Shorter can be better: Balancing length and predictive power when measuring noncognitive skills to predict academic outcomes

Shuaizhang Feng,Yu Gan,Yujie Han,Tim Kautz
DOI: https://doi.org/10.1016/j.econlet.2024.111598
IF: 1.469
2024-02-11
Economics Letters
Abstract:We develop shorter versions of a Big Five survey designed to measure students' noncognitive skills and predict students' later academic outcomes. We find that measures with fewer items can better predict students' outcomes, suggesting that using shorter versions of a Big Five Inventory may be cost-effective in large-scale social surveys.
economics
What problem does this paper attempt to address?