Limited Imagination: Effects of Constraints and Individual Differences in Working Memory Capacity on Idea Generation

ABSTRACT

Recent evidence suggests that constraints can facilitate creative thinking rather than hinder it. This research tested how individual differences in working memory capacity (WMC) affect the creativity of ideas generated under conditions of low and high constraint. Participants generated short sentences with either low or high constraints placed on the choice of initial letters used. Under the high-constraint condition, participants were less likely to successfully produce sentences but the sentences were judged as more creative. There was an interaction between constraint condition and working memory capacity such that WMC and creativity were positively associated in the low constraint condition, but not in the high constraint condition. An examination of strategy reports also indicated that higher WMC participants may have made better use of strategies which were related to higher creativity in the low-constraint condition. This research adds to findings that constraints can be beneficial for creativity, and is consistent with suggestions that WMC enhances efficient search and retrieval from long-term memory and use of strategies.

Acknowledgments

We are grateful to Melany Carvalho, Nuo Chen, Lydia Hogan, and Lili Mihaljovits for their help with data coding.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

https://osf.io/sdt6f/?view_only=810fa8668b6841219ea99300a1a76866

Notes

1. Approximately 40% of participants were paid. This was entered as a factor in the mixed model for creativity. However, payment had no effect, β = 0.05, SE = 0.08, t = 0.59, p = .556, suggesting comparable performance across motivation for monetary payment vs. motivation for course credit requirements.

2. Because of variability in creativity scores, participants were also divided into “low” and “high” creative groups based on a median split, and two separate mixed models were conducted for the WMC x constraint interaction for the low and high creative groups. These analyses demonstrate a few findings. First, the main effect of constraints is present in both the low (β = 0.71, SE = 0.06, t = 13.16, p < .001) and high creative groups (β = 0.79, SE = 0.06, t = 10.97, p < .001). Interestingly, for the low-creative group there was no interaction between WMC and constraint condition, (β = −0.003, SE = 0.01, t = −0.51, p = .61) with no predictive effect of WMC on creativity in this group. However, for the high-creative group, the interaction is present (β = −0.04, SE = 0.01, t = −3.82, p < .001) and mirrored the whole-group analysis – WMC was predictive of creativity in the low constraint condition, but not in the high constraint condition.