Abstract
This study aims to investigate the effects of item order used with computerized adaptive testing (CAT) in the domain of personality assessment. Based on items from the German version of the Eysenck Personality Profiler, four test versions were designed from seven scales and presented to four experimental groups (n=362). One conventional and three adaptive versions were created, each beginning with items representing either medium or extreme trait levels. Results showed significant differences in average gained person parameters for three out of seven scales between the adaptive versions, as well as differences in average reaction times for answering the items. Results indicated context effects as a problem for using CAT to measure personality.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
