Data and analysis code for: Stimulus luminance, color, and perceived size expansion

This dataset contains the individual size-matching judgments of observers O1–O4 (Experiments I–III; CSV), the Experiment III stimulus images, the Python analysis code that reproduces all reported statistics and Fig. 3, the analysis results, and the figures for the article "Stimulus luminance, color, and perceived size expansion" by A. Bertulis and A. Bielevičius (Lithuanian University of Health Sciences). See README.md for file contents and variable definitions. Study abstract: Apparent size expansion, the overestimation of a figure's length or height relative to an equal empty interval, depends on stimulus shape and size. We asked whether it also depends on luminance and color. Four observers adjusted an empty test interval to match the length of a 192 × 36 arc min reference rectangle. In Experiment I, rectangle luminance varied from 0 to 50 cd/m² on a 5 cd/m² background; in Experiment II, a 6 cd/m² rectangle was shown on backgrounds of 0 to 100 cd/m²; in Experiment III, rectangles of seven colors (six hues and light gray; 37 to 94 cd/m²) with and without a white frame, an empty white frame, and a three-stripe control were presented on a 26 cd/m² background. Observers overestimated the rectangle length in 139 of 140 observer × condition means (group means 6.6 to 28.8 arc min), whereas the control produced only 3.4 arc min. Expansion increased with positive luminance contrast in Experiments I and II, changed little with negative contrast, and did not depend on color in Experiment III. The results suggest that apparent size expansion is driven by the contour that defines the figure rather than by the luminance or color of its surface.