The Impact of Attribute Size on the Performance of Item Selection Algorithms

The use of cognitive diagnosis models (CDMs) and computerized adaptive testing (CAT) has been increasing in both education and other fields such as psychology. To date, most of the research in CAT has been done using item response theory models which provide summative scores. However, formative assessment techniques (e.g., CDMs) that provide more detailed information about individuals’ strengths and weaknesses have become popular in the recent years. The use of cognitive diagnosis computerized adaptive testing (CD-CAT) can produce more diagnostic information with an efficient testing design. This paper aims to investigate the impact of attribute size on the performance of item selection indices in terms of average test lengths. The result of this study showed that increasing the attribute size resulted in longer average test lengths; however, using different CDMs did not change the average test lengths