Electrophysiological evidence for failures of item individuation in crowded visual displays

Abstract

Visual perception is strongly impaired when peripheral targets are surrounded by nearby distractors, a phenomenon known as visual crowding. One common behavioral signature of visual crowding is an increased tendency for observers to mistakenly report the features of nearby distractors instead of the target item. Here, our goal was to distinguish between two possible explanations of such substitution errors. On the one hand, crowding may have its effects after the deployment of attention toward-and individuation of-targets and flankers, such that multiple individuated perceptual representations compete to guide the behavioral response. On the other hand, crowding may prevent the individuation of closely spaced stimuli, thereby reducing the number of apprehended items. We attempted to distinguish these alternatives using the N2pc, an ERP that has been shown to track the deployment of spatial attention and index the number of individuated items within a hemifield. N2pc amplitude increased monotonically with set size in uncrowded displays, but this set size effect was abolished in crowded visual displays. Moreover, these crowding-induced declines in N2pc amplitude predicted individual differences in the rate of substitution errors. Thus, crowding-induced confusions between targets and distractors may be a consequence of failures to individuate target and distractor stimuli during early stages of visual selection.

Figure 1

Task description. (A) Experiment 1 displays. A 500-msec cueing display (white arrow, in this example) was followed by the presentation of the stimulus display for 150 msec. After a 500-msec delay period, observers were instructed to report the orientation of the target T. (B) Experiment 2 displays. Observers were instructed to apprehend the orientation of the clock stimulus on the cued side of the display (right hemifield) and later report its orientation by clicking on the probe ring. All timing parameters were identical to Experiment 1. (C) Configuration of stimulus displays. For a given flanker distance condition (near, far), stimuli were presented in one of three triplet configurations, each demarcated by a unique color. For example, on a single trial in the far flanker condition (left side), three items could be distributed among the three red locations; likewise, in the near flanker condition (right side), three items could be distributed among the three blue locations.

Figure 2

Experiment 1 behavior. Response accuracy for reporting the orientation of the target T declined with set size (p < .001), which was overall higher when flankers were far, relative to near, with respect to the location of the target (p < .001). Because there was almost no effect of set size on response accuracy in the far flanker condition (dashed line), whereas a large decline in response accuracy was observed in the near flanker condition (solid line), the interaction was significant (p < .001).

Figure 3

Experiment 1 electrophysiology. (A, B) Grand-averaged contralateral and ipsilateral waves from OL/OR electrodes recorded during near (A) and far (B) flanker conditions, with a phasic negative deflection indicative of the N2pc at 200–275 msec (gray window). (C) Grand-averaged difference waves from OL/OR electrodes. N2pc amplitudes were measured from 200 to 275 msec (gray window) after stimulus onset (black wedge indicates stimulus duration). (D) N2pc amplitude as a function of set size and flanker condition. N2pc amplitudes rose monotonically with set size in the far flanker condition (dashed line; p < .05), whereas N2pc amplitudes were statistically indistinguishable across set sizes (solid line; p > .4).

Figure 4

Experiment 2 response error distributions: Far flanker condition. (A) Target-related response error distributions across set sizes in the far flanker condition. Error distributions were created by subtracting the participant’s response value from the target value. (B) Flanker-related response error distributions across set sizes in the far flanker condition. Error distributions were created by subtracting the participant’s response value from flanker values. These distributions allow for the visual inspection of feature substitution, which is indicated by a central tendency over distractor values.

Figure 5

Experiment 2 response error distributions: Near flanker condition. (A) Target-related response error distributions across set sizes in the near flanker condition. (B) Flanker-related response error distributions across set sizes in the near flanker condition. Central tendency over flanker values (0, 180°) was apparent in set size 3 displays, indicating the presence of feature substitution.

Figure 6

Parameter estimates. Each distribution (Figures 4A and 5A) was fitted with a three-parameter model using MLE, and the best fitting model parameters were evaluated. (A) P_targ, corresponding to the proportion of trials in which a target-related response was made, declined as a function of set size (p < .001), and more target-related responses were made in the far flanker condition compared with the near flanker condition (p < .001). (B) P_flank, corresponding to the proportion of trials in which a flanker-related response was made (Figures 4B and 5B), increased as a function of set size (p < .001). The largest difference in feature substitution errors was observed at set size 3, in which more flanker-related responses were made during near flanker displays than far flanker displays (p < .001). (C) P_drop, corresponding to the trials in which participants failed to report a feature value present in the display, increased as a function of set size (p < .001), and participants failed to encode target and distractor representations more frequently in near flanker displays than far flanker displays (p < .001). (D) SD, corresponding to the inverse of the resolution of the target representation, increased (resolution decreased) with set size (p < .001), and lower target resolution was observed in near flanker displays than far flanker displays (p < .001).

Figure 7

Experiment 2 electrophysiology. (A, B) Grand-averaged contralateral and ipsilateral waves from OL/OR electrodes recorded during near (A) and far (B) flanker conditions, with a phasic negative deflection indicative of the N2pc at 200–275 msec (gray window). (C) Grand-averaged difference waves from OL/OR electrodes. N2pc amplitudes were measured from 200 to 275 msec (gray window) after stimulus onset (black wedge indicates stimulus duration). (D) N2pc amplitude as a function of set size and flanker condition. N2pc amplitudes rose monotonically with set size in the far flanker condition (dashed line; p < .05), whereas N2pc amplitudes were statistically indistinguishable across set sizes (solid line).

Figure 8

Linking electrophysiology with behavior. We examined flanker-related changes in set size 3 N2pc amplitudes as a function of flanker-related changes in behavior. We observed a significant relationship between crowding-related changes in N2pc amplitudes and observed increases in P_flank in crowded displays (B; p < .01), such that individuals who demonstrated a difference in N2pc amplitudes made more feature substitution errors. We failed to observe a significant relationship between differences in N2pc amplitudes and changes in P_targ (A; p = .89), P_drop (C; p = .25), and SD (D; p = 27).