Two kinds of bias in visual comparison illustrate the role of location and holistic/analytic processing differences

Abstract

A number of studies have shown that two stimuli appearing successively at the same spatial location are more likely to be perceived as the same, even though location is irrelevant to the task. This bias to respond "same" when stimuli are at the same location is termed spatial congruency bias. The experiments reported here demonstrate that the spatial congruency bias extends to letter strings: Participants tend to respond "same" when comparing two strings appearing successively at the same location. This bias may arise because successive stimuli at the same location are more likely to be perceived as a single object. Bias is also affected by the nature of the comparison task. We show that if letters must be compared individually (analytical comparison), there is a bias to respond "different," but if letter strings are compared as unified wholes (holistic comparison), there is no bias or a bias to respond "same." This analytical bias is apparently separate from the spatial congruency bias. It appears whether the task requires localization of differences between strings, or counting the number of differences, or ignoring differences in some parts of the stimuli while attending to others. All of these analytical comparison tasks require that letters be selected individually, and the analytical bias may reflect difficulty in preventing interference from neighboring letters in this selection process. Each type of bias reflects a different aspect of visual processing, and both can be measured to probe how processing changes across different tasks.

The influence of spatial location on same-different judgments of facial identity and expression.

“…A related question is whether the physical differences in stimuli across tasks (e.g., availability of color image information, hair/background cropping, and gender variation) might have contributed to a different processing strategy and/or attentional focus across tasks. A recent paper by Cave and Chen (2017) identified an additional type of bias related to analytic versus holistic processing strategies during visual comparison tasks. They found that when stimuli could be compared as unified wholes, participants tended to use a holistic strategy, resulting in an overall greater bias to report the items as the “same,” whereas participants tended to use an analytic comparison strategy when parts of the stimuli must be compared individually, resulting in an overall bias to say “different.” Critically, this analytic/holistic bias appears to be orthogonal to the spatial congruency bias; in Cave and Chen’s study, a spatial congruency bias (difference in bias) was reported for all experiments, regardless of the direction of the overall analytic/holistic bias (Cave & Chen, 2017).…”

“…A recent paper by Cave and Chen (2017) identified an additional type of bias related to analytic versus holistic processing strategies during visual comparison tasks. They found that when stimuli could be compared as unified wholes, participants tended to use a holistic strategy, resulting in an overall greater bias to report the items as the “same,” whereas participants tended to use an analytic comparison strategy when parts of the stimuli must be compared individually, resulting in an overall bias to say “different.” Critically, this analytic/holistic bias appears to be orthogonal to the spatial congruency bias; in Cave and Chen’s study, a spatial congruency bias (difference in bias) was reported for all experiments, regardless of the direction of the overall analytic/holistic bias (Cave & Chen, 2017). In contrast, in the current study we found a complete lack of spatial congruency bias in Experiment 2.…”

“…This makes sense from an ecological perspective, since objects in our environment that occupy the same location in space are generally more likely to be the same object, that is, outside the experimental context, an object remaining in the same location is likely to maintain the same identity. Previous studies have found that a pair of objects appearing in the same visual location is more likely to be judged as the same orientation, letter string, or shape, compared to a pair of objects appearing in different visual locations (Cave & Chen, 2017;Golomb et al, 2014;Shafer-Skelton et al, 2017). In addition to these lower-level object judgments, Shafer-Skelton et al ( 2017) also found that participants were more likely to judge two faces as being of the same identity if they were presented in the same location.…”

“…Previous studies have found that a pair of objects appearing in the same visual location is more likely to be judged as the same orientation, letter string, or shape, compared to a pair of objects appearing in different visual locations (Cave & Chen, 2017;Golomb et al, 2014;Shafer-Skelton et al, 2017). In addition to these lower-level object judgments, Shafer-Skelton et al…”

“…The “spatial congruency bias,” first reported by Golomb et al (2014), reveals the tendency for observers to judge two sequential items presented in the same location as being the same identity, even when the items are not identical. The spatial congruency bias has been observed with simple stimuli (e.g., oriented gabors, letter strings) and stimuli as complex as computer-generated faces (Cave & Chen, 2017; Golomb et al, 2014; Shafer-Skelton, Kupitz, & Golomb, 2017). The spatial congruency bias effect is measured as a difference in response bias when stimuli are presented in the same versus different locations, such that two objects presented in the same location are more likely to be perceived as having the same orientation, color, shape, and even facial identity, compared to if the objects appeared in different spatial locations.…”