Retention of Vocational Students in the Danish VET System
The purpose of this quantitative study is to compare the success of Danish vocational colleges in terms of student retention, after allowing for differences in the composition of the student body and external factors, and in the light of the findings to point to colleges that are success-ful and less successful in that respect for the purposes of the research project’s qualitative studies. The research project concerned is entitled “Retention of Vocational Students in the Danish VET System” and examines the susceptible factors that can contribute to reduce the rate of drop-out from vocational education and training (VET).
A corrected comparison of vocational college drop-out rates
It is important to bear in mind that students may drop out for many different reasons. The project “Retention of Vocational Students in the Danish VET System” focuses on vocational colleges and their performance. Since student drop-out is not solely due to conditions at the colleges, it is important, when comparing course completion rates, to correct for external fac-tors that cause students to drop out and over which the individual college has no influence.
The study is therefore based on a statistical model that takes into account the compo-sition of the student body and external factors as well as possible by making use of indicators describing the students’ characteristics, family background and previous education. This al-lows us to assume that the ranking of the colleges is to a high degree a reflection of condi-tions at the individual colleges, including susceptible factors in the teaching and college envi-ronment and not differences in, for example, the colleges’ student recruitment base.
However, it is important to emphasise the fact that although relevant corrections have been carried out, there will be a number of other factors which affect retention rates and over which the colleges have no influence, and which therefore ideally ought to be incorporated into the model. For instance, not all factors that play a part in students’ discontinuation of their course – e.g., inner motivation to gain a qualification and commitment to the course – are included among the register variables. However, we consider that the findings of the present study are less affected by this problem than those of many other studies in the area, as we have taken into account the students’ performance by including their grades obtained at school exit exams in 9th grade (or 10th) before entering VET, which is to some extent corre-lated with their motivation and “will” to gain a qualification, and possible personal problems.
At the same time it must be emphasised that it is not the intention to assess whether a vocational college can be said generally to be good or less good, for example in terms of the quality of the skills that its students carry away with them when they complete their course, but only to identify colleges that have good or less good retention rates when selected objec-tive background factors are taken into account.
Results of the analyses for commercial, technical and social and health care col-leges
The general background to the research project is the Danish policy aim that by 2015, 95 per cent of a youth cohort will have completed a higher secondary education programme (either VET or general education). These terms of reference affect which vocational courses were se-lected for inclusion in the study data, and also the definition of drop-out rate. Thus, we fo-cused on the students’ last vocational course in order to avoid course-switching being counted as dropping out. The reason for this was that a large proportion of students at the commercial and technical colleges begin the basic programme for more than one course, and as our study is concerned with the 95 per cent aim, we thought it important to focus on the rate of drop-out from the last course students embarked upon as definitive of the colleges‘ re-sults. Altogether 46 different college main sites offering commercial qualifications, 38 differ-ent college main sites offering technical qualifications, and 28 colleges offering qualifications in the social and health care field were included in the study.
The actual drop-out rate from the basic programmes at colleges offering commercial vocational courses lay between 10 and 48 per cent, and the mean drop-out rate was 21 per cent. In order to assess the success of the colleges, given the composition of their student body, we looked at the difference between the actual drop-out rate and the expected drop-out rate (as calculated using the model). Amongst the commercial colleges, the college with the best performance had a drop-out rate that was 9 percentage points lower than the expected rate, while the college with the worst performance had a drop-out rate that was 12 percentage points higher than the expected rate.
The actual drop-out rate between commencement of basic programme and commenced main programme at colleges offering technical vocational courses lay between 32 and 70 per cent, and the mean drop-out rate was 44 per cent. Amongst these colleges, the college with the best performance had a drop-out rate that was 12 percentage points lower than the ex-pected rate, while the college with the worst performance had a drop-out rate that was 12 percentage points higher than the expected rate.
Finally, the actual drop-out rate from the main programme for the social and health care helper qualification lay between 17 and 36 per cent, and the mean drop-out rate was 26 per cent. The social and health care college with the best performance had a drop-out rate that was 10 percentage points lower than the expected rate, while the college with the worst performance had a drop-out rate that was 8.5 percentage points higher than the expected rate.
In all three analyses it is the case that the great majority of colleges with significant indi-cators are found in the first and fifth quintiles, where the differences in relation to the national average are greatest, while among the middle-ranked colleges, where the differences between the actual and expected drop-out rates lay closer to zero, there are only a small number of colleges that can be said with statistical certainty to deviate from the national av-erage. For the purpose of identifying colleges, it is a key result that there were considerable and statistically significant drop-out rate differences for colleges in the first and fifth quin-tiles.
The effect of applying the corrections to the colleges’ results was most marked for the techni-cal colleges. The results for the majority of these were substantially different when compo-sition of student body and basic programmes offered at the VET college were taken into ac-count. That signifies that to some extent the differences between colleges in actual drop-out rates can be attributed to differences in the composition of the student body. The correction is therefore important when comparing drop-out rates from technical vocational courses. In the case of the colleges offering commercial and social and health courses, however, the cor-rections had only a minor effect on the results.
As regards the overall design of the project – where the selection of colleges for the quali-tative studies will be based on the findings of the present study – it will be important to focus on the trends at the colleges in recent years. As there is a time gap between this register-based study and the subsequent studies of the colleges, it is relevant to know whether any radical changes that deviate substantially from the trends generally obtained at the voca-tional colleges have taken place in the intervening period with regard to the composition of the student body, the actual drop-out rate or measures to improve student retention.
Danish Institute of Governmental Research | Købmagergade 22 | 1150 København K | E-mail: