Writing Produces Better Calibrated Learners Than Rereading

The rationale behind the present study consisted of evidence reported to date for Galbraith et al.’s (1999, 2018, 2023) dual-process model of writing, suggesting that not only do writers engage in knowledge transformation, but in the development of new knowledge. An additional aspect of Galbraith et al.’s (2023) work is their proposal and validation of a novel subjective knowledge measure, tailored to those two processes, with potential to be used as a tool for calibration, knowledge activation, and learning. The purpose of the present study was to (a) investigate knowledge development comparing two different writing tasks relative to a comparison task of rereading a text passage, (b) explore patterns in subjective knowledge, confidence, and situational interest ratings throughout engagement with such tasks, (c) examine the predictive power of those ratings for post-intervention knowledge, and (d) compare confidence ratings with evidence of knowledge, that is, calculating calibration scores. The study used a pretest-posttest repeated measures intervention design, in which 146 undergraduate students, enrolled in human development and psychology research methodology courses, were randomly assigned to experimental or comparison conditions. Students in all conditions started by reading a text on the topic of research design, after which students in the experimental conditions engaged in two writing activities, consisting of a free-write (for both experimental conditions) and either an explanatory or persuasive writing task. Simultaneously, students in the comparison condition reread the initial text twice while being tasked with, first, surface-level strategies and, second, deep-level reading strategies. At least a week after the intervention, students in all conditions completed a transfer test, consisting of an argument writing task. Students rated their subjective knowledge about the topic (using an adapted version of Galbraith et al.’s [2023] instrument), confidence in their knowledge about the topic, and situational interest in the topic at hand multiple times throughout the study. The study occurred in real classrooms, using materials akin to existing course materials, on a topic already part of existing course curricula but not yet covered, which contributed to its high ecological validity. Exploratory factor analyses indicated that the two subscales of subjective knowledge ratings and the single-item confidence rating needed to be combined into one factor (Subjective Knowledge/Confidence; SKC) and treated as such in all analyses. Further, tests of condition regarding knowledge gains, one of the primary hypotheses, needed to be adjusted because of a failure of randomization between groups that was observed upon analyzing initial between-group equivalence. Despite random assignment to conditions, significant differences between conditions on the primary dependent variable of conceptual knowledge were found at pretest for the comparison (control) group. Because such a difference at pretest would invalidate any causal conclusions drawn from comparisons between the experimental and comparison conditions, further comparisons were made only between the two experimental groups in addressing those research questions that pertained to the effect of condition on changes in knowledge and the subjective factors measured, as well as the predictive value of those subjective factors for post-intervention knowledge levels. Findings indicated that the writing intervention central to the present study had a positive, significant effect on learning about the topic of research design for students in both experimental conditions (i.e., explanation and persuasion) relative to their pretest knowledge levels. Additionally, students in the persuasion group were significantly better calibrated than students in the rereading group, and SKC ratings at posttest were a significant predictor of transfer-test knowledge scores for both the explanation and the persuasion groups, indicating an improved relationship between confidence and actual knowledge levels. The findings of this study underscore the importance of providing students with a range of learning strategies, including rereading and writing, to help them acquire knowledge. Educators can use these findings to inform their instructional decisions, recognizing that students’ individual needs will vary and that a combination of strategies may be most effective in promoting knowledge development.