Item

Abstract

The standard psychometric approach to studying creativity uses divergent thinking (DT) tests, where participants are typically asked to generate multiple responses to open-ended questions. Several studies have found correlations between DT and real-life creativity, but the effect sizes have been underwhelming. One reason could be that DT tests are often designed to be domain-general, while real-life creativity depends heavily on domain-specific expertise and related information processing. Thus, developing more domain-specific DT tests could potentially target more relevant cognitive processing and creative potential. For this purpose, a musical divergent thinking test (MDTT) was designed, in which musicians had to create 30 improvised continuations to each of five prime melodies. Each improvisation was scored on complexity and edit distance to the prime melody. The latter measure was explored as an index of melodic distance, similar to how semantic distance is used to indicate associative ability in verbal DT. The MDTT was compared to a traditional DT test (Alternate Uses Test) in a sample of 30 amateur and 32 professional pianists. Professionals significantly outperformed the amateurs on the MDTT, but no group difference was found on the traditional DT test. Interestingly, performance on the MDTT produced the well-known serial order effect with a stronger effect in professionals than amateurs. These results suggest that DT tests tailored to a particular field of expertise tap into creative processes relevant to real-life creative achievement. Accordingly, such tests could be used to study domain-specific creative cognition and potentially predict future creative achievement.