2006:45 R

Examinations with quality - a survey of examination procedures at some Swedish higher education institutions

A number of reports have been presented, mainly focusing on the form taken by examinations and the way in which they are conducted. The interest directed at the quality of examinations has been relatively scanty.

In this report, the intention has been to study examinations and examination procedures in four degree programmes, medicine, psychology, master´s programmes in engineering and teacher training programmes. Particular emphasis has been placed on their quality. The study was undertaken during the period from 1 April 2005 until 30 March 2006.

Two samples were taken in each programme, one at the beginning and one in a later section. The study involved review of the contents of the examinations (test items and a selection of tasks, requirements and instructions) interviews with teachers and students and analysis of data from the examination results (basic psychometric analysis with a focus on reliability and validity).  

Reliance on written examinations is clearest in the programme in medicine. Teacher training programmes seem to some extent to be their antithesis. In our samples we found no traditional examinations (taken in examination rooms) but only take-home tests or the like. The programme in psychology reflects elements of both of these programmes. The master´s programme in engineering, finally, focuses mainly on traditional written examinations. Here, too, “quizzes" feature prominently.

Questions of reliability and validity do not seem to play any role in the choice of test or examination form. It seems as if no attention is paid at all to the question of the quality of examinations in higher education in Sweden. The examinations are taken for granted and considered to be objective. This seems to apply to all four degree programmes.
The fundamental approach to the question of whether students should pass or fail seems most frequently to allow them to pass. Resits do take place, but these also usually result in passes, not infrequently for grounds other than those stated in the directives for the examinations.
In Sweden there are firm links between attitudes to examination and the way in which examinations are organised and assessed. The current situation is that the state funding received by each higher education institution is related to the credits attained by students and at the same time students must pass their examinations to get their loans and grants. There is an obvious risk that today´s system makes it advantageous to pass students irrespective of the grounds on which this may be based.

We can determine that there is no discussion of the quality of examinations, and of examination procedures, in higher education in Sweden. This lack of attention may be the outcome of an organisational system in which positive results benefit both the individual students (loans and grants) and the higher education institutions (more funding from the state). There is however a risk that this system could lead to lack of attention to examination requirements, so that they could be reduced. The effect in the long run would be a gradual decline in standards.

In our opinion greater weight must be placed in higher education on the issues of reliability and validity per se and in particular when determining examination procedures. One of the conclusions reached in our study is that examinations should be organised on a two-stage basis. The first would require written evidence of command of conceptual systems, understanding of the fundamental principles and theories and some degree of awareness, for instance, of significant individuals in a subject. The second would involve tasks of different kinds to probe higher order cognitive skills of the kind described by Bloom and others.