Doctoral Degrees (Computer Science and Informatics)

Permanent URI for this collection

http://hdl.handle.net/11660/410

Browse

Browsing Doctoral Degrees (Computer Science and Informatics) by Author "De Wet, Lizette"

Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    Determining the usability of pedagogical interface agents in the context of adult computer literacy training: a South African perspective
    (University of the Free State, 2016) Mabanza, Ntima; De Wet, Lizette; Schall, Robert
    𝑬𝒏𝒈𝒍𝒊𝒔𝒉 The problem of high computer illiteracy amongst adults in 3rd world countries, such as South Africa (SA), is preventing these individuals from becoming part of workforce as computer skills have become one of the major requirements to enter the job market. The other challenge for this group is that most of the current adult computer literacy training is conducted using approaches such as conventional training and e-learning. Often these approaches pose intellectual challenges to people who have never been exposed to a computer. In line with this, this research study was conducted in an effort to find new approaches to assist adult computer illiterate people, specifically in SA, to acquire the skills necessary (i.e. basic computer literacy) to promote employability and the improvement of their lives. The use of Pedagogical Interface Agents (PIAs) is one example of such a new approach. Over the years, many research studies on the potential of PIAs in promoting learning have been carried out. The results obtained from these studies showed that PIAs can improve student learning, engagement and motivation. However, most of these studies have focused on age groups ranging from childhood to undergraduate college students. Little attention has been paid to the use of PIAs in adult learning, and in computer literacy training, specifically in the word processing environment. The aim of this research study was to establish, through user testing, whether the incorporation of a variety of PIAs could improve and facilitate the computer literacy training (specifically in a word processing environment) for adult computer illiterates in SA. To achieve this aim, Simulated Microsoft Office Word System (SMOS), a simulated word processor system incorporating 10 PIAs (varying in terms of appearance, gender, voice and reality) was used in conjunction with MS Word to collect data needed for this research. A summative usability evaluation was performed in this study. The pragmatist paradigm was selected as a suitable paradigm for the current study. This research study employed the convergent parallel mixed-methods design. Study participants consisted of a group of adults from Mangaung-University of the Free State Community Partnership Programme (MUCPP) who could understand and speak English, had little or no formal education and had never been exposed to a computer before. In order to better assess the usability of the 10 PIAs incorporated in SMOS, participants were divided into two groups, namely the test group (trained with SMOS, which incorporated PIAs) and the control group (trained with MS Word, without PIAs). Afterwards, participants from both groups were given 11 basic word processing tasks to complete using their respective systems (i.e. SMOS or MS Word) as part of the user assessment. In this study, reliability and validity were established through the use of multiple sources of data, which were directly linked to the study aims and objectives. A pilot study was conducted to check the appropriateness of data collection techniques used in this study. The user test performed in this study sought to measure three usability attributes, namely efficiency (measured by means of task efforts), effectiveness (measured by means of both individual and total number of errors) and satisfaction (measured by means of the user experience questionnaire related to the amount learnt from the study overall, enjoyment in being part of the study, and willingness to participate in a similar study in future). Observations and interviews were used to complement and better understand the effectiveness, efficiency and satisfaction data. All collected user test data were analysed using comparative statistical test techniques. The comparative analyses‟ results showed that test group participants (trained with SMOS, which incorporated PIAs) outperformed the control group participants (trained with MS Word, without PIAs) in both efficiency and effectiveness. With regard to the satisfaction, no difference was noted between participants from the two groups. Based on the findings of this study, it can be inferred that the use of PIAs could assist in overcoming many problems faced by adult computer illiterate users during computer literacy training in a 3rd world environment like in SA. However, it is suggested that key aspects, such as characteristics of adult learners, PIAs‟ external observable characteristics, and adult learners‟ satisfaction/ability to use the system, need to be systematically explored. ___________________________________________________________________
  • Thumbnail Image
    ItemOpen Access
    Evaluating and improving the usability of e-government websites in Sub-Saharan Africa for enhancing citizen adoption and usage
    (University of the Free State, 2017) Verkijika, Silas Formunyuy; De Wet, Lizette
    𝑬𝒏𝒈𝒍𝒊𝒔𝒉 E-government has been widely known to be an important factor for fostering a nation’s economic and social development. It holds enormous potential for improving the administrative efficiency of public institutions, encouraging democratic governance, deracinating public sector corruption, and building trust between citizens/private sector and governments. However, most e-government initiatives to date have failed to attain their full potential, because they are increasingly plagued by usability issues. The case is worse for e-government initiatives in Sub-Saharan Africa (SSA) as many researchers have posited that this region is the least developed with regards to e-government diffusion. Moreover, its e-government systems are mostly characterised by poor usability. Consequently, advancing e-government in SSA necessitates advancing the usability of current e-government systems in the region, as usability has been shown to be a vital precondition for e-government progress. As such, this study had as primary objective to develop a model for improving the usability of e-government websites in SSA. The study focused on e-government websites as these sites are generally acknowledged to be the primary platform for government interaction with citizens and other stakeholders. In order to achieve this goal, the design science research (DSR) approach was used as the overall research method for this thesis. The DSR was selected due to its core focus on developing useful artefacts that can effectively address DSR problems. This research approach was further supported by mixed methods encompassing usability evaluation (heuristic evaluation and automated testing) and cross-sectional analysis of national indicators. Using a six-dimensional framework as the core theoretical framework for assessing the usability of e-government websites in SSA, the study concluded that SSA e-government websites were currently characterised by poor usability. After evaluating 279 e-government websites from 31 SSA countries, it was observed that the average usability score for the websites was 36.2%, with the most usable website having a score of 64.8%, while the least usable website scored 10.8%. The poor level of usability was consistent for all the six dimensions. Out of a weighted score of 16.7 for each dimension, the following mean scores were obtained: 7.6 for online services, 6.3 for user-help and feedback, 6.2 for navigation, 5.7 for legitimacy, 5.4 for information architecture, and 3.3 for accessibility accommodation. The study further identified 8 national indicators (corruption, cybersecurity, gender inequality, global competitiveness, human development, innovation, national income, and population age distribution) with significant associations to both e-government development and the usability of e-government websites in SSA. Based on these findings, a policy-ingrained model was proposed for advancing the usability of e-government websites in SSA. The model comprised of four mental models (government, designer, evaluator, and user), with each having both general and specific strategies for improving e-government website usability in SSA. Additionally, the model presented the practical and policy implications regarding the role that the identified national indicators could play in advancing e-government website usability in the region. Moreover, the model was accompanied by a quick assessment checklist that could be used by IT staff of government agencies to evaluate their websites in order to determine which strategies from the model could be applied. The outcomes of this study could contribute to the development of practical and policy-based solutions for improving the usability of e-government websites in SSA, as well as advance the theoretical knowledge base on the use of information systems in government. ___________________________________________________________________