A Comparison of Traditional and Computer-Assisted Methods for the Development of a German Need for Cognition Short Scale

The current study presents the development and validation of a Need for Cognition (NEC) short scale. In Study 1 (N = 282), traditional item selection procedures based on classic test theory were used as well as an innovative computational approach, our "full information approach." The procedures led to three different short scales with comparable psychometric quality, which were validated in Study 2 (N = 530). Based on different validation criteria, one of the three scales obtained from the full information approach was selected as the final short scale NFC-K. This NFC-K achieved a Cronbach's a of .69, a correlation of .81 with the original scale, and showed no significant correlation with social desirability. Besides presenting a systematic and objective selection and validation of the NFC-K, this article represents a case study of the challenges of developing a short scale comparing both traditional and computational approaches.