Do Creativity-Supportive School Environments Shape Students’ Creativity Mindset and Outcomes? Multi-level Evidence from PISA 2022

• In Event: Situating Growth Mindset across Contexts and Cultures: Experiential, Environmental, and Developmental Influences

Abstract

Objectives
This study examines how students’ perceptions of teacher support for creativity influences their creative self-efficacy and engagement via their creativity mindset. While prior research has established the role of growth mindset in academic motivation and performance (Dweck, 1999; Blackwell et al., 2007), fewer studies have explored how these mechanisms operate in the domain of creativity, particularly using large-scale international data. In this study, we tested a multilevel mediation model, comparing student-level and cross-level pathways linking environmental support, creativity mindset, and creativity outcomes.

Theoretical Framework
Grounded in the mindset × context perspective (Hecht et al., 2021; Yeager & Dweck, 2020), this study conceptualizes creativity mindset—the belief that creativity can be cultivated—as a malleable belief shaped by students’ perceptions of their environment. This framework posits that mindsets are not static traits but interact dynamically with supportive or constraining environments to influence motivation and behavior. Extending this model to creativity, we examine how shared school climate and personal perceptions of creativity support promote mindset-driven pathways to creative self-efficacy and engagement.

Methods
This study analyzed data from the 2022 Programme for International Student Assessment (PISA), comprising 491,460 15-year-old students nested within 19,417 schools across 70 countries/territories. Students were included based on availability of student-reported data on four core constructs: perceptions of a creativity-supportive environment (6 items), creativity mindset (1 item; reverse-coded from fixed mindset), creative self-efficacy (10 items), and creative engagement (8 items). Analyses controlled for socioeconomic status (ESCS), gender, and sampling weight. Doubly-latent multilevel structural equation modeling (ML-SEM) was used to test direct and indirect effects across 1-1-1 and 2-1-1 mediation pathways.

Results
ML-SEM supported both student-level (1-1-1) and cross-level (2-1-1) mediation models. At the student level (1-1-1), perceptions of creativity-supportive environments predicted creativity mindset (β=.19), which in turn predicted self-efficacy (β=.23) and engagement (β=.11). Direct effects from environment to outcomes were significant (β=.28 for self-efficacy; β=.34 for engagement). Significant indirect effects were found for both outcomes: self-efficacy (β=.11, 95% CI [.10, .12], p < .001) and engagement (β=.08, 95% CI [.06, .10], p < .01). At the school level (2-1-1), aggregated school climate predicted students’ creativity mindset (β=.13), which significantly predicted engagement (β=.10) but not self-efficacy (β=.02, ns). Indirect effects on engagement were significant (β=.06, 95% CI [.04, .08], p < .01), while those on self-efficacy were not (β=.003, ns).

Significance
Findings advance mindset research by applying the mindset × context model to the creativity domain and by highlighting how both individual and shared environmental perceptions influence motivational beliefs and outcomes. Our cross-level results echo prior work showing that mindset effects depend on the affordances of the context (Yeager & Dweck, 2020; Walton & Yeager, 2020). This research contributes by focusing on creativity-related constructs across diverse sociocultural settings and emphasizing the need for policies and school climates that promote malleable views of creativity. It also aligns with meta-analytic findings on the variability of growth mindset interventions (Burnette et al., 2023), highlighting the importance of perceived environmental support in intervention effectiveness.

Authors

• Norman B. Mendoza, The Education University of Hong Kong

• Hyun Ji Lee, Hunter College, The City University of New York

• Ronnel B. King, Chinese University of Hong Kong