The primed Stroop is not a Stroop task: Evidence from delta plots

The Stroop interference effect - slower color naming on incongruent trials (e.g., GREEN printed in red) compared to neutral trials (e.g., PRESS printed in red) - is widely regarded as a hallmark of selective attention failure. We (Curtis et al., Psychonomic Bulletin & Review, 32, 1328-1336, 2025) have previously argued that the integrated nature of the Stroop stimulus is a main factor contributing towards the robustness of this interference, which shows a characteristic positively sloped delta plot pattern. Here, we tested this claim by directly comparing the standard, integrated Stroop task with the primed Stroop task, in which the word and color are separated temporally (e.g., GREEN in monochrome followed by XXXX in red). Participants completed randomly intermixed standard (integrated) and primed Stroop trials: In Experiment 1, the prime-target stimulus-onset asynchrony (SOA) was 550 ms; and in Experiment 2, it was 250 ms. In the standard task, interference was large and increased across quantiles, replicating the classic delta plot pattern. In contrast, interference in the primed Stroop task was smaller and remained constant across the RT distribution - producing a flat delta slope, with both long and short SOAs. We take these results as support for our claim that the robustness of Stroop interference stems from the perceptual integration of word and color into a single visual object, and when this integration is broken as in the primed Stroop task, the resulting interference effect reflects a qualitatively different mechanism; thus, the effect in the primed Stroop task is not a "Stroop effect."