Investigating the neural correlates of visual attention and response selection in contextual cueing

Abstract

Response times during visual search are facilitated when target-distracter spatial configurations are repeated, a phenomenon known as contextual cueing (Chun & Jiang, 1998). According to one account, the influence of context arises from the guidance of attention to the target location in repeated contexts by long-term memory representations (Chun & Jiang, 1998), and this claim is supported by studies showing that the N2pc event-related potential (ERP) component is larger for repeated search contexts than for novel contexts (Johnson et al., 2007). According to a contrasting view, however, the effect of context is thought to be largely mediated by response selection processes (Kunar et al., 2007) rather than attentional guidance, and studies showing enhanced response-related readiness potentials (i.e., LRP) in the absence of an N2pc enhancement for repeated contexts appear to support this claim (Schankin & Schubo, 2010; Schankin et al., 2011). These conflicting results highlight that the roles of visual attention and response selection in contextual cueing remain unknown. Here we investigated the effects of repeated context on ERP components related to attention (N2pc) and response preparation (LRP). Thirteen subjects performed a visual search task in two sessions (256 trials, 50% repeated contexts) separated by one week and with 64-channel EEG recorded. Behaviorally, there was no contextual cueing effect in session 1, but there was in session 2 (p <0.01). The N2pc also did not differ between repeated and novel contexts in session 1, but in session 2 an enhanced N2pc emerged for repeated displays (p <0.05). Conversely, the amplitude for the response-locked LRP component did not differ between repeated and novel displays in either session 1 or session 2. These results support the notion that visual attention mechanisms, as indexed by the N2pc ERP component, play a key role in the contextual cueing effect. Meeting abstract presented at VSS 2012