Using Decision Tree Analysis to Understand Foundation Science Student Performance. Insight Gained at One South African University

Abstract

The Foundation Programme of the Centre for Science Access at the University of KwaZulu-Natal, South Africa provides access to tertiary science studies to educationally disadvantaged students who do not meet formal faculty entrance requirements. The low number of students proceeding from the programme into mainstream is of concern, particularly given the national imperative to increase participation and levels of performance in tertiary-level science. An attempt was made to understand foundation student performance in a campus of this university, with the view to identifying challenges and opportunities for remediation in the curriculum and processes of selection into the programme. A classification and regression tree analysis was used to identify which variables best described student performance. The explanatory variables included biographical and school-history data, performance in selection tests, and socio-economic data pertaining to their year in the programme. The results illustrate the prognostic reliability of the model used to select students, raise concerns about the inefficiency of school performance indicators as a measure of students' academic potential in the Foundation Programme, and highlight the importance of accommodation arrangements and financial support for student success in their access year.

Keywords:

• Access to tertiary science studies
• Foundation Programme
• Factors affecting performance
• Tree analysis

Notes

1. The term ‘Black African’ used here is consistent with the ‘population group’ category used by Statistics South Africa (2013). South African government policy documents tend to use the term ‘black' as it has been used in its historical sense to include Africans, Coloureds and Indians. The term ‘Black African' excludes Coloureds and Indians.

2. In South Africa educational disadvantage is typically connected to low socio-economic status (as measured by parents' educational, occupational and economic achievements), second language problems and family breakdown (Zaaiman, 1998). It is also a result of the inferior education offered by the schools they attended which are generally in high-density urban areas, informal settlements or rural villages (Rollnick, 2010). Educational disadvantage in this paper implies socio-economic disadvantage. In South Africa, alternative access programmes are exclusively for students who come from such disadvantaged backgrounds. In other parts of the world, access programmes are typically for ‘non-traditional' students who may not necessarily be disadvantaged according to the South African criteria (see Schuetze & Slowey, 2002).

3. To replace the pre-democracy National Assembly Training & Education Department (NATED) 550 curriculum of the secondary school system, the National Curriculum Statement (NCS) was introduced into the General Education and Training phase in 1998, and into the FET phase including Grades 10, 11 and 12 in 2006, 2007 and 2008, respectively (Reddy, 2006). The end of 2008 saw all South African learners in Grade 12 write common national exams for the first time (Grussendorff et al., 2010). This common national exam, the NSC replaced the SC.

4. This research was part of Kirby (2013)'s doctoral study. This author is particularly indebted to Professor Colleen Downs of the School of Life Sciences, UKZN, for her mentorship, encouragement and award of funding for this research. The first author is also very grateful to Mr Craig Morris for his patience and guidance whilst providing initial guidance with regression and classification tree analysis, a method which this author feels has much potential in elucidating the issues around student performance, and informing curriculum development.