A structured five-step collaborative inquiry approach to mathematical reasoning in elementary school: effects and implementation evidence.

Mathematical reasoning is a core goal of elementary mathematics, yet classroom opportunities to conjecture, justify, and evaluate claims remain uneven. This study examined the effects and implementation of a Structured Five-Step Collaborative Inquiry-Based Instructional Model (CIBI) in Grade 4 mathematics. Using a pretest-posttest nonequivalent control-group design, 96 students (Class A, n = 47; Class B, n = 49) in a public school in Chengdu, China received either 5 weeks of CIBI (20 lessons) or business-as-usual instruction. Reasoning was assessed with a 10-item open-ended Mathematical Reasoning Test using parallel forms (αA = 0.964, αB = 0.967; form equivalence r = 0.993) and analytic scoring (ICC = 0.994). Observation-based fidelity differentiated conditions (M = 1.10 vs. 0.48, 0-3). Mixed-effects modeling showed greater pre-post gains under CIBI (time × condition b = 3.76, p < 0.001), and posttest differences favored CIBI (d = 0.66); the adjusted advantage at the mean pretest was 3.84 points. Findings were robust to covarying midterm achievement and indicated improvements across conjecturing, evidence use, justification, and near-transfer. Overall, CIBI offers a practicable scaffold for strengthening elementary reasoning.