The Stroop interference effect-the slower response to color in an incongruent Stroop stimulus (e.g., ) relative to a neutral Stroop stimulus (e.g., ) is usually highly robust. The present study investigated the role of selective attention in the Stroop task by priming the distractor word. Replicating previous studies using the verbal (color-naming) task, priming the distractor word produced a substantial speedup of response to the color in a Stroop stimulus in our manual Stroop task. Importantly, priming the distractor completely eliminated the Stroop interference effect (Incongruent = Neutral, e.g., ), and brought about a sizable facilitation effect (Congruent <
Neutral, e.g., ) that was absent in the standard (control-primed) Stroop trials. RT distribution analysis showed that the pattern of facilitation and interference effects was changed radically by priming the distractor: In the standard Stroop task, the Stroop interference effect increased across quantiles, and the facilitation effect was absent throughout the quantiles
in contrast, in the distractor-primed Stroop task, the interference effect was eliminated, and the large facilitation effect that emerged remained constant across the quantiles. We interpret these results in terms of a "Trojan horse" account that suggests that in a Stroop stimulus, color and word form are integrated into an object
hence, when object-based attention is deployed to attend to the color, the word form "sneaks in." Priming the distractor breaks this integration, allowing attention to disengage from the irrelevant word dimension and eliminating Stroop interference.