Creativity at the task and process level: What is responsible for age differences

Laboratory measures of creativity have often emphasized divergent thinking, however, the age differences obtained on such measures may have been largely due to speed. The themes of this dissertation were that quality of problem solution is an important aspect of creativity, along with quantity, and that convergent (evaluating answers) and divergent (generating answer) thinking may both be involved in creativity. Fifty younger (M = 20) and fifty older (M = 73) adults completed Mednick's Remote Associates Test (RAT) and Practical Creativity Problems (PCP). The creativity measures were decomposed into component tasks of divergent (i.e., generate answers without evaluating the quality of the answers) and convergent (i.e., evaluate the quality of the answers without generating answers) thinking, for both the RAT and PCP.; Age differences favoring younger adults were found for RAT proportion correct, PCP answer quality, and the total number of answers generated for both RAT and PCP. For decomposed process measures, younger adults generated more answers (divergent thinking), and younger adults were more sensitive in distinguishing between adequate and inadequate solutions (convergent thinking). RAT proportion correct was predicted by decomposed process measures of convergent and divergent thinking; the decomposed process measures did not mediate the effects of ability, although they did mediate the effects of age. PCP answer quality was predicted by divergent thinking; age was mediated by divergent thinking. Support for convergent thinking compensating for divergent thinking was found for the PCP, but not for the RAT, where divergent thinking served as a minimum necessary process. Older adults may be limited by divergent thinking in the RAT, but they appear to compensate with convergent thinking on the PCP.