Answers are grounded in this paper and cite their sources.
Cultural experience shapes attention, memory, and some aspects of visual processing, but it remains unclear whether it influences perceptual selection when incompatible images compete for awareness. We preregistered a binocular rivalry study to test whether Eastern and Western observers differ in the relative weighting of low- versus high-spatial-frequency information during initial perceptual competition. Forty-nine Western participants (U.S.-born, ≤ 2 years abroad) and 43 Eastern participants (recent East Asian arrivals, ≤ 5 years in the U.S.) viewed dichoptic pairs of spatially filtered faces (Albert Einstein, Marilyn Monroe). Image size was varied across five levels to shift retinal spatial frequency content, allowing estimation of each observer’s “balance point”—the size at which low-pass and high-pass filtered faces were equally dominant. In a second component of the study, a filtered image of one face was pitted against an unfiltered image of the other face to assess the competitive strength of isolated spatial-frequency bands against broadband-image rivals. Increasing retinal image size enhanced the dominance of high-pass information. However, mean balance points did not differ between cultural groups, and no reliable association emerged with measures of participants’ levels of acculturation or cultural values. The results suggest that the initial outcome of perceptual competition in binocular rivalry is largely independent of cultural background and that the mechanisms governing this competition operate similarly across observers from different cultural backgrounds. Open practices statement The study was preregistered online ( https://osf.io/pgyqv ). Final analyses deviated from the plan in that we used the quickpsy Package in R (instead of fitting a 2nd-degree polynomial using MATLAB’s curve fitting toolbox) to generate a cumulative normal curve which better characterized the relationship between image sizes and task performance. This also allowed us to use t tests rather than ANOVAs. In addition, sample sizes exceeded the original plan to allow for exclusions and for exploratory individual difference analyses. After reviewing the stereopsis test measure—but without comparing data from the two groups—the team decided that getting six of the nine circles correct (instead of the preregistered criterion requiring nine out of nine circles) would be sufficient to demonstrate stereopsis. We did not use response times as an exclusion criterion for individual trials, and our additional exploratory analyses focused on the self-report measures of culture rather than those included in the preregistration. We deemed these exploratory analyses to be more of interest, given the focus of the paper, than those originally planned or the sample characteristics did not support further follow-up analyses (e.g., uneven distributions of urban/rural background and gender). Data and code are available at https://osf.io/st473 and https://doi.org/10.17605/OSF.IO/ST473 .
No takes yet. Share an insight, caveat, or question.
ZHU et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: