Abstract
Given a set of simple objects, lab-based studies show that our visual working memory capacity drops from 3–4 down to only 1–2 when the display rotates (Xu & Franconeri, 2015). But real-world STEM experts somehow overcome these limits: chemists rotate novel multi-object molecules. What are lab-based studies missing? The chemistry education literature suggests that these experts don’t store all information in its raw form, but instead compress sets of objects across repeated features. We tested how leveraging repeated features could improve mental rotation performance in a visual working memory task. METHOD: Participants briefly saw a layout of four colored shapes, either all distinct or with repetitions of color, shape, or paired color+shape (e.g., two green squares among a blue triangle and a yellow diamond), with a concurrent verbal suppression task. A rotating ‘windmill’ cued that the layout was to be rotated 90° clockwise or counterclockwise, and participants reported potential swaps (layout change/no change). RESULTS: In experiments 1A–1C, repetition improved performance for color, shape, and paired color+shape. But critically, Experiments 2A–2B found that the benefits of multiple repetitions fell apart when color and shape repetitions were split across different objects (e.g., green square, green triangle, red triangle, along with a yellow diamond). The combined repetition needs to fall on the same objects to substantially improve mental rotation performance. CONCLUSION: Visual compression is an effective encoding strategy (Morey et al., 2015) that may spatially tag identical objects that repeat (see also Brady & Tenenbaum, 2013), creating the visual equivalent of a ZIP file format. But this format cannot leverage repetitions of features separated across multiple objects. Improving student performance in STEM education requires that we understand these limits of visual compression, and individual differences in who can and cannot leverage it.