The effects of process-oriented feedback and revisions on middle schools' fraction arithmetic performance
Date
2025
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Foundational fraction knowledge is critical for students’ long-term success in mathematics and in both STEM and non-STEM careers (Barbieri et al., 2019; Handel, 2016). The middle school years represent a key developmental stage when math becomes more challenging and achievement often begins to decline (Cleary & Chen, 2009; Wang & Pomerantz, 2009). Prior research suggests that two strategies, process-oriented feedback and opportunities to revise, can improve student learning, but few studies have examined their combined effects in middle school mathematics classrooms. This study extends existing research by experimentally testing the effects of varying feedback types and revision opportunities on students’ fraction arithmetic performance. ☐ Using an experimental design, 103 sixth- to eighth-grade students from two schools in the Northeastern U.S. were randomly assigned to one of three conditions: (1) process-oriented feedback with revision (POF+OTR), (2) answer feedback with revision (AF+OTR), or (3) process-oriented feedback without revision (POF). Students completed a pretest, participated in a one-on-one audio-recorded training session where they reviewed their pretest responses, and then completed a posttest. I analyzed the main effects of conditions on posttest scores using multiple linear regression, controlling for pretest performance and class-level fixed effects. I also explored aptitude-by-treatment interactions to assess whether prior knowledge moderated the effects of the conditions. To complement the quantitative analysis, I conducted an inductive thematic analysis of think-aloud transcripts and written responses from 25 students, identifying four key themes: feedback engagement, feedback interpretation, strategies for correction, and emotional/metacognitive responses. Additionally, I conducted an error analysis to track the persistence or change in error types (e.g., strategy and execution errors) from pretest to revised pretest and posttest. ☐ Results revealed no statistically significant differences in posttest performance between the three conditions. Additionally, students’ prior knowledge did not significantly moderate the effects of the three conditions. The qualitative analyses revealed that students who engaged more deeply with feedback tended to revise more effectively. These qualitative findings suggest that combining process-oriented feedback with revision may be more effective for addressing execution errors or procedural misunderstandings, while errors based on students' strategy use may require more intensive or sustained support. Notably, strategy errors remained more resistant to change across conditions.
Description
Keywords
Foundational fraction knowledge, Key developmental stage, Feedback engagement