Measuring creativity: Assessing self-reported creativity and divergent thinking across school aged children

• In Event: Building on divergent thinking: Expanding approaches to measuring children’s creativity

Abstract

Creativity is a vital 21st century skill (Lucas, 2016), yet it is also an elusive construct that has proved difficult to study during childhood. Two main questions that challenge the field are 1) whether children’s self-ratings of creativity reflect their actual creative potential, and 2) whether children’s creativity changes with age. Research addressing both questions has produced mixed findings. Some studies find that children overestimate their artistic creative abilities (Baer, 2010), while another showed associations between children’s creativity ratings for domain specific items and corresponding tasks (Kaufman et al, 2016). Torrance coined the phrase “4th grade slump” in 1968 to describe a decline in creativity that begins as children enter the upper elementary school years. Many studies have examined this premise, with a recent meta-analysis finding that creativity scores do seem to decline as children progress through school, although this likely begins in 7th grade (Said-Metwaly, et al., 2021). We investigate these two questions by 1) examining the associations between self-ratings of creativity and a divergent thinking task and 2) exploring whether self-ratings of creativity varied between elementary and middle school students.

Kindergarten through 7th-grade students (N=94) completed a survey that measured creativity and other intellectual virtues. Creativity was measured with 5 items (Table 1) that were averaged into a composite score (a=.75) A subset set of students (n=29, K, 2nd, 3rd grade) also completed the Instances task (Wallach & Kogan, 1965), a verbal measure of divergent thinking, a common measure of creativity. Children were asked to generate as many, different, and usual ideas as they could in response to three prompts, with responses scored for fluency (number of responses) and originality (unique responses) and summed across the three items.

We found no associations between students’ self-reported creativity scores and divergent thinking fluency (p=.75) and originality scores (p=.56).

Children were divided into three groups, lower elementary (K-2nd grade), upper elementary (3rd-5th grade), and middle school (6th-7th grade), to examine differences in self-reported creativity by age. A One-way ANOVA indicated a significant group difference of creativity by age (F(2,91)=13.19, p<.01, ηp2=.23), and Bonferroni post hoc analysis found that lower elementary students rated their creativity higher than upper elementary (p=.04) and middle school students, (p<.01; Figure 1 for means).

Our findings indicate that younger elementary children’s self-ratings of creativity were unrelated to their divergent thinking performance, despite alignment between the survey items and the direct measure. Self-ratings at this age were significantly higher than older children’s, suggesting that the lack of associations between the self-ratings and direct measure could be related to children overestimating their creativity, but may also be due to limited variability and power to detect an association, which we plan to explore with a larger sample. Unfortunately, we did not have divergent thinking data to assess associations with self-ratings for the older children, so further research should examine whether declines in self-rated creativity with age is related to older children’s more accurate appraisals of their own creativity.

Authors

• Natalie Evans, University of Virginia - Charlottesville
  Presenting Author

• Shoronda Erica Matthews, Boston University

• Jamie J Jirout, University of Virginia - Charlottesville