BEDÖMNING AV KUNSKAPER OM SYSTEMATISKA UNDERSÖKNINGAR I NATURVETENSKAP

Abstract

Aim: Teachers working in the Swedish Compulsory School are facing challenges regarding assessment overall and particular in assessing knowledge in Scientific Inuiry (Systematisk undersökning i naturvetenskap). The object of this research project is a quantitative method of assessment called Cognitive Diagnostic Assessment combined with Cognitive Diagnostic Models (CDA/CDM). The aim of the project is to examine if well enough measure regarding reliability of the method, is reachable with sample size originating from an ordinary Swedish classroom. The aim is also to examine what diagnostic information about the students knowledge a teacher might get from a CDA/CDM, in the context of Scientific Inquiry.

Theory: Cognitivism describes learning as constructing structures that in turn impacts our thinking. The structures are unobservable, but we can observe signs of those structures. Knowledge is regarded as a latent variable. We cannot measure knowledge as we can measure shoe size, instead we must measure signs of knowledge. Psychometrics are ways of measure signs of latent variables. Cognitivism interacts with psychometrics in CDA to seek knowledge about a person´s true knowledge.

Method: The aim of this research project is to examine the assessment method CDA/CDM but at the same time CDA/CDM is the method of the project. The CDA/CDM is conducted using the software RStudio with the package ”GDINA”. Both statistics and knowledge in subject matter didactics in science education are used to analyse the results generated in RStudio.

Results: This research project shows that the method CDA/CDM generate well enough measurement regarding model fit and item fit with small samples (as in the classroom) if the actual sample is supplemented with simulated data. The results also indicate that this particular CDA/CDM only can be applied to the group of students, not to the individual student. In regard to this particular CDA/CDM it turned out the students could benefit from education concerning the design of a scientific inquiry (undersökningsdesign) and the concept of researchable question.