The Contributions of Confusion and Position Swapping to Crowding Between Letter-Like Symbols: Evidence and a Confusion-and-Position-Swapping Model

L.-J. Cong,C. Yu,L. Liu

DOI: https://doi.org/10.1167/12.9.338

2012-01-01

Journal of Vision

Abstract:Most current explanations of visual crowding are descriptive. Greenwood et al. (2009, 2010) built a feature position average (FPA) model to explain crowding with letter-like symbols. Here we used crowding errors observed with a similar experimental paradigm to compare this and an alternative model. A T target in one of four cardinal orientations 12.5° above the fixation was paired with a flanker which was one of 25 "cross-like" symbols (including four Ts). The target and flanker had four configurations (flanker on the top, bottom, left, or right side of the target). Observers reported the T target orientation. T target identification rate was 92.3% when presented alone and 68.2% when accompanied by a flanker. When the target and flanker were both Ts, 62.7% of the errors were the flanker being reported as the target. When the flanker was a No-T symbol, a T symbol that was more confusable to the flanker than to the target was often reported, accounting for 50-70% of errors across stimulus configurations. A confusion and position swapping (CPS) model was thus built in which a No-T flanker was allowed first to be misperceived as a T symbol according to an empirical confusion rate and then to swap position with the target at a to-be-determined rate, PSR, the only model parameter. Greenwood’s three-parameter FPA model was also tested. The correlations between observed and model generated errors were 0.442 and 0.604 for FPA and CPS, respectively, suggesting the CPS model being able to better explain the experimental data. The best-fitting PSRs of the CPS model were 0.33, 0.21, 0.28 and 0.27 for the top, bottom, left, and right flanker configurations, respectively indicating the strongest and weakest symbol-level position swapping with outer and inner flankers, respectively, and average swapping when the target and flanker were equal distance from the fovea. Meeting abstract presented at VSS 2012