National Examinations in the Netherlands: Standard- Setting Procedures and the Effects of Innovations

The Dutch secondary education system is divided into a number of pre-vocational education levels and two pre-university education levels. To complete one of these tracks successfully, pupils have to pass several examinations. These examinations are a mixture of national and school-based assessments in a number of subjects. This paper will address the specific goals and characteristics of the national examinations. Emphasis will be given to intra and inter-year comparability of examination results. Within a year, a national examination is comparable among all pupils, regardless of the school they attend. Interyear comparability is obtained by using linking procedures. However, there are no common items among examinations of different years. For a large proportion of examinations, the linking is thus based on a statistical-result comparison among the population of two consecutive years, under the assumption of random equivalent groups. Under this assumption, the percentile ranks of cut-off scores on the examinations in previous years can be used to set the cut-off scores of the new examinations. Alternatively, for some examinations a non-equivalent group-linking procedure is applied. A number of pilot studies have recently been set up to investigate possible changes to the national examination system aimed at more flexibility for schools and pupils. A key issue is that the intra and inter-year results comparability must remain intact. Changes to the national examination system may involve changes in time, place, content and type of examination. Recent developments mostly concern changes in time, some changes towards computer-based examinations and, to some extent, changes in content and standards. The details of these innovations and their impact on the appropriateness of traditional standard setting methods will also be discussed in this paper.