Design Thinking (DT) originated in professional design practice as an iterative approach used by designers to meet user needs (Liedtka, 2015). In the information age, DT has expanded beyond traditional design domains and has become an essential method for non-designers to address complex, technology-driven environments that require frequent problem solving. Research has documented DT applications in social innovation, healthcare, business, and engineering (Razzouk & Shute, 2012). Interest in integrating DT into early childhood education is also increasing, as DT-related interventions have been shown to positively influence young children’s STEM (Science, Technology, Engineering, and Mathematics) dispositions (Ata-Aktürk & Demircan, 2021), problem-solving skills (Yalçın & Erden, 2021), and 21st-century skills (Yalçın & Öztürk, 2022). These competencies are important for helping children thrive in a rapidly changing, technology-driven world.

Despite this growing recognition, research on DT among preschool children remains constrained by the lack of appropriate measurement tools. As a result, most existing studies have relied primarily on qualitative methods (e.g., Hatzigianni et al., 2021), and a systematic quantitative evaluation framework has not yet been established. In other educational stages, however, scholars have developed scales assessing students’ DT dispositions (Tsai & Wang, 2021), DT skills (Ergin & Coşkun, 2024), and DT mindsets (Vignoli et al., 2023). Developing a DT scale specifically for preschool children would advance quantitative research in this area, facilitate empirical studies, and provide an evidence-based foundation for pedagogical practice. Therefore, this study aimed to fill the research gap by developing and validating a DT scale with a sample of Chinese preschool teachers and children.

Prior research has highlighted the relevance of the IDEO (Innovation Design Engineering Organization) model and the Stanford d. school (Hasso Plattner Institute of Design at Stanford) model for early childhood DT education (Hatzigianni et al., 2021;Yalçın & Öztürk, 2022). The former is derived from the DT toolkit launched by IDEO (2012) for educators, in which DT is defined as a human-centered mindset and follows the process of discovery, interpretation, ideation, experimentation, and evolution (p. 14). The Stanford d. School’s (2018) DT model also consists of five steps: empathize, define, ideate, prototype, and test (p. 1), primarily aims at promoting the academic research and educational dissemination of DT. These two institutions share a close history and collaborative relationship, and their core methods and tools are deeply interconnected (Brown, 2009). Given that the Stanford d. school’s DT model is widely used in the field of educational dissemination, this study adopts it as the theoretical foundation for scale development.

A review of the existing literature on the development and validation of DT scales reveals that most scales are derived from the DT models proposed by IDEO and Stanford d. School (Cai & Yang, 2023; Chen et al., 2023; Ergin & Coşkun, 2024; Tsai & Wang, 2021). In these scales, the stages of the DT model are typically used as the primary dimensions. Additionally, some scholars have also considered the developmental appropriateness of the target measurement population, incorporating other competencies that frequently emerge during DT activities as additional dimensions (e.g., Vignoli et al., 2023). Furthermore, items in the existing DT scales are usually designed in Likert-scale formats (e.g., Chen et al., 2023). However, it is worth noting that there is a research gap in the availability of developmentally appropriate and psychometrically reliable DT assessments explicitly tailored for preschool children. Therefore, the primary aim of the current study is to address this gap by developing a DT measurement scale that is both developmentally appropriate and psychometrically sound for preschool children.

This study aimed to develop and validate the Early Design Thinking Scale (EDTS), a teacher-report instrument for assessing DT competence in preschool children. The EDTS was grounded in the Stanford d. school DT model (2018) and adapted to ensure developmental appropriateness. Following established scale development procedures (DeVellis & Thorpe, 2022), we employed a five-step process: (1) item generation, (2) expert evaluation, (3) target-population evaluation, (4) model refinement and construct validity testing, and (5) internal reliability assessment and descriptive analyses. We first constructed a five-dimension framework aligned with the DT model (empathize, define, ideate, prototype, test) and drafted 20 items accordingly. To better reflect preschoolers’ DT, we incorporated DT-related mindsets, competencies, and behaviors from prior research. We searched Web of Science, Scopus, ERIC, and EBSCOhost using “design thinking” AND “preschoolers,” yielding 1,925 records; 39 English, peer-reviewed empirical studies were retained. Two researchers independently coded DT-related words/phrases that appeared explicitly in the articles and reconciled discrepancies to produce a consolidated code set. An additional coder assessed inter-coder agreement on 20 randomly selected articles (Cohen’s κ = .71). We identified 20 code categories (e.g., empathy, creativity, concrete thinking) and excluded five low-frequency categories (n < 10). Using inductive and deductive coding (Azungah, 2018), we mapped codes to the five dimensions (e.g., making/building and technology proficiency under prototyping), while treating communication and problem-solving as cross-cutting across stages. Integrating the initial items, code interpretations, and relevant scales, we produced EDTS Version 1 (five dimensions, 30 items). Two early childhood DT measurement experts reviewed Version 1; seven items were removed due to weak DT relevance or redundancy. A researcher with extensive preschool experience refined item wording for developmental appropriateness and teacher usability, yielding Version 2 (23 items). Version 2 was administered in two public kindergartens in southern China. Thirty-five teachers rated 484 children (ages 3–6) via an online questionnaire platform. Of 450 questionnaires returned (92.98%), 417 were valid after excluding 33 uniform-response cases. Using R 4.5.0, we split the sample for cross-validation: EFA (sample A; n = 167) and CFA (sample B; n = 250), with no significant group differences in gender (χ² = .04, p = .64) or age (t(415) = 1.09, p = .28). EFA used PAF with oblique rotation; CFA evaluated model fit, while convergent and discriminant validity were examined using AVE and the Fornell–Larcker criterion. Reliability and descriptive analyses were then conducted on the full sample.

Kaiser–Meyer–Olkin (KMO = .95) and Bartlett’s test (χ² (253) = 4023.90, p < .001) supported the suitability of Sample A for factor analysis. Using PAF with direct oblique rotation, the initial 23-item solution produced five factors explaining 81.63% of the variance, with communalities ranging from .65 to .89 (> .60). Two items were removed due to low primary loadings (D5 = .29; T4 = .24), minimal separation from cross-loadings (< .15), and unexpected loading on Empathize dimension. The revised 21-item PAF retained a five-factor structure and explained 83.04% of the variance, and no psychological concerns related to factor loadings were observed. CFA was conducted on Sample B to confirm the factor structure suggested by the EFA. CFA used robust maximum likelihood (MLR) given acceptable distributional characteristics (skewness = −.34; kurtosis = −.68). A first-order five-factor model showed good fit: χ² (179) = 351.21, χ²/df = 1.96, CFI = .97, TLI = .96, RMSEA = .06, SRMR = .03. A second-order model also fit well: χ² (184) = 354.58, χ²/df = 1.93, CFI = .97, TLI = .96, RMSEA = .06, SRMR = .04. All standardized loadings were significant (p < .001) and > .50. Although the differences in model fit were small, we retained the first-order model as the final EDTS because it had stronger theoretical and empirical support, and DT is typically assessed as a set of related yet distinct processes (Brown, 2009; Chen et al., 2023). Convergent validity was supported by AVE values for Empathize (.72), Define (.82), Ideate (.73), Prototype (.73), and Test (.76), and discriminant validity by the Fornell–Larcker criterion. Using the full sample (n = 417), internal reliability was high (ω = .92–.96 across dimensions; overall α = .96). The mean EDTS score was 71.33 (SD = 16.21); skewness (−.34) and kurtosis (−.74) suggested approximate normality, and item means ranged from 2.99 to 4.16. Taken together, these results further support retaining the first-order five-factor model.

Ata-Aktürk, A., & Demircan, H. Ö. (2021). Supporting Preschool Children’s STEM Learning with Parent-Involved Early Engineering Education. Early Childhood Education Journal, 49(4), 607–621. https://doi.org/10.1007/s10643-020-01100-1 Brown, T. (2009). Change by design: How design thinking transforms organizations and inspires Innovation. HarperBusiness. Cai, Y., & Yang, Y. (2023). The development and validation of the scale of design thinking for teaching (SDTT). Thinking Skills and Creativity, 48, Article 101255. https://doi.org/10.1016/j.tsc.2023.101255 Chen, N., Wu, D., Hong, H.-Y., Chai, C. S., & Liang, J.-C. (2023). Developing a design thinking engagement scale for early childhood teachers. International Journal of Technology and Design Education, 33(5), 2045–2069. https://doi.org/10.1007/s10798-023-09807-z DeVellis, R. F., & Thorpe, C. T. (2022). Scale development: theory and applications (5th ed.). SAGE Publications, Inc. Ergin, A., & Coşkun, Y. D. (2024). Design thinking scale development: Assessing reliability and validity. International Journal on Social and Education Sciences (IJonSES), 6(3), 319–333. https://doi.org/10.46328/ijonses.670 Hatzigianni, M., Stevenson, M., Falloon, G., Bower, M., & Forbes, A. (2021). Young children’s design thinking skills in makerspaces. International Journal of Child-Computer Interaction, 27 (2021), Article 100216. https://doi.org/10.1016/j.ijcci.2020.100216. Hasso Plattner Institute of Design at Stanford [Stanford d. School]. (2018). Bootcamp Bootleg. Retrieved online February 25, 2025, from: https://static1.squarespace.com/static/57c6b79629687fde090a0fdd/t/58890239db29d6cc6c3338f7/1485374014340/METHODCARDS-v3-slim.pdf Innovation Design Engineering Organization [IDEO]. (2012). Design thinking for educators toolkit (2nd ed.). Author. Retrieved online February 25, 2025, from: https://page.ideo.com/design-thinking-edu-toolkit Liedtka, J. (2015). Perspective: Linking Design Thinking with Innovation Outcomes through Cognitive Bias Reduction. The Journal of Product Innovation Management, 32(6), 925–938. https://doi.org/10.1111/jpim.12163 Razzouk, R., & Shute, V. (2012). What Is Design Thinking and Why Is It Important? Review of Educational Research, 82(3), 330–348. https://doi.org/10.3102/0034654312457429 Tsai, M. J., & Wang, C. Y. (2021). Assessing young students’ design thinking disposition and its relationship with computer programming self-efficacy. Journal of Educational Computing Research, 59(3), 410–428. https://doi.org/10.1177/0735633120967326 Vignoli, M., Dosi, C., & Balboni, B. (2023). Design thinking mindset: scale development and validation. Studies in Higher Education (Dorchester-on-Thames), 48(6), 926–940. https://doi.org/10.1080/03075079.2023.2172566 Yalçın, V., & Erden, Ş. (2021). The Effect of STEM Activities Prepared According to the Design Thinking Model on Preschool Children’s Creativity and Problem-Solving Skills. Thinking Skills and Creativity, 41, Article 100864. https://doi.org/10.1016/j.tsc.2021.100864 Yalçın, V., & Öztürk, O. (2022). Examination of the effects of design-oriented STEM activities on the 21st century skills of pre-school children aged 3-4. Southeast Asia Early Childhood Journal, 11(2), 1–20. https://doi.org/10.37134/saecj.vol11.2.1.2022