A discussion of some methodological issues in international assessments

Three topics addressed in the previous chapters are identified and discussed from a somewhat different perspective from that of the chapter authors. The topics are: the level of scoring in assessment studies, translation of test items, and sampling of curriculum content. Based on the analysis of these topics, five recommendations are offered. International assessments should be scored and reported at a more specific level that is currently the practice. There is a need of sound statistical checks on the quality of item translations. Rather than sampling the curriculum only once or twice, sampling could be in real time and on a permanent basis. Tests could be administered with open books and a well-chosen time limit per item. Finally, schools could be instructed to prepare their students for the assessment.