What Belongs Together Retrieves Together - The Role of Perceptual Grouping in Stimulus-Response Binding and Retrieval

Abstract

Nowadays there is consensus that stimulus and response features are partially represented in the same coding format furthering the binding of these features into event files. If some or all features comprised in an event file repeat later, the whole file can be retrieved thereby modulating ongoing performance (leading to so-called stimulus-response binding effects). Stimulus-response binding effects are usually investigated in sequential priming paradigms where it is assumed that binding occurs in the prime and retrieval in the probe. Importantly, binding and retrieval are not exclusive for targets but also apply to distractor stimuli. A previous study showed that distractor-binding effects were affected by perceptual grouping: Binding effects were significantly larger when stimuli were grouped compared to ungrouped stimuli. Recent theorizing suggests that binding and retrieval are two separate processes that can be individually modulated. Against this background, it is not possible to pinpoint the modulating influence of perceptual grouping on either process at this point in time. Therefore, we adapted the previous study design in two experiments to observe the effect of perceptual grouping on both processes in isolation. Results indicate that perceptual grouping did not impact binding but retrieval: Distractor-response retrieval was reduced when target and distractor were presented in separate objects. Our results thus support recent theorizing on the separation of binding and retrieval.

Keywords: S-R binding; gestalt principles; perceptual grouping.

Figure 1

Exemplary experimental trial. Distractor repetition with response change trial. a) Experiment 1, prime manipulation. b) Experiment 2, probe manipulation. Responses were made towards the letters identity. Perceptual grouping was manipulated by presenting target and distractor within a single box or in two separate boxes. Note that stimuli are not drawn to scale.

Figure 2

Mean Performance for Experiment 1 and 2. Reaction times (left panels) and error rates (right panels) for a) Experiment 1 and b) Experiment 2. The prime-probe relation reflects the interaction between response relation (repetition vs. change) and distractor relation (repetition vs. change). Response repetition/change trials are abbreviated with RR/RC. Distractor repetition/change trials are abbreviated with DR/DC. Thus, this leads to the four trial types response repetition with distractor repetition (RRDR), response repetition with distractor change (RRDC), response change with distractor repetition (RCDR), and response change with distractor change (RCDC). Note that performance should be impaired in partial repetition trials (RRDC and RCDR) compared to complete repetition or complete change trials (RRDR and RCDC). Error bars indicate within-participant error of the mean (Morey, 2008).

Figure A1

Difference Scores of Individual Binding Effects. Raincloud plots (Allen et al., 2019) for differences of individual binding effects (grouped – ungrouped) for a) Reaction Times and b) Error Rates in Experiment 1 (Prime) and Experiment 2 (Probe). The solid horizontal line in each boxplot represents the median of the distribution, the dashed line represents the mean of the distribution. Upper and lower whiskers extend to the largest/smallest value above/below the respective hinge but at most 1.5 times the interquartile range above and below the third and first quartiles (McGill et al., 1978).