Abstract
The fact that the Stroop effect is replicated much more often than the reverse Stroop effect has been attributed to the notion that in the traditional paradigm, which requires a verbal response, responding to the visual color with a color name requires translation of the perceptual stimulus to a verbal code, while reading the word does not (Durgin, 2000). Alternatively, Blais and Besner (2006) argue that the relative robustness of the Stroop compared to the reverse Stroop effect can be predicted by the strength of association between the features and the processing typically associated with the task. Accordingly, if identification of the target’s color or the target word (as in the traditional Stroop paradigm) is more strongly associated with semantic processing than perceptual processing, the target’s meaning should interfere with identification of the target’s color more than the target’s color interferes with reporting the word. In contrast, if localization is more strongly associated with perceptual processing, the target’s color should interfere with localizing the target word more than vice versa. Participants viewed color words in either the congruent or incongruent ink color, and were asked to either identify the meaning of target words or their color (experiment 1), or localize target words among other words based on either word or visual color cues (experiment 2). Consistent with the strength-of-association explanation, the Stroop effect was larger than the reverse Stroop for identification, but smaller than the reverse Stroop for localization. Experiments 3 and 4 replicated the larger reverse compared to Stroop effect in localization, while also controlling for any contribution of faster overall RTs to the smaller Stroop interference in experiment 2. Thus, whereas the Stroop effect may dominate in identification tasks, the reverse Stroop effect appears to be reliably elicited by localization, supporting the strength-of-association account of Stroop interference.