Designing Competency-Oriented Learning Games for Grade 3 Natural and Social Sciences: Supporting Observation, Evidence-Based Reasoning, and Problem Solving

Abstract

Learning games are common in primary classrooms, yet they are often used mainly for motivation or review rather than as structured environments for inquiry. This study developed and preliminarily evaluated a competency-oriented system of learning games for Grade 3 Natural and Social Sciences in Vietnam. The study comprised three phases: a descriptive needs analysis with 32 teachers and 118 Grade 3 students; development of a five-step design process and a 12-game system; and a seven-lesson quasi-experimental pilot with 77 students in two intact classes (experimental n = 41; control n = 36). For international interpretation, the original inquiry indicators were analytically mapped onto three interconnected functions: purposeful observation; evidence-based reasoning, including simple prediction or judgment when tasks permitted; and contextual problem solving. The mapping was conceptual and did not create new psychometric subscales. Before the intervention, no statistically significant difference in total test scores was detected between groups, Welch's t(74.57) = 0.30, p = .763, d = 0.07. At post-test, the experimental class scored higher (M = 8.22, SD = 1.13) than the control class (M = 7.06, SD = 1.35), Welch's t(68.55) = 4.07, p < .001, 95% CI [0.59, 1.73], d = 0.94. Process records indicated that the intervention created more structured opportunities for students to generate questions, record observations, justify classifications, and defend decisions. Because the available dataset contains total scores rather than validated skill-specific subscales, the findings support the feasibility and integrated inquiry value of the system, not separate causal effects on observation, reasoning, or problem solving.