Creative self-belief responses versus manual and automated alternate use task scoring: A cross-sectional study

Abstract

While behavioral tasks like the Alternate Use task (AUT) quantify creative potential levels, clarifying how they relate to subjective self-report ratings would contribute to the creativity assessment field. Valid and reliable single-item measures to efficiently assess one's creativity level could allow researchers and educators with limited time and resources to use the simpler and shorter self-report item. This study's primary objective was to evaluate the construct validity of a single-item creative self-belief (CSB) measure by comparing it with AUT fluency, flexibility, elaboration, and originality scores that were scored using manual and automated methods. It also aimed to assess the single-item CSB's convergent validity and test-retest reliability. In addition, the relationship between the manual and automated AUT scoring methods was evaluated. Data from 1,179 adult participants collected in a more extensive parent study were used for these analyses. CSB was weakly correlated to manual fluency (rho = 0.13, p = 0.004, n-505) and manual originality (rho = 0.11, p = 0.01) but no other AUT measures. CSB was correlated with the personality indices of openness to experience (rho = 0.49, p <0.001), extraversion (rho = 0.20, p <0.001), neuroticism (rho = -0.20, p <0.001), agreeableness (rho = 0.14, p <0.001), and conscientiousness (rho = 0.14, p <0.001). CSB test-retest reliability, assessed using entries from participants who completed two sessions, was high (Intraclass Correlation 78.6). The manual elaboration score was strongly correlated with the automated Open Creativity Scoring with Artificial Intelligence (OCSAI) elaboration score (rho = 0.76, p < 0.001), and manual originality scores were correlated with OCSAI originality scores but less strongly (rho = 0.21, p < 0.001). These findings support using multiple measures to assess aspects of the creative process, not relying solely on this single-item CSB measure. However, the single-item CSB item may be helpful in limited-time situations and has demonstrated positive content validity, test-retest reliability, and a significant, albeit weak, correlation to AUT fluency and originality. This study also supports the continued use of OCSAI to score elaboration and originality in AUT.