An analytical model for assessing the knowledge of statistical procedures amongst postgraduate students in a higher educational environment

Philosophiae Doctor - PhDOver the past decades, the use and application of statistical concepts for university students have been a big challenge learned from their previous courses. Aftermath of democracy, South African higher education focused on redressing issues of reparation and social imbalances inherited from Apartheid with the commitment to reconstruct a comprehensive educational quality framework. Growing activities lead to new models emphasised to support students and universities in their attempts to demonstrate evidence of enthusiastic statistics learning, with an acceptable degree of accuracy. This study combines quantitative and qualitative research approaches to assess the knowledge of postgraduate students in applying suitable statistical procedures in higher education (HE). The quantitative data were randomly collected from the postgraduate students (n1=307) while the qualitative data were collected through semi-structured interviews (n2=19) from two institutions (University of Cape Town [UCT] and University of the Western Cape [UWC]) in the Western Cape, South Africa. The SPSS V24 statistical package was used for quantitative data analysis and the explorative design was selected as a theoretical framework to guide the investigation, analysis and interpretation of the qualitative findings. UCT model achieved for all combined categories 73% high prediction accuracy. The UWC model revealed similar results, with ask for help, worth of statistics, fear of statistics monitors, affect, cognitive competence, support from significant others, marital status, ethnic groups and type of study as significant predictors with a high prediction accuracy of 75.49%. Additionally, the ethnic groups, marital status, postgraduate programmes, experiences in statistics and effort were significant contributed factors of SELS beliefs while findings of the combined data of UCT and UWC significantly explained the variation observed in SELS beliefs with only 60% model accuracy. Nevertheless, the qualitative data outcomes indicated that the comments of the participants provided a rich understanding of the perceived failure to choose a relevant statistical test. The results further indicated that confusion and frustration characterised the attitude of students during the selection of a suitable statistical test. The original value of this current study is bridging the inequity gap, in terms of statistics learning, and building a substantial input to the achievement of the objectives of UNESCO, the World Education Forum and the White Paper 3, while ultimately, contributing to the sustainable development of learning statistics at universities in the Western Cape, South Africa. By logical extrapolation, this current study proffers significant insights to the rest of the universities in Africa, and beyond