Developmental differences in repeated visual search as modulated by signal to noise ratio

Abstract

This dissertation was designed to study developmental differences in the way simultaneous and sequential signal to noise ratios impact contextual cueing effects. Contextual cueing refers to a form of implicit associative learning of the target location and its context. Over repeated exposures, participants typically respond faster to repeated displays than to new displays that are not repeated. Researchers have assumed that this reflected the learning of associations between the location of the target and the locations of the distracters in the repeated displays. Couperus et al. (2011) found that 10 year olds were able to show intact contextual cueing when the ratio between the predictive distracters and the unpredictive distracters within each display was 75:25, but not when the ratio was 50:50. In contrast, adults showed significant contextual cueing effects when the ratio was 50:50. Hence, it seems that children and adults are differentially sensitive to noise (irrelevant distracters) in the displays. The current study incorporated two forms of signal to noise ratio (S/N): simultaneous S/N, defined as the ratio of predictive and unpredictive distracters within each display; and sequential S/N, defined as the ratio of repeated and new displays within each block. It was predicted that low S/N might be more detrimental to the learning of contextual cueing effects for children than it was for adults. Three age groups participated in the study: 6-8 year old children, 10-12 year old children and college students. In the simultaneous condition, 20 participants in each group were included in the final analysis. The results suggested that all three groups demonstrated significant and comparable contextual cueing effects across three S/N ratio conditions. In addition, the analysis of search efficiency suggested that all three groups demonstrated guided search. This was indicated by faster search slopes to the repeated displays than to the new displays as a function of set size. Therefore, no developmental difference was found in the simultaneous condition. In the sequential condition, 22 participants in each group were included in the final analysis. The results suggested that adults demonstrated significant contextual cueing effects across all three ratio conditions. Older children demonstrated significant contextual cueing effects in the high and medium conditions but only marginally significant learning in the low condition. By contrast, younger children only demonstrated significant learning in the high and medium conditions, but they did not show significant contextual learning in the low S/N condition. There was a significant developmental difference in the sequential condition. Explicit memory tests conducted after the experiment suggested no conscious awareness about the repetition for any age group in any condition. First, the results suggested that adults have an intact ability to extract repeated information from the information stream, as long as it is at least 33% predictive. Contextual cueing is hence a relatively robust form of implicit learning. The developmental difference found in the sequential but not in the simultaneous condition suggested that the presentation mode of irrelevant information impacted the acquisition of contextual cueing in children. It is likely that children's intact learning in the simultaneous condition reflected their relatively mature selective attention mechanisms. They were able to selectively attend to the predictive information when the unpredictive information was presented on the same scene. The developmental difference observed in the sequential condition might be due to children's immature working memory, especially as it applies to younger children. The practical and methodological implications of this dissertation were also discussed.