Development of the evidence evaluation component for the animated scientific creativity test for children and examination of its psychometric properties

In the study, animated scientific creativity test for K2 students was revised by adding an evidence evaluation component and the psychometric properties of the revised test (TOSCAC-R) were examined. TOSCAC-R consists of three subtests: hypothesis development, experiment design, and evidence evaluation, with each subtest containing three items, resulting in a total of nine items. In the study, which included 252 participants, an Exploratory Graph Analysis (EGA) was conducted to assess construct validity, and the expected three-factor structure was obtained. Both Exploratory Structural Equation Modeling (ESEM) and Confirmatory Factor Analysis (CFA) confirmed the three-factor structure of TOSCAC-R, with excellent fit indices for both approaches. The Average Variance Extracted (AVE) values ranged from .518 to .548, Composite Reliability (CR) values ranged from .756 to .784, and √AVE values ranged from .72 to .74. These results demonstrated that TOSCAC-R has a good level of construct validity. It was observed that subtest and total test scores of TOSCAC-R were significantly higher in favor of upper-grade students. Significant correlation coefficients above .50 were found between the SAGES-2 battery and TOSCAC-R, indicating a good level of criterion validity. Item bias analyses were conducted using measurement invariance models to examine gender differences, and it was determined that TOSCAC-R items did not exhibit bias in favor of either boys or girls. As part of the reliability analyses, Cronbach's Alpha and McDonald's Omega coefficients ranged between .765 and .784, while interscorer reliability was rICC = .998, indicating excellent reliability. Keywords: Scientific creativity, TOSCAC, ÇABİYAT, evidence evaluation, exploratory structural equation modelling, exploratory graph analysis