An Application of Cognitive Diagnostic Assessment on TIMMS-2007 8th Grade Mathematics Items

The least squares distance method (LSDM) was used in a cognitive diagnostic analysis of TIMSS (Trends in International Mathematics and Science Study) items administered to 4,498 8th-grade students from seven geographical regions of Turkey, extending analysis of attributes from content to process and skill attributes. Logit item positions were compared between data for Turkey and the full international data. The Pearson correlation between item logit positions was r = 0.82, though several items were not invariant across the datasets. Results indicated that the majority of the TIMSS items were well explained by a set of 20 attributes (R[superscript 2] = 0.65). This study provides an extension of work by Dimitrov (2007) in use of LSDM in cognitive diagnostic analysis. An appendix contains The Content, Cognitive Process and Skill Attributes for the TIMSS-R (1999). (Contains 3 tables and 5 figures.)