Development of an ethnoscience-based REA learning model in natural science lessons changes in the form of objects: Analysis of validity and reliability

This study seeks to assess the credibility and trustworthiness of an ethnoscience-based REA (Read, Explore, Apply) learning model framed for teaching the transitions in states of matter in science education. To assess the model’s validity and reliability, the study involved three subject matter experts who provided evaluations focused on content and construct validity. Two questionnaires assessed content and construct validity and reliability; expert responses were averaged for validity and percentages used for reliability. Results showed that out of the 12 content-related items, 11 were in the very valid category range (score interval: 3.25 < VS ≤ 4.00), while one item was in the valid category (2.50 < VS ≤ 3.25). For content validity, all 12 items met the threshold of 0.75 reliability score, thus achieving marked consistency. Regarding construct validity, 14 of 18 items scored as very valid while the remaining four were deemed simply valid. In summary, the investigation found that the ethnoscience based REA learning model is highly valid and reliable for teaching science specifically, for teaching phase changes. This study points to a shift toward learning models that reflect students' daily experiences. This study points to a shift toward learning models that reflect students' daily experiences.