Abstract
One limitation of our visual system is spatial crowding - target identification is impaired when other items appear in spatial proximity to the target. Target identification is also impaired when other items appear before and after the target at the same spatial location – crowding in the temporal domain. This impairment is observed with a relatively long Stimulus-Onset-Asynchrony (SOA , 200 – 400 ms.) which exceeds the time-course of classical masking (<150 ms.). Here, we examined which aspects of visual processing are impaired by temporal crowding, the degree of interference generated by each distractor, and critically, whether temporal crowding is unique to peripheral vision . The participants were shown a sequence of three orientated lines, separated by either short or long SOAs (200 or 400 ms.). All stimuli were presented to the center of the screen. The participants adjusted a probe to match the orientation of the second line in the sequence (the target). In addition, we measured baseline performance by showing only the target instead of a sequence of three stimuli. The distributions of report errors were analyzed with a Mixture-Model Analysis. Experiment 1 used high-contrast stimuli embedded in noise, and Experiment 2 used low contrast stimuli without noise; both were conducted online. The results of both experiments replicated those found with peripheral presentation: Temporal crowding decreased the precision of target report and increased substitution errors. With low-contrast stimuli, we also found an increased guessing rate. Unlike peripheral presentation, the last distractor was more detrimental than the first distractor. These results were replicated with the controlled setting of the lab (Experiment 3). Thus, temporal crowding occurs with foveal presentation and is robust enough to be studied with online experiments. Currently, we examine how the target’s position within the sequence affects performance.