Effectiveness of Argument-Driven Inquiry Combined with Peer Assessment in Developing Epistemic Cognition and Laboratory Report Writing Skills in General Chemistry: A Randomized Cluster Trial

Abstract

We ran a cluster-randomized controlled trial with 228 undergraduate general chemistry students, separating what Argument-Driven Inquiry (ADI) and structured peer assessment each contribute alone, and what they contribute together, to epistemic cognition and laboratory report writing quality. Randomization was at the section level. Twelve laboratory sections went to one of three conditions: ADI only (n = 79), traditional cookbook-style instruction (n = 77), or ADI with structured peer assessment (ADI+PA; n = 72). Writing quality was scored on a six-dimension analytic rubric; for epistemic cognition we adapted a scenario-based instrument from Barzilai and Weinstock (2015). In two-level hierarchical linear models with Kenward–Roger small-sample corrections, ADI+PA students scored significantly higher than both comparison groups on epistemic cognition at posttest (g = 1.48, 95% CI [0.89, 2.07], p < .001), and five weeks after the intervention ended that advantage, attenuated by then, was still present. Writing quality followed the same ordering (ADI+PA > ADI > Traditional), with especially large effects in claim quality and counter-argument consideration. Causal interpretation is constrained, though, by several methodological caveats: instructor–treatment confounding, rubric familiarity effects, and time-on-task differences we did not control. Across sessions, qualitative discourse analysis of the peer assessment dialogues traced a progressive shift away from surface-level comments toward substantive epistemic engagement. Pairing ADI with peer review may therefore be synergistic, but replication with crossed instructor designs and larger cluster counts is needed before these estimates can be confirmed.