Nestor Matthews¹ (), Megan Broderick²; ¹Denison University, ²The Pennsylvania State University

Introduction: Detecting statistically significant patterns in graphs requires ensemble perception—the ability to extract summaries rapidly from groups of similar elements. Prior work found systematic ensemble perception biases in younger adults (<65 years) when learning to detect statistical significance patterns in graphs (Matthews, Broderick & Kozlowski, 2025). Here, we examined whether those same biases extend to adults aged 65 years or older. Methods: We recruited 480 older adults (aged 65+) from Prolific. Participants viewed 2×2 graphs that plotted a y-axis variable against two predictors: one on the x-axis (Factor A) and one in the legend (Factor B). After random assignment to evaluate Factor A, Factor B, or their interaction, participants learned by trial-and-error to classify line or bar graphs as statistically significant or non-significant. Line and bar graphs displayed identical numeric information. Results: The older adults showed the same ensemble perceptual biases observed in younger adults. First, line graphs yielded significantly greater accuracy than bar graphs when detecting interactions or Factor B main effects. No such line graph advantage emerged for Factor A main effects, which produced low accuracy for both graph types. Second, participants tasked with detecting main effects in bar graphs often applied an incorrect cognitive strategy, misconstruing significant interactions as significant main effects. This incorrect classification rule produced more systematic errors (both false positives, and false negatives) in bar graphs than in line graphs, whether detecting Factor A or Factor B main effects. Conclusion: The findings show that these perceptual biases in graph reading persist across the adult lifespan. They also guide data visualization efforts to optimize graph readability and inform educators about what students likely miss—and incorrectly “see”—when learning to interpret 2×2 graphs.

Acknowledgements: Denison University Research Foundation