When being narrow minded is a good thing: Locally biased people show stronger contextual cueing

Abstract

Repeated contexts allow us to find relevant information more easily. Learning such contexts has been proposed to depend upon either global processing of the repeated contexts, or alternatively processing of the local region surrounding the target information. In this study, we measured the extent to which observers were by default biased to process towards a more global or local level. The findings showed that the ability to use context to help guide their search was strongly related to an observer's local/global processing bias. Locally biased people could use context to help improve their search better than globally biased people. The results suggest that the extent to which context can be used depends crucially on the observer's attentional bias and thus also to factors and influences that can change this bias.

Keywords:

• Contextual cueing
• Visual search
• Attentional bias.

Notes

¹Splitting the data into two groups might suggest that there is a strict qualitative dichotomy between “local” and “global” participants. However, the data suggest that scale of processing is a continuous variable with no clearly defined boundary between global and local. Thus the median split procedure provides two groups of participants who are separated by being more or less globally biased (rather than processing at a strictly local or global level).

²One reviewer asked whether the relationship between attentional bias and contextual cueing depends on overall differences in reaction time. There was only a relatively small difference between the two groups (91 ms), t(38) = 0.72, ns. Furthermore, normalizing the contextual cueing scores (see Howard, Howard, Dennis, Yankovich, & Vaidya, 2004) did not change the overall pattern of result: The contextual cueing effects (averaged across Epochs 3 and 4) were 12.7% and 5.8% in the locally biased and globally biased groups, respectively. Finally, the correlation between attentional bias and normalized contextual cueing was also not different (r = −.57, p < .001). This suggests that group differences were not affected by overall reaction time.