# Measuring scientific creative thinking: development and validation of a process-oriented instrument

## TL;DR

This paper introduces a new tool to measure scientific creative thinking in students, validated through rigorous statistical methods.

## Contribution

The paper presents a novel, process-oriented instrument for assessing scientific creative thinking with strong psychometric properties.

## Key findings

- The Scientific Creative Thinking Scale has a stable three-factor structure confirmed by exploratory and confirmatory factor analyses.
- The scale shows robust reliability with high internal consistency and inter-rater reliability.
- The SCTS correlates significantly with science academic achievement, supporting its validity.

## Abstract

Scientific creative thinking (SCT) is a critical cognitive process that drives scientific discovery and innovation. To address the limitations of existing assessment tools in capturing its multi-stage, domain-specific nature, this study developed and validated the Scientific Creative Thinking Scale (SCTS). Grounded in an integrative psychological framework, the SCTS combines explicit indicators of divergent thinking with implicit theories of within-domain appropriateness. The scale employs three open-ended tasks based on real-world scientific scenarios to systematically assess the three core stages of SCT: problem identification, hypothesis construction, and experimental verification. Using a two-stage independent sample design with 401 tenth-grade students, the scale was rigorously validated. Exploratory factor analysis (n = 182) employing Principal Axis Factoring (PAF) with parallel analysis extracted a stable three-factor structure, accounting for 70.189% of the total variance. Confirmatory factor analysis with an independent sample (n = 219) confirmed the three-factor model with good fit indices (CFI = 0.948, TLI = 0.922, RMSEA = 0.076, SRMR = 0.032). The scale demonstrated robust reliability: internal consistency was satisfactory (Cronbach’s α = 0.87; McDonald’s ω = 0.89), and inter-rater reliability was excellent (ICC ≥ 0.95). Composite reliability for each dimension exceeded 0.86, and average variance extracted exceeded 0.67, indicating good convergent validity. The Fornell-Larcker criterion was met, as the square root of AVE for each dimension exceeded its correlations with other factors, supporting discriminant validity. Additionally, the SCTS total score showed a significant positive correlation with science academic achievement (ρ = 0.55, p < 0.001), providing preliminary evidence for criterion-related validity. This study offers a process-oriented, psychometrically sound instrument for measuring SCT, laying a methodological foundation for future research on its cognitive mechanisms, individual differences, and developmental trajectories.