship with self-esteem. Contrary to these findings, van Heerden (1995) did, in fact, find a significant positive correlation between family size and the self -concept of adolescents. In terms of the educational status of parents, she found no link between this variable and the self-concept of adolescents. Holland and Andre (1994) also did not find parents’ marital status and educational level to be linked to adolescent self- esteem. Corcoran and Franklin (2002) reported that the highest levels of self -esteem in their study were to be found among adolescents living with both parents, followed by adolescents in single-parent families. This suggests that family structure and the marital status of parents do play a role in the development of self-esteem. It is possible that aspects such as high levels of parental conflict have a greater impact on adolescent self-esteem than the actual demographic aspects of the family (Pawlak & Klein, 1997). The adolescent’s subjective perceptions of his family’s functioning may carry more weight than the reality of family functioning in terms of the development of self-esteem. For independent variables consisting of two groups, t-tests for independent groups were conducted. The results of these tests are shown in Table 5. 155 Table 5 Results of t-tests with self-esteem as dependent variable Family variable N X S t p Cohen D Employment status – father: Employed 616 30,17 4,41 Unemployed 120 29,69 4,04 1,10 0,2719 - Employment status – mother: Employed 834 30,09 4,36 Unemployed 280 30,13 4,27 -0,13 0,8975 Time spent with father: up to six hours 414 29,93 4,44 more t han six hours 363 30,25 4,21 -1,02 0,3086 Time spent with mother: up to six hours 305 29,81 4,49 more than six hours 984 30,11 4,22 -1,05 0,2961 Taking care of members with severe disabilities or serious chronic illnesses: Yes 635 29,29 4,29 No 905 30,39 4,31 -4,93 0,0000** 0,26 ** p = 0,01 * p = 0,05 Table 5 indicates that there is a significant relationship between self-esteem and taking care of household members with severe disabilities or chronic illnesses (p = 0,01). Adolescents in such households have significantly lower self-esteem than those in households who do not take care of disabled or chronically ill members. The corresponding effect size is small. Siblings of children with disabilities and chronic illnesses often experience feelings such as resentment, guilt and embarrassment (Makola, 1994). They often feel that they have more than their fair share of responsibility and that they are being neglected (Senel & Akkök, 1996). In addition, they may have a need for support and for an opportunity to express their concerns (cf. Naylor & Prescott, 2004). It is thus likely that such a situation in a household can lead to lowered self-esteem in the involved adolescents. Research has linked father absence with reduced self-esteem in adolescents (Magane, 2000). However, the current study did not support these findings. Less than six hours a week of contact with parents does not necessarily imply that this is perceived as 156 parental absence by the adolescent. If contact with parents is minimal owing to factors such as migratory labour, this is probably not perceived by the adolescent as rejection, but as an inevitable situation that does not take away from the quality of the parent-adolescent relationship. The absence of significant findings in terms of self-esteem and family variables does not necessarily mean that the family is not important for adolescent well-being. It is more likely that aspects other than those measured also play a role in adolescent well- being, for example, parent-child conflict (Portes & Zady, 2002) and spousal conflict (Pawlak & Klein, 1997). In addition, the sources of self-esteem may vary among different groups. For example, Corcoran and Franklin (2002) found teenage pregnancy to be a source of self-esteem among low socio-economic -status adolescents, but not among high socio -economic -status adolescents. The average score for life satisfaction is 24,13, with a standard deviance of 6,22. This score indicates that the adolescents in the study are slightly satisfied with their lives (cf. Pavot & Diener, 1993). The results of the analysis of variance, with satisfaction with life as the dependent variable, are shown in Table 6. Table 6 Results of ANOVAs with life satisfaction as dependent variable Family variable F P R² Family structure 1,23 0,2961 - Marital status of parents 2,36 0,0951 - Household size 5,18 0,0057** 0,005 Educational level – father 0,61 0,6078 - Educational level – mother 0,16 0,9228 - ** p = 0,01 * p = 0,05 Table 6 indicates that significant differences were found in terms of satisfaction with life and household size (p = 0,01). The corresponding effect size is small. A post t- 157 test was nevertheless conducted by means of the Scheffé procedure in an attempt to identify specific group differences. The results thereof appear in Table 7. Table 7 Scheffé results regarding household size Groups that are N X s Groups different Group 1 from 3 343 23,57 6,61 1 - 3 (group 1) Group 2 from 3 1311 24,03 6,23 4 – 6 (group 2) 349 25,03 5,69 7 or more (group 3) According to the results contained in Table 7, adolescents in big households (seven or more members) experience significantly higher levels of life satisfaction than adolescents in households with six or fewer members. Interpretation of this result should be conducted cautiously. Firstly, not all household members are necessarily close – or blood relations. Secondly, a large household does not necessarily imply a loss of parental atte ntion, but may actually imply increased attention from different persons in the household. Thus, a larger household can lead to increased life satisfaction. The quality of the home environment of adolescents has been linked to perceived life satisfaction (cf. Huebner, Suldo, Smith, et al., 2004). In addition, earlier research established that subjective perception of the quality of the household is more important in determining life satisfaction than a demographic variable such as actual household size (Westaway et al., 2003). 158 For independent variables consisting of two categories, t-tests were conducted; and these results appear in Table 8. Table 8 Results of t-tests with life satisfaction as dependent variable Family variable N X s t P Cohen D Employment status – father: Employed 750 24,85 5,87 Unemployed 145 23,57 6,20 2,39 0,0168* 0,22 Employment status – mother: Employed 1014 24,35 6,11 Unemployed 358 24,64 5,89 -0,79 0,4307 - Time spent with father: up to six hours 511 23,95 5,78 more than six hours 457 24,95 6,31 -2,59 0,0098** 0,17 Time spent with mother: up to six hours 376 23,66 6,25 more than six hours 1253 24,54 6,12 -2,43 0,0151* 0,14 Taking care of members with severe disabilities or serious chronic illnesses: Yes 800 24,10 6,14 No 1123 24,24 6,16 -0,50 0,6139 - ** p = 0,01 * p = 0,05 According to Table 8, significant relationships were found between life satisfaction and three family variables, namely time spent with mothers, the employment status of fathers (p = 0,05) and time spent with fathers (p ≤ 0,01). Corresponding effect sizes are small. The fact that adolescents who spend more time with their parents report higher life satisfaction is informative. Young et al. (1995) indicated that adolescent perception of parental support is positively correlated with life satisfaction. Phinney and Ong (2002) also found the parent-adolescent relationship to influence life satisfaction. Spending more time with parents is probably viewed in a positive light by the adolescents in this study and also appears to improve the quality of the parent- child relationship, thereby increasing life satisfaction. 159 Respondents with employed fathers also reported higher life satisfaction than those with unemployed fathers. Employment status is a determinant of socio-economic status; and the socio-economic status of the family of origin has been found to be a predictor of life satisfaction for Appalachian young adults (Wilson et al., 1997). Westaway et al. (2003) reported that life satisfaction among black and white adults in South Africa is related to employment status. Diener and Diener (1995) also established that satisfaction with finances correlates significantly with life satisfaction among adults in different countries. It may thus be true that fathers who are employed experience higher life satisfaction, which in turn impacts positively on the experience of life satisfaction among other family members. The fathe r’s employment status may be more important than the mother’s employment status. This may be firstly because many males in South Africa still earn more than females and secondly, because many rural traditional black families are still patrilinear in nature. CONCLUSION In this study the only variable that correlated significantly with self-esteem was that of caring for disabled/chronically ill household members. Household size, time spent with mothers and fathers and the employment status of fathers showed positive correlations with life satisfaction. A finding of particular interest is that although adolescents perceive more time spent with parents as positive (thereby increasing their life satisfaction), less time spent with parents does not necessarily have a negative impact on their self-esteem. This suggests, firstly, that the reasons why parents do not spend time with their children are possibly taken into account by the children and secondly, that the overall quality of the parent-adolescent relationship carries more weight than specific family variables. For example, if a parent spends less time with an adolescent because of job demands, the adolescent probably does not perceive this as a lack of caring on the part of the parent. On the other hand, if the adolescents could choose, they would prefer to spend more time with their parents. This confirms that parental involvement is important to the adolescents in this study. 160 The fact that significant links were not established between self-esteem and most family variables (except that of disabled/chronically ill household members), does not imply that family variables are not of importance. As a matter of fact, this points towards the possibility that there may be other mediating factors in the relationship between self -esteem and family variables, such as inter-parental conflict (Pawlak & Klein, 1997), parent-child conflict (Portes & Zady, 2002) and paternal and maternal nurturing (Ohanessian et al., 1996). According to Baumeister, Campbell, Krueger and Vohs (2005), findings regarding self-esteem should be interpreted with caution, as self-report questionnaires are usually utilized to assess self-esteem. Such findings thus merely reflect how people rate themselves and do not take into account how people rate each other. Furthermore, research has not been able to establish conclusively that adolescents with low self -esteem are more prone to conduct problems such as substance abuse and early sexual activity. The possibility that self- esteem is not as important as the literature has indicated it to be, should therefore be considered. A shortcoming of the research is that self-esteem was not investigated separately for boys and girls. If this had been done, the research might have yielded more significant results. A longitudinal study could yield interesting results concerning changes in self -esteem and life satisfaction over time. The internal consistencies for both the self-esteem scale and the life satisfaction scale were not high. This is probably owing to the fact that questionnaires were not completed in the participants’ home language, and may have influenced results. The participants in this study are all high school students; thus, the sample group can be considered a select group. Results can thus not simply be generalized to include all adolescents in Mangaung. As time spent with parents appears to be of importance to the adolescents in this study, future research should focus on the characteristics of the parent -adolescent relationship. In addition, the factors that influence the relationship between family variables on the one hand and self-esteem and life satisfaction on the other hand should also be investigated. Sources of self-esteem and life satisfaction for black adolescents should be explored. The nature of disabilities and chronic illnesses that blacks have to cope with in their families, and the impact thereof, particularly on children in the family, should also be researched. 161 By and large, it appears that black families are coping better than one might have expected. However, there is still much to be done to improve and strengthen the family as an important pillar of society. Billingsley’s (1992) comment that the African-American family is both strong and weak, also applies to the South African black family. In addition, it also seems as if the adolescents in the current study are coping reasonably well. It should be borne in mind, though, that positivity bias might have led them to give positive answers to questions, even if these answers are not necessarily true in reality. According to the Department of Social Development (2003), campaigning for a supportive family life for all members of society is essential. Hopefully, studies such as the present one will continue to investigate challenges facing black South African families, thereby not only relieving the plight of these families, but also strengthening the already-present resilience which has proved to be so valuable. This in turn will improve the psychological well-being of children and adolescents in these families. 162 LIST OF REFERENCES Andrews, A.B. & Morrison, B. (1997). The changing American family: Impact on South Carolina’s children. Columbia: Institute for Families in Society. Baumeister, R.F., Campbell, J.D., Krueger, J.I. & Vohs, K.D. (2005). Exploding the self-esteem myth. Scientific America Online. Retrieved January 10, 2005, from http://www.freerepublic.com/focus/f-news/1306149/posts Billingsley, A. (1992). Climbing Jacob’s ladder: The enduring legacy of African- American families. New York: Simon & Schuster. Blake, P.C. & Slate, J.R. (1993). A preliminary investigation into the relationship between adolescent self-esteem and parental verbal interaction [Electronic version]. School Counselor, 41(2), 81-85. Boulter, S.L. (1995). Factors affecting standard six adjustment to high school. Unpublished master’s dissertation. Bellville: University of the Western Cape. Brage, D. & Meredith, W. (1994). A causal model of a dolescent depression. The Journal of Psychology, 128(4), 455-468. Chapman, P.L. & Mullis, R.L. (2000). Racial differences in adolescent coping and self-esteem. Journal of Genetic Psychology, 161(2), 152-160. Retrieved September 14, 2004, from Academic Search Premier database. Chubb, N.H., Fertman, C.I. & Ross, J.L. (1997). Adolescent self-esteem and locus of control: A longitudinal study of gender and age differences. Adolescence, 32(125), 113-129. Retrieved September 14, 2004, from Academic Searc h Premier database. 163 Corcoran, J. & Franklin, C. (2002). Multi-systemic risk factors predicting depression, self-esteem, and stress in low SES and culturally diverse adolescents. Journal of Human Behavior in the Social Environment, 5(2), 61-76. Department of Social Develoment (2003). 1st Draft baseline document for the development of a national policy for families. Retrieved February 24, 2005, from http://www.socdev.gov.za/Documents/2004/IYF/1base.htm DeSimone, A., Murray, P. & Lester, D. (1994). Alc ohol use, self-esteem, depression and suicidality in high school students. Adolescence, 29(116), 939-942. Retrieved September 14, 2004, from Academic Search Premier database. Diener, E. & Diener, M. (1995). Cross-cultural correlates of life satisfaction and self- esteem. Journal of Personality and Social Psychology, 68 (4), 653-663. Diener, E., Emmons, R.A., Larsen, R.J. & Griffin, S. (1985). The Satisfaction with Life scale. Journal of Personality Assessment, 49(1), 71-75. Dori, G.A. & Overholser , J.C. (1999). Depression, hopelessness and self-esteem: Accounting for suicidality in adolescent psychiatric inpatients. Suicide and Life-threatening Behavior, 29(4), 309-318. Dumont, M. & Provost, M.A. (1999). Resilience in adolescents: Protective role of social support, coping strategies, self-esteem, and social activities on experience of stress and depression. Journal of Youth and Adolescence, 28(3), 343-363. Fernandez, M.L. & Castro, Y.R. (2003). Age and sex differences in self-esteem among Spanish adolescents. Psychological Reports, 93, 876-878. Ferreira, G.V. & Monyemorathwe, S.M. (1993). The influence of parents and peers on the self-concept formation of the urban black adolescent. Journal of Pedagogics, 14 (1), 8-14. 164 Field, T., Lang, H.C., Yando, R. & Bendell, D. (1995). Adolescents’ intimacy with parents and friends. Adolescence, 30(117), 133-140. Retrieved September 14, 2004, from Academic Search Premier database. Gilligan, T.D. & Huebner, E.S. (2002). Multidimensional life satisfaction reports of adolescents: A multitrait -multimethod study. Personality and Individual Differences, 32 (7), 1149-1155. Gilman, R. (2001). The relationship between life satisfaction, social interest, and frequency of extracurricular activities among adolescent students. Journal of Youth and Adolescence, 30(6), 749-767. Grossman, M. & Rowat, K.M. (1995). Parental relationships, coping strategies, received support, and well-being in adolescents of separated or divorced and married parents. Research in Nursing and Health, 18, 249-261. Harter, S. (1993). Causes and consequences of low self -esteem in children and adolescents. In R.F. Baumeister (Ed.), Self-esteem: The puzzle of low self- regard. New York: Plenum Press. Heinonen, K., Räikkönen, K. & Keltikangas-Järvinen, L. (2003). Maternal perceptions and adolescent self-esteem: A six-year longitudinal study. Adolescence, 38 (152), 669-687. Hickson, J. & Kriegler, S. (1991). Childshock: The effects of apartheid on the mental health of South Africa’s children. International Journal for the Advancement of Counselling, 14 (2), 141-154. Holland, A. & Andre, T. (1994). The relationship of self-esteem to selected personal and environmental resources of adolescents. Adolescence, 29(114), 345-360. Retrieved September 14, 2004, from Academic Search Premier database. Houlihan, B., Fitzgerald, M. & O’Regan, M. (1994). Self -esteem, depression and hostility in Irish adolescents. Journal of Adolescence, 17, 565-577. 165 Howell, D.C. (2002). Statistical methods for psychology (5th ed.). Johannesburg: Duxbury. Huebner, E.S., Suldo, S.M., Smith, L.C. & McKnight, C.G. (2004). Life satisfaction in children and youth: Empirical foundations and implications for school psychologists. Psychology in the Schools, 41 (1), 81-93. Huebner, E.S., Suldo, S., Valois, R.F., Drane, J.W. & Zullig, K. (2004). Brief multidimensional students’ life satisfaction scale: Sex, race, and grade effects for a high school sample. Psychological Reports, 94, 351-356. Huysamen, G.K. (1983). Inferensiële statistiek en navorsingsontwerp: ‘n Inleiding. Kaapstad: Academica. Kang, S.M., Shaver, P.R., Sue, S., Min, K.H. & Jing, H. (2003). Culture-specific patterns in the prediction of life satisfaction: Roles of emotion, relationship quality and self-esteem. Personality and Social Psychology Bulletin, 29(12), 1596-1608. Leung, J. & Leung, K. (1992). Life satisfaction, self -concept, and relationship with parents in adolescence. Journal of Youth and Adolescence, 21(6), 653-665. Magane, M.S. (2000). Father absence: Psychological experiences of black, rural adolescents. Unpublished master’s dissertation. Pretoria: University of Pretoria. Mahabeer, M. (1993). Correlations between mothers’ and children’s self-esteem and perceived familial relationships among intact, widowed, and divorced families. Psychological Reports, 73, 483-489. Makola, D. (1994). The effects of a chronic disorder on the child, his family and environment. South African Paediatric Medicine, 11, 24-30. 166 Marcotte, D., Fortin, L., Potvin, P. & Papillon, M. (2002). Gender differences in depressive symptoms during adolescence: Role of gender-typed characteristics, self-esteem, body image, stressful life events, and pubertal status. Journal of Emotional and Behavioral Disorders, 10(1), 29-42. Retrieved October 6, 2004, from http://www.findarticles.com/p/articles/mi_m0FCB/is_1_10/ai_83790824/ Mboya, M.M. (1998). Family relations and the self-concept of African adolescents: Gender-related differences. Journal of Comparative Family Studies, 29(1), 201-213. McFarlane, A.H., Bellissimo, A. & Norman, G.R. (1995). The role of family and peers in social self-efficacy: Links to depression in adolescence. American Journal of Orthopsychiatry, 65 (3), 402-410. Mokgatlhe, B.P. & Schoeman, J.B. (1998). Predictors of satisfaction with life: The role of racial identity, collective self-esteem and gender -role attitudes. South African Journal of Psychology, 28(1), 28-35. Møller, V. (1988). The relevance of personal domain satisfaction for the quality of life in South Africa. South African Journal of Psychology, 18, 69-75. Møller, V. (1996). Life satisfaction and expectations for the future in a sample of university students: A research note. South African Journal of Sociology, 27(1), 16-25. Naylor, A. & Prescott, P. (2004). Invisible children? The need for support groups for siblings of disabled children. British Journal of Special Educa tion, 31 (4), 199-206. Nillsen, D.M. & Metha, A. (1994). Parental behaviour and adolescent self-esteem in clinical and nonclinical samples. Adolescence, 29(115), 525-542. Retrieved September 14, 2004, from Academic Search Premier database. 167 Nolan, S. (2002). Effect size. Retrieved November 30, 2004, from http://members.aol.com/johnp71/javastat.html Ohanessian, C.M., Lerner, R.M., von Eye, A. & Lerner, J. (1996). Direct and indirect relations between perceived parental acceptance, perceptions of the self, and emotional adjustment during early adolescence. Family and Consumer Sciences Research Journal, 25 (2), 159-183. Pavot, W. & Diener, E. (1993). Review of the Satisfaction with Life Scale. Psychological Assessment, 5(2), 164-172. Pawlak, J.L. & Klein, H.A. (1997). Parental conflict and self-esteem: The rest of the story. Journal of Genetic Psychology, 158 (3), 303-313. Retrieved September 14, 2004, from Academic Search Premier database. Phinney, J.S. & Ong, A.D. (2002). Adolescent-parent dis agreements and life satisfaction in families from Vietnamese and European-American backgrounds. International Journal of Behavioral Development, 26 (6), 556- 561. Portes, P.R. & Zady, M.F. (2002). Self-esteem in the adaptation of Spanish-speaking adolescents: The role of immigration, family conflict and depression. Hispanic Journal of Behavioral Sciences, 24(3), 296-318. Rosenberg, M. (1965). Society and the adolescent self -image. New Jersey: Princeton University Press. Schimmack, U., Diener, E. & Oishi, S. (2002). Life-satisfaction is a momentary judgement and a stable personality characteristic: The use of chronically accessible and stable sources. Journal of Personality, 70(3), 345-384. Senel, H.G. & Akkök, F. (1996). Stress levels and attitudes of normal siblings of children with disabilities. International Journal for the Advancement of Counseling, 18, 61-68. 168 Shek, D.T.L. (2002a). Family functioning and psychological well-being, school adjustment, and problem behavior in Chinese adolescents with and without economic disadvantage. Journal of Genetic Psychology, 163 (4), 497-502. Shek, D.T.L. (2002b). The relation of parental qualities to psychological well-being, school adjustment and problem behavior in Chinese adolescents with economic disadvantage. American Journal of Family Therapy, 30(3), 215- 230. Southall, D. & Roberts, J.E. (2002). Attributional style and self-esteem in vulnerability to adolescent depressive symptoms follow ing life stress: A 14- week prospective study. Cognitive Therapy and Research, 26(5), 563-579. Steyn, A.F. (1993, 1994). Family structures in the RSA. Pretoria: Human Sciences Research Council. Steyn, H.S. (1999). Praktiese beduidendheid: Die gebruik van effekgroottes. Potchefstroom: Publikasiebeheerkomitee, PU vir CHO. Suldo, S.M. & Huebner, E.S. (2004). Does life satisfaction moderate the effects of stressful life events on psychopathological behavior during adolescence? School Psychology Quarterly, 19(2), 93-105. Taylor, S., Field, T., Yando, R., Gonzalez, K.P., Harding, J., Lasko, D., Mueller, C. & Bendell, D. (1997). Adolescents’ perceptions of family responsibility-taking. Adolescence, 32 (128), 969-976. Retrieved September 14, 2004, from Academic Search Premier database. Utsey, S.O., Ponteretto, J.G., Reynolds, A.L. & Cancelli, A.A. (2000). Racial discrimination, coping, life satisfaction, and self-esteem among African Americans. Journal of Counseling and Development, 78(1), 72-80. 169 Valois, R.F., Zullig, K.J., Huebner, E.S. & Drane, J.W. (2004). Physical activity behaviors and perceived life satisfaction among public high school adolescents. Journal of School Health, 74(2), 59-65. Van Heerden, M. (1995). Die verband tussen ‘n groep laat-adolessente se persepsie van hul gesinsfunksionering en hul selfkonsepontwikkeling. Ongepubliseerde meestersverhandeling. Stellenbosch: Universiteit van Stellenbosch. Van Wyk, N.L. (1998). Gesinsfunksionering en die selfkonsepontwikkeling van ‘n groep adolessente. Ongepubliseerde meestersverhandeling. Stellenbosch: Universiteit van Stellenbosch. Westaway, M.S., Maritz, C. & Golele, N.J. (2003). Empirical testing of the Satisfaction with Life Scale: A South African pilot study. Psychological Reports, 92(2), 551-554. Williams, N.B. (1999). Self-esteem and scholastic functioning: The impact of parental involvement, parent-child interactions and socio-economic status. Unpublished master’s dissertation. Johannesburg: University of the Witwatersrand. Wilson, S.M., Henry, C.S. & Peterson, G.W. (1997). Life satisfaction among low- income rural youth from Appalachia. Journal of Adolescence, 20, 443-459. Young, M.H., Miller, B.C., Norton, M.C. & Hill, E.J. (1995). The effect of parental supportive behaviors on life satisfaction of adolescent offspring. Journal of Marriage and the Family, 57(3), 813-822. 170 SUMMARY The black South African family has been neglected in research, although attempts have been made in recent years to rectify this situation. In addition, little information exists regarding black adolescents in South Africa. The aim of this research project was thus to supplement existing data on black adolescents within their family contexts. To this end the relationship between selected family variables and the psychological well-being of black adolescents was investigated. The traditional black South African family and socio-political factors impacting on it, was explored through a literature review. Subsequently, data was colle cted from 2 505 black adolescents. Information regarding the following family variables were obtained: family structure, marital status of parents, household size, educational status of both parents, employment status of both parents, time spent with eac h parent, and taking care of household members who are severely disabled or chronically ill. Variables relating to psychological well-being of adolescents that were measured, were depression, perceptions of social support, life satisfaction and self-esteem. Regarding family variables, interesting trends were observed. The extended family was found to be the predominant family structure, although only marginally so. Divorce rates were low and the average household size was larger than expected. Furthermore adolescents spend considerably more time with their mothers than with their fathers. A majority of adolescents indicated that they do not spend enough time with their fathers. A large proportion of participants indicated that they are caring for severely disabled or chronically ill members in their households. A high depression rate of 26% was found. Results showed that family variables had virtually no significant impact on adolescent depression. Maternal employment and the presence of household members who are severely disabled or chronically ill were the only two variables that showed statistically significant relationships with depression. 171 It seemed that adolescents experience less perceived support in female -headed extended families. Intact families, employed fathers, and more time spent with both mothers and fathers appeared to be linked to higher levels of perceived familial support. Identified trends appear to suggest that what is generally considered to be a stable family environment (e.g. intact family and sufficient family time) has a positive impact on adolescents’ well-being. The only family variable that showed a statistically significant relationship with self- esteem was taking care of household members with severe disabilities or chronic illnesses. The study also found that larger households, paternal employment and more time spent with parents were statistically significantly linked to higher levels of life satisfaction. Results of the study revealed that the black families in this study are reasonably stable despite pressures exerted on it. This confirms the existence of innate strengths within black families, and shows their resilience. Several prominent issues that deserves further attention both in terms of future research and in terms of intervention strategies, came to light in the course of the research. The high depression rate amongst adolescents needs serious attention. Time spent with parents also deserves further consideration. Black families need assistance in overcoming the negative effects of South Africa’s socio-political history and in finding a synthesis between their traditional family values and the demands of modern society. In this regard, inherent strengths of black families need to be investigated. Finally, the emergence of new family structures, and the causes and effects thereof, should be examined. The current project adds greatly to the existing database on black families and the psychological well-being of black adolescents. It is hoped that this will stimulate further research in this field, and that information will be utilized in designing and steering intervention programmes. 172 Key words psychological well-being of adolescents; black adolescents; adolescent self-esteem; adolescent depression; adolescent perceptions of social support; adolescent life satisfaction; black family structure; family variables and psychological well-being; black family variables 173 OPSOMMING Die swart Suid-Afrikaanse gesin het in die verlede weinig aandag van navorsers ontvang, en pogings om hierdie toedrag van sake aan te spreek het eers redelik onlangs begin. Min data bestaan ve ral ook ten opsigte van swart adolessente. Die huidige navorsing het gepoog om die beskikbare inligting rakende swart adolessente binne gesinsverband aan te vul. Die spesifieke doel van die projek was om die verband tussen geselekteerde gesinsveranderlikes en die geestesgesondheid van swart adolessente te ondersoek. Die tradisionele swart Suid-Afrikaanse gesin, en sosio-politiese faktore wat die gesin beïnvloed het, is ondersoek deur middel van ‘n literatuurstudie. In die empiriese ondersoek is data vanaf 2 505 adolessente in Mangaung, ‘n stadswyk (“township”) buite Bloemfontein in die Vrystaat provinsie, verkry. Die volgende gesinsveranderlikes is ondersoek: gesinstruktuur; huwelikstatus van ouers; grootte van die huishouding; opvoedkundige peil van beide ouers; werkstatus van ouers; tyd wat spandeer word met beide ouers; asook die versorging van ernstig gestremde of chronies siek lede in die huishouding. Veranderlikes wat verband hou met die geestesgesondheid van die adolessente wat in hierdie studie ingesluit is, is depressie, persepsies van sosiale ondersteuning, lewenstevredenheid en eiewaarde. In terme van gesinsveranderlikes is ‘n aantal interessante bevindings verkry. Die uitgebreide gesin was die dominante gesinstruktuur, maar met slegs ‘n geringe meerderheid. Egskeidingsyfers was baie laag en die gemiddelde grootte van huishoudings groter as verwag. ‘n Minderheid adolessente het aangetoon dat hulle voldoende tyd met hulle vaders spandeer. Voorts het ‘n groot proporsie adolessente aangetoon dat hulle vir ernstig gestremde of chronies siek lede in hul huishoudings moet sorg. ‘n Baie hoë depressiesyfer van 26% is verkry. Resultate het feitlik geen statisties beduidende verband getoon tussen gesinsveranderlikes en depressie nie. Moeders wat werk het en die versorging van ernstig gestremde of chronies siek lede in 174 huishoudings was die enigste veranderlikes wat ‘n statisties beduidende invloed gehad het op depressie. Die studie het ook getoon dat adolessente persepsies van verminderde gesinsondersteuning het in uitgebreide gesinne met vroulike hoofde. Getroude ouers, vaders wat werk het en meer tyd wat gespandeer word by beide vaders en moeders blyk die geestesgesondheid van adolessente positief te beïnvloed. Die enigste gesinsveranderlike wat ‘n statisties beduidende verhouding getoon het met eiewaarde, was die versorging van lede met ernstige gestremdhede of chroniese siektes in die huishouding. Voorts het die studie ook gevind dat groter huishoudings, vaders wat werk het en meer tyd wat saam met ouers spandeer word, statisties beduidend verband hou met hoër vlakke van lewenstevredenheid. In die algemeen blyk dit uit hierdie navorsing dat die swart gesin daarin geslaag het om redelik stabiel te bly ten spyte van baie negatiewe faktor e wat in die verlede daarop ingewerk het, en ook faktore wat vandag steeds eise aan die swart gesin stel. Dit bevestig die teenwoordigheid van inherente sterktes in die swart gesin, en sy vermoë om te herstel van terugslae. Verskeie prominente aspekte wat verdere aandag moet geniet beide in terme van verdere navorsing sowel as in terme van intervensie, het in die studie vorendag gekom. Die hoë depressiesyfer verdien ernstige en onmiddellike aandag. Tyd wat adolessente by hulle ouers deurbring moet ook verder ondersoek word. Swart gesinne benodig ondersteuning en hulp om die negatiewe uitwerking van Suid-Afrika se sosio-politiese verlede te oorkom, en om tradisionele waardes met moderne waardes te versoen. In hierdie opsig moet die inherente sterktes va n die swart gesin nagevors en aangewend word. Die ontwikkeling van nuwe gesinstrukture, en die oorsake en gevolge daarvan, verdien ook aandag. Die huidige projek dra aansienlik by tot beskikbare data ten opsigte van swart gesinne en die geestesgesondheid van swart adolessente. Die hoop word uitgespreek dat hierdie studie verdere navorsing sal stimuleer, en dat bevindinge geïnkorporeer sal word in die ontwikkeling en implementering van relevante intervensie. 175 Sleutelwoorde: geestesgesondheid van adolessente; swart adolessente; eiewaarde van adolessente; depressie by adolessente; adolessente persepsie van sosiale ondersteuning; adolessente lewenstevredenheid; swart gesinstrukture; gesinsveranderlikes en geestesgesondheid; gesinsveranderlikes by swart gesinne 176