Divergent thinking tasks are commonly used as indicators of creative potential, but traditional scoring methods of ideational originality face persistent problems such as low reliability and lack of convergent and discriminant validity. Silvia et al. (2008) have proposed a subjective top-2 scoring method, where participants are asked to select their two most creative ideas, which then are evaluated for creativity. This method was found to avoid problems with discriminant validity, and to outperform other scoring methods in terms of convergent validity. These findings motivate a more general, systematic analysis of the subjective top-scoring method. Therefore, this study examined how reliability and validity of the originality and fluency scores depend on the number of top-ideas and on time-on-task. The findings confirm that subjective top-scoring avoids the confounding of originality with fluency. The originality score showed good internal consistency, and evidence of reliability was found to increase as a function of the number of top-ideas and of time-on-task. Convergent validity evidence, however, was highest for a time-on-task of about 2 to 3 minutes and when using a medium number of about three top-ideas. Reasons for these findings are discussed together possible limitations of this study and future directions. The article also presents some general recommendations for the assessment of divergent thinking with the subjective top-scoring method.
Keywords: Originality; fluency; reliability; time-on-task; validity.