The Gratton effect remains after controlling for contingencies and stimulus repetitions

Abstract

The conflict monitoring hypothesis signals the need for cognitive controlThe Gratton effect is a key result attributed to the conflict monitoring hypothesisSome argue that controlling binding confounds eliminates the Gratton effect A Gratton effect remains in a vocal Stroop task after eliminating confounds The Gratton effect, the observation that the size of the Stroop effect is larger following a congruent trial compared to an incongruent trial, is one pivotal observation in support of the conflict-monitoring hypothesis. Previous reports have demonstrated that non-conflict components, such as feature binding, also contribute to this effect. Critically, Schmidt and De Houwer (2011) report a flanker task and a button-press Stroop task suggesting that there is no conflict adaptation in the Gratton effect; it is entirely caused by feature binding. The current investigation attempts to replicate and extend this important finding across two experiments using a canonical four-choice Stroop task with vocal responses. In contrast to Schmidt and De Houwer, we observe reliable conflict adaptation after controlling for feature binding. We argue that the overall strength of conflict is critical for determining whether a conflict adaptation component will remain in the Gratton effect after explaining binding components.

Keywords: Gratton effect; Stroop effect; cognitive control; cognitive control mechanisms; conflict adaptation; conflict monitoring; feature binding.

Figure 1

The size of the Gratton effect as a function of the proportion of congruent trials. Black lines represent data in which all trials are included in the estimate. Gray lines represent data in which only the trials without feature repetitions are included in the estimate. The panel to the left (A) reflects response time difference scores and the panel to the right (B) reflects error rate difference scores. A positive slope is consistent with the hypothesis that the contingency between the color and the word contributes to the size of the Gratton effect. Error bars represent the standard error of the mean.