Interpreting data tables: Can variable symmetry scaffold performance? [Author Accepted Manuscript]

Understanding how individuals interpret data is crucial for enhancing reasoning and data-driven decision-making across various domains. Prior research suggests that table structure – specifically, whether variables are symmetric (e.g., treatment A vs. treatment B) or asymmetric (e.g., treatment vs. no treatment) – influences reasoning accuracy and interpretation strategies, with symmetric tables supporting more accurate interpretation and normative interpretation strategies. The current study examined whether prior experience interpreting symmetric tables scaffolds subsequent data interpretation performance when individuals later encounter asymmetric tables. Undergraduate participants were randomly assigned to interpret an initial set of three symmetric or asymmetric tables. All participants then interpreted a second set of three asymmetric tables. Compared to participants who initially interpreted asymmetric tables, those who first interpreted symmetric tables did not show higher accuracy or greater use of proportion-based strategies on the second set. However, participants in the symmetric-first condition rated the cells in the bottom row of the table as more important for their reasoning. This pattern was especially pronounced among participants who sometimes used frequency-based strategies; participants who consistently used proportion-based strategies tended to rate all four cells as equally important. These findings extend prior work by showing that prior exposure to symmetric table structure influences how attention is allocated within tables, even when accuracy outcomes do not differ. The results inform research on table data interpretation and have implications for instructional approaches aimed at supporting proportion-based reasoning in data interpretation.