- Title
- Exploring the early identification of first year accounting at-risk students
- Creator
- De Villiers, Lorelle
- Subject
- Accounting -- Study and teaching (Higher) -- South Africa
- Subject
- Accounting -- South Africa -- Students Accounting
- Date Issued
- 2017
- Date
- 2017
- Type
- Thesis
- Type
- Masters
- Type
- MCom
- Identifier
- http://hdl.handle.net/10948/14895
- Identifier
- vital:27897
- Description
- It is well documented that tertiary institutions in South Africa are reporting high failure rates in accounting courses; several calls have been made to address this in recent years. Various reasons are given for this high failure rate, such as an increase in student numbers and diversity, a mismatch between programmes and students, unequal schooling and inequity of access to tertiary institutions, an increase in the number of underprepared students for tertiary education, and difficulties with language. As a result, there is a growing interest in the factors predicting academic performance, and several studies on predicting academic performance have been undertaken both internationally and in South Africa. Scholars have specifically identified several factors that influence failure in accounting in the first year of tertiary education and in first year accounting in particular. However, the focus of the current study was on several biographical and educational factors only, factors that have commonly been found to influence student performance in their first year, namely: Gender, Age, Ethnicity, Home language, School category, School language, Nationality, Degree programme, Repeating (the accounting module), Admission Point Score (APS), Matric LAMN (combined score for Matric language, accounting, mathematics and numeracy), whether the student studied Accounting in Matric, and Matric year. The primary objective of this study was to develop a predictive model based on biographical and educational secondary data for identifying students at risk in first year accounting at Nelson Mandela Metropolitan University (NMMU). A quantitative research design was adopted and a non-experimental, descriptive study of a cross-sectional and deductive nature was undertaken. The non-probability sampling technique of criterion sampling was used and the sample consisted of all students enrolled for the R101, RNC101, R102, RG102 and RNC102 first year accounting modules in the Department of Accounting Sciences at NMMU. Historical data, both biographical and educational, was collected on which to undertake the data analysis. The data was analysed by means of descriptive statistics and logistic regression analyses. A separate logistic regression analysis was undertaken for each module group because of the diversity of curriculum content and participants of each module group. The findings show that as a whole, the levels of model accuracy in predicting the at-risk and not-at-risk categories were high. The R101 model showed the highest overall prediction success rate with 80.10 per cent of students being accurately classified into either the at-risk or not-at-risk category. The RG102 model showed the lowest overall prediction success rate of 73.91 per cent. The predictor variables of Matric accounting and Home language were identified as the most significant factors in predicting at-risk first year accounting students because they predicted at-risk students in three of the five logistic regression models, while School language, APS, Matric LAMN, Gender and Age were significant in predicting at-risk students in two of the five logistic regression models. The findings of the current study are interesting in highlighting that Matric accounting is not a predictor for students who intend majoring in accounting. They also highlight that for students who intend majoring in accounting in order to become CAs, English language proficiency is an important predictor for at-risk students. For students who intend majoring in accounting and becoming a CA, the findings show that African-home-language students are the most likely to be at risk in the first semester of first year accounting, while Afrikaans-home-language students are the least likely to be at risk in first year accounting. Several observations and recommendations are made and those that are of particular importance to the R101 module relate to Matric accounting, language, APS and Matric LAMN, as well as Gender and Age. For the R102 module, it is those relating to language and Matric LAMN, and for the RG102 module, it is those relating to APS. For the RNC101 module, the observations and recommendations that are of particular importance are those relating to Matric accounting and language as well as Gender and Age, while for the RNC102 module, it is those relating to Matric accounting. The ability to identify at-risk students in first year accounting is of great value to universities across South Africa at institutional, faculty and programme level. The predictive model developed assists in identifying at-risk students timeously, and early identification allows for targeted interventions and support, which could assist these students in overcoming their challenges and ultimately improving pass rates.
- Format
- xiii, 220 leaves
- Format
- Publisher
- Nelson Mandela University
- Publisher
- Faculty of Business and Economic Sciences
- Language
- English
- Rights
- Nelson Mandela University
- Hits: 1339
- Visitors: 1368
- Downloads: 122
Thumbnail | File | Description | Size | Format | |||
---|---|---|---|---|---|---|---|
View Details Download | SOURCE1 | Exploring the early identification of first year accounting at-risk students | 1 MB | Adobe Acrobat PDF | View Details Download |