CHAPTER THREE LITERATURE REVIEW
STUDENT ATTRITION FROM TERTIARY INSTITUTIONS
THE EXTENT OF STUDENT ATTRITION
As a start to this study which would attempt to reduce attrition rates by the early identification of students at risk of failure, it was first necessary to establish whether attrition was regarded as a problem serious enough to warrant time and other resources spent on it. A review of the available literature revealed that the problem of student attrition was widespread, and that the rate of student attrition in most institutions of higher education had remained high for more than 70 years and was still high at the time of this study. With the exception of a few national surveys, most of the published studies reviewed took the form of self-evaluative research by individual tertiary institutions.
STUDENT ATTRITION ABROAD
In the United States, Summerskill (1965) reviewed 35 different studies of student attrition published between 1913 and 1962. He found that colleges lost, on average, about half of their students in the four years after matriculation, and that only 40% of the college students graduated on schedule. In this report it was concluded that the attrition rate showed no appreciable change between 1920 and 1962.
A number of excellent, comprehensive literature reviews since that date such as those by Spady (1970), Tinto (1975) and Pantages and Creedon (1978) all reported the fact that student attrition was a serious problem. Cambell and Dickson ( 1966) conducted a 10-year review using integrative review and meta-analysis of studies investigating student success in baccalaureate nursing programmes. They reported that 44% of all students who were admitted to baccalaureate nursing programmes failed to complete their programme successfully.
number of more recent self-evaluative research studies by individual tertiary institutions in the United States reported serious attrition. From all reports it also appeared as though the problem of attrition was similar across all types of tertiary educational institutions. Anderson et al. (1985) for example found that 46% of all students who enrolled in business programmes at the Athens postsecondary area vocational technical school in Georgia either quit or were dismissed in the 12-month period of studies. Bassin and Sellner (1992) reported that only 58.9% of Bachelor of Science (Business Administration) students who entered the Shippensburg University, Pennsylvania programme in the fall of 1986 remained in the programme four years later.
From Canada, Anderson et al. (1994) reported that approximately 60% of the 4 500 students enrolling annually for the introductory economics course over all three campuses of Toronto University, were prevented from further study in economics because they either failed, dropped out, or did not obtain at least a C grade. Because of the perceived seriousness of the problem of attrition in Canada, a Commission of Inquiry on Canadian Education subsequently investigated the problem. The report by this commission stated that approximately 42% of all full-time undergraduate students who entered Canadian universities in 1985 failed to obtain a degree within 5 years (Johnson & Buck 1995).
Contrary to North America where it appeared to be routine for educational institutions to conduct studies of their rates of retention of students over time, little consolidated research into non-completion rates was available from the United Kingdom. It appeared as though a number of institutions had conducted ‘in-house’ surveys in recent years, but relatively little emerged into print. In a national study Johnes and Taylor (1989) investigated the differences between the undergraduate non-completion rates of different UK universities. Using the 1979 and 1980 undergraduate entry cohorts they calculated a single measure of non-completion based upon the proportion of any given cohort of undergraduates who failed to complete their degree course at the university where they originally registered, using a six-year cut-off date (in other words double the minimum time needed to complete). The results reported indicated substantial differences between non-completion rates of different institutions (between 3.5% for Cambridge and 22.8% for Heriot Watt) and an average of approximately 13.5% for all universities per intake. The differences between different universities as far as non-completion rates were concerned were of interest. It was assumed that the low rate at Cambridge University could in part be attributed to the fact that greater competition for places would exist there than at perhaps less esteemed institutions. A high correlation existed between the non-completion rates obtained for the 1979 cohort and those obtained for the 1980 cohort.
Baumgart and Johnstone (1977) reported that the rates at which students discontinued their studies from higher education in Australia were ‘unfortunately very high’ and remained ‘remarkably stable’ in tertiary institutions. At Macquarie University in Sydney, for example, the rates remained fairly constant for each intake of new students since the university began classes in 1967, at values approaching 40% for undergraduate discontinuation. The authors mentioned that this rate of discontinuation was similar to rates at other Australian tertiary institutions.
From Nigeria, Young (1989) reported growing concern among the authorities of the University of Benin, especially the Science Faculty Board, about the continuing decline in the performance of students in their first year examinations. Failure rates at this university, for example, increased from 22. 7% for the 1977/1978 intake to 52.2% for the 1980/1981 intake, while attrition rates increased from 6.2% in 1977/1978 to 12.6% in 1980/1981. Unfortunately the authors gave no definitions for the rates mentioned. It was assumed that ‘failure rate’ referred to actual subject failures, whilst ‘attrition rate’ referred to loss from the institution.
STUDENT ATTRITION IN MEDICAL SCHOOLS
Published results on attrition rates in medical schools were not easy to compare. This was primarily due to the fact that in the United States of America all medical students are ‘older’ in that they first complete pre-professional studies before selection into medical school. One would therefore expect that the academic failure rates of such students would be low as ‘natural selection’ has already taken place to a large extent during their first few years at college. Most other countries have a combined baccalaureate/medicine course where young school leavers are selected straight into medical school. One would expect that such first-time freshmen, despite the academic attributes that make them acceptable in a very competitive field, would tend to face more adaptation and personal problems than the academically more mature American students. One would thus expect higher attrition rates amongst such first-time freshmen.
However, a review of the literature on attrition rates in medical schools in the United States indicated that failures still occurred despite the very stringent screening methods employed by these schools. Croen et al. (1991) stated that despite careful reviews of applicants’ academic records and letters of recommendation, medical schools each year accepted a number of students who encountered substantial difficulty in coping with the curriculum. They conducted a review of the attrition rates in 30 USA medical schools and found that 6.1°/o of the students had their graduation delayed or withheld because of academic failure.
Cariaga-Lo et al. (1997) conducted a study on all 658 students from the entering classes of 1987-1991 at the Bowman Gray School of Medicine of Wake Forest University. They categorized 5.5% of these students as failures (i.e. those experiencing attrition or academic difficulty) in the first year, with another 10.3% in the second year and an additional I% in the third year. Considering the fact that these three years equate more or less to the third, fourth and fifth years of study in the combined baccalaureate/medicine courses in other countries, it can be argued that these reported attrition rates must be seen as high.
A large national study was undertaken by Koenig et al. (1998) on 11 279 students who entered medical school in 1992 and who took the United States Medical Licensing Examination (USMLE) Step 1 examination in June 1994. They reported a pass rate of between 88. 9% and 96. 8% for the different race groups, except for the African American students who showed a pass rate of only 69.3%. Again the attrition rate was high, considering the fact that all students must have completed pre-professional tertiary studies.
From those countries where medical students are selected straight from school, as is the case in South Africa, the published studies also indicated a serious problem. In the United Kingdom, Simpson and Budd (1996) attempted to assess the extent of the problem of attrition at the School of Medicine, University of Leeds. They conducted a retrospective analysis of the records of all students who failed to complete in the 10-year period 1983 to 1992. Reasons for leaving the course were assessed for each student and thereafter classified as academic failure, personal problems, or ill health. They also noted transfers to other courses if they were known. They reported that 14% of the students failed to complete the course. A possible criticism of this study (as with almost every other study reviewed) is the fact that the researchers only looked at non-completion rates, in other words those students who left the medical school without graduating. They gave no indication of the numbers of students (if any) who graduated behind their initial group. They also gave no indication of whether a particular year of study proved to be more problematical than other years.
In Israel, Lazin and Neumann ( 1991) investigated data for the first 10-year period ( 1974-1983) of the medical school at the Ben Gurion University, Beer-Sheva. They reported a permanent dropout rate (students who failed to complete their studies due to academic failure) of 12.6%, with another 11.2% dropping back (having delayed graduation). An additional 2.5% graduated on time, but at other medical schools. The figures reported for student attrition at Ben Gurion medical school were in actual fact averages and the actual figures at the time of reporting appeared to be much higher than those stated by the authors. A careful scrutiny of their paper indicated that for some reason the researchers divided the 10-year period under investigation into 3 sections, namely 1974 to 1977 (4 years), 1978 to 1980 (3 years) and 1981 to 1983 (3 years). The percentage permanent dropouts as well as the percentage dropbacks over the first two reported periods ( 1974 to 1977 and 1978 to 1980) appeared to be within the averages quoted by the authors. However, for the last reported period (1981to1983), these figures more than doubled to 19% permanent dropouts and 23 .1 % drop backs. This gave a total of 42.1 % of medical students in serious academic difficulty over the period 1981 to 1983. This was a considerable increase over the average figures of 12.6% and 11.2% respectively quoted by Lazin and Neumann (1991).
Nnodim (1994) reported an average failure rate of31.3% in first year anatomy for students at the Faculty of Medicine, University of Benin, Nigeria for the 5-year period 1989-1992. This result was not directly comparable to that of the other studies mentioned above, as it only reported on failures in a single subject. However, it must be remembered that anatomy is a major subject in most medical programmes and failure in anatomy would thus seriously affect total first year pass rates.
Unfortunately no publications which gave attrition rates from South African medical schools could be found. However, the literature reviewed indicated clearly that attrition rates from medical schools around the world were high, and there was no doubt this phenomenon was a source of serious concern.
STUDENT ATTRITION IN SOUTHERN AFRICA
Ayaya (1996) reported serious concern about the high failure rate in the B Comm programme at the National University of Lesotho. Average failure rates of more than 50% occurred in second, third, as well as fourth years of study in this programme. A criticism of this study is again the fact that no clear indication was given by the researcher of exactly what was meant by ‘failure rate’. It appeared from the publication as though failure rate was given per subject, and there was therefore no indication what the actual attrition rates were. According to the author the rates referred only to the first sitting of the end-of-year examination, which means that some of these students could still have passed the supplementary examination and thus the actual failure rate per subject may have been lower than stated. It would have been more appropriate perhaps to have indicated failure rates as the combination of the two examinations. It cannot be denied, though, that such high failure rates, particularly so in the more senior years, highlighted a very serious matter.
In an investigation on failure rates amongst first-year students at the University of the Transkei, Sawyer (1994) reported failure rates that exceeded 50% in six of the nine programmes examined. Sawyer defined a failure as a student who passed in no more than one subject, or no subjects at all in their first year of study at the University of the Transkei.
The definition of failure rate used in the Sawyer (1994) paper, i.e. to define success or failure per subject registered for, appeared to be common to a number of institutions in Southern Africa. In reality student failure rates could be much higher than indicated by such a definition. For example, a student who passed two subjects and was therefore classified as having passed by this method might actually have failed, or he/she may even have been an academic exclusion due to specific departmental or institutional rules. In some cases a student might for example only be allowed to ‘carry’ a number of first-year subjects to the next level, and in other cases the failure of a specific number of first year subjects might lead to academic exclusion.
CHAPTER 1 INTRODUCTION
1.1 Identification of the problem
1.2 Statement of the problem
1.3 Specific research objectives
1.4 Clarification of terms
1.5 Overview of the thesis layout
CHAPTER 2 THEORETICAL BACKGROUND
2.1 Determinants of student attrition
CHAPTER 3 LITERATURE REVIEW
3.1 Student attrition from tertiary institutions
3.2 Management of attrition
3.3 Summary and Conclusion
CHAPTER 4 RESEARCH METHOD
4.1 Study design
4.2 Selection of subjects
4.4 Statistical techniques
CHAPTER 5 RESULTS OF THE STUDY
5.1 STUDY 1 : THE CURRENT REALITY
5.2 STUDY 2 : TEST 1 RESULTS
5.3 STUDY 3: TEST 2 RESULTS
5.4 STUDY 4: DISCRIMINANT ANALYSIS
5.5 SUMMARY AND CONCLUSION
CHAPTER 6 DISCUSSION
6.1 STUDY 1: THE CURRENT REALITY
6.2 STUDY 2: TEST 1 RESULTS
6.3 STUDY 3: TEST 2 RESULTS
6.4 STUDY 4: DISCRIMINANT ANALYSIS
CHAPTER 7 RECOMMENDATIONS
GET THE COMPLETE PROJECT
TOWARDS THE DEVELOPMENT OF AN EARLY WARNING SYSTEM FOR THE IDENTIFICATION OF THE STUDENT AT RISK OF FAILING THE FIRST YEAR OF HIGHER EDUCATION