Att betygsätta elevers kunskaper – möjligheter och begränsningar <br/> Grading pupils' knowledge – possibilities and limitations

Abstract

National tests in the natural sciences for grade 6 were carried out 2013 - 2015. One aim was to provide information about students’ scientific knowledge according to the “abilities”: communicating, exploring and explaining, as stated in the syllabus. Are these “abilities” possible and reasonable to distinguish from each other? Answers (60,000) were used for exploratory and confirmatory factor analyses and controlled for dimensionality in a Rasch model. The results show that the tests mainly capture one overall ability. It was not possible to separate the three “abilities” in a reliable way. Thus, our analyses show that dividing the test into subscales for grading the “abilities” jeopardize test validity. A more reliable option might be to let students’ strong and weak performances in relation to different parts of the syllabus compensate for each other. Lastly we discuss our results in relation to the Swedish educational legislation.

Keywords

• national curriculum
• science school subject
• national tests
• grading
• factor analysis
• Rasch modelling