Fig 1.
Example of the visual world setup.
Object positions were balanced such that each object type was equally likely to appear in each corner of the display. Actual images were derived primarily from the Bank of Standardized Stimuli (BOSS [62]). Images in the visual world are republished from Brodeur et al. [62] under a CC BY license, with permission from Matthieu Brodeur, original copyright (2009, 2010).
Fig 2.
Mean proportion of fixations over time from word onset (in 10-ms bins, averaged across participants) for manipulation trials (top), shape trials (middle), and filler trials (bottom).
Blue lines indicate looks to the target, yellow lines indicate looks to the related item (for manipulation and shape trials), and red lines indicate looks to the average of two unrelated items. Bands around each line represent the standard error at each 10-ms bin. The gray area denotes the window within which relatedness effects were analyzed, and the inset graphs show a zoomed-in view of the relatedness effect during that time window, which—as expected—is evident for both shape and manipulation trials, but not fillers.
Fig 3.
Correlation between total AQ score and the relatedness effect (proportion fixations to related item–proportion fixations to unrelated items) for (A) trials with a manipulation-related object (e.g., faucet–jar) and (B) trials with a shape-related object (e.g., rope–snake). The relatedness effect was calculated as the average of fixation proportions in the critical 600- to 1100-ms time window (see Fig 2). Each dot (in each panel) represents one subject. Pearson’s r correlations are superimposed on each panel.
Table 1.
Model Results Predicting Relatedness Effects from AQ and Competitor Type.
Table A1.
Manipulation and Shape Trial Stimuli.