Abstract: Interference in conflict tasks has been shown to be modulated by the frequency of congruent trials, a pattern that is usually explained by strategic attentional control. For example, the conflict effect increases or decreases as the proportion of congruent to incongruent trials increases or decreases, a pattern termed the proportion congruency effect. Even when distracting stimuli are mostly congruent in one context (e.g., location) but mostly incongruent in another context (e.g., another location), the congruency effect is modulated by context, creating a context-specific proportion congruent (CSPC) effect. CSPC effects are often attributed to the attentional setting being bound to contextual cues—allowing rapid, context-driven modulation of attention. However, the theory that this effect is due to attentional modulation has been challenged by contingency learning accounts, which hold that the CSPC effect is entirely explained by cognitive control or contingency learning. In the present research we employed the Hedge and Marsh task to clarify the contribution of contingency learning (strengthened S-R associations) to the CSPC effect. We conducted two experiments to investigate this issue. We adopted a 2 (proportion congruency: 75/25 [MC], 25/75 [MI]) × 2 (S-R color mapping: SC rule, AC rule) × 2 (spatial congruency: congruent, incongruent) within-subjects design. In experiment 1, the spatial location of the stimulus was designed as the context cue, 32 participants were recruited to participant in this experiment. In experiment 2, the shape of the stimulus was designed as the context cue; another 32 participants were recruited to participant in this experiment. The two experiments results showed that there was a significant two-way interaction of proportion congruency and spatial congruency, p < .001. The analyses found that in the MC condition (i.e., when the proportion of congruent trials was higher), the standard Simon effect for the SC rule condition increased, whereas the reversed Simon effect for the AC rule condition was no longer apparent, with the Simon effect being significantly smaller in the AC rule condition, p < .001. In contrast, in the MI condition (i.e., when the proportion of congruent trials was lower), the standard Simon effect for the SC rule condition reversed direction (incongruent condition was faster than congruent), and the reversed Simon effect for the AC rule condition increased. The results indicated that interference effects were reversed mainly due to strengthened S-R associations when the stimulus-irrelevant dimension was set as the contextual cue. These results supported the main role of S-R associations on CSPC effect. In conclusion, when the location and shape of the stimuli was as the context cue, the positive Simon effect in the SC rule was reversed in the mostly incongruent condition compared with the mostly congruent condition; in the AC rule, the reversed Simon effect disappeared in the mostly congruent condition compared with the mostly incongruent condition. These results were difficult to be interpreted by the attentional modulation from cognitive control, and it strongly supported the main role of S-R associations on CSPC effect.

Key words: cognitive control, context specific proportion congruency, stimulus-response association, contingency learning, attentional modulation

摘要： 采用Hedge和Marsh任务与比例一致性操纵相结合，分别以刺激的空间位置和形状作为比例一致操纵的情境线索，考察刺激-反应联结学习与注意调节在比例一致效应中的作用。结果发现，在情境比例一致操纵下，冲突效应的效应量受到比例一致操纵的影响产生反转。表明刺激-反应联结学习在情境特异比例一致效应中起到主要作用。

关键词: 认知控制, 情境比例一致效应, 刺激-反应联结, 可能性学习, 注意调节