Investigating the Relation Between Peer Critiquing and Model Revisions in Modeling-Based Learning

Modeling-based learning engages students in model generation, evaluation, revision, and application. This study specifically investigated the relation between peer critiquing and model revisions so that suggestions can be made for the design of curricula and interventions to support students in model evaluation and revisions. This study engaged 158 high school students in 47 groups in evaluating and revising their models on the topic of carbon cycling. The students' comments received in peer critiques and subsequent changes to their models were coded and analyzed. Qualitative coding and rating were conducted to indicate students' epistemic criteria for critiquing, and changes in model quality after critiquing. Epistemic network analysis (ENA) was employed to investigate the students' epistemic framing demonstrated during peer critiquing, and to relate it to the students' model revision behaviors. A total of 14 different types of epistemic criteria used by students during peer critiquing were identified. Thirty-eight of the 47 groups (81%) revised their models, but only 10 (21%) improved their model quality in terms of the model's scientific merit. The relation between model critiquing and revision or improvement was identified, including: (1) peer critiques focusing on whether the model was complete, original, accurate, provided sufficient information, could support explanation, or needed to be more aesthetic, were more likely to lead to model revisions, and (2) peer critiques employing the completeness- or quantity-centered epistemic framing were better able to elicit significant model improvement compared to those employing the organization-centered epistemic framing. It is concluded that with scaffolding, high school students can conduct peer critiquing with various epistemic criteria and distinct epistemic framing. Instruction needs to pay attention to student language in light of its epistemic framing in peer critiquing, and can start by focusing on criteria and framing that appeal to students, such as the completeness- or quantity-centered epistemic framing, which can lead to productive model improvement. Future research needs to investigate the interplay among peer critiquing and teacher guidance and feedback to augment the effects of peer critiquing. The study also contributes to the literature by combining qualitative coding and quantitative analysis to identify students' epistemic framing significant to model evaluation and revision.