Situation selection and cognitive conflict: explicit knowledge is necessary for conflict avoidance

ABSTRACT

Humans transform their environment in order to regulate their own affect. One way to do so is to avoid situations that come with negative rather than positive affect. This selection might not solely bear on expectations of full-blown emotions, but may also be invoked by anticipating the aversiveness of cognitive conflict, when a situation suggests competing behavioural responses. If cognitive conflict is indeed aversive, it may trigger affect regulation goals, which in turn influence choices of situations depending on the magnitude of conflict they contain. People should prefer actions that produce conflict-free situations to actions that produce conflicting situations. In three experiments, participants had to solve a Stroop task by freely choosing between response keys that were either associated with low-conflict or high-conflict in the subsequent trial. We find that people do not automatically prefer actions associated with conflict-free situations to actions that are associated with conflicting situations. They only do so, when they are explicitly informed about the contingency between action and congruency of an upcoming situation. This suggests that cognitive conflict, at least at the level of a standard conflict task as used here, is insufficient to invoke affect regulation processes.

KEYWORDS:

• Cognitive control
• affect regulation
• demand avoidance

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 As mentioned in the methods sections, we did a manipulation check of awareness by asking participants several questions, either at the end of the Experiment (1 & 2) or at the end of the implicit block (Experiment 3). One of those questions inquired whether participants discovered any differences between the two keys they could use to respond to a specific color. If they answered yes, they were asked to provide the differences they discovered in an open question. In Experiment 3, 39 participants answered that they did not discover any differences between the keys. Of the 9 participants who did, we looked at the differences they indicated and there were 4 participants whose answers could be classified as having figured out the contingency (e.g., “for the right responses, the color fitted with the word”). An exploratory analysis shows that their mean low conflict choice in the implicit condition was 65%, which appears to deviate from chance performance (a t-test is not significant, but severely underpowered due to the small sample size). However, these participants are included in the reported main Bayes analysis, which provides evidence for the hypothesis that overall performance is about chance level. So if anything, the test for the implicit condition in Experiment 3 is very liberal and should have found an effect if there was one. We did not preregister any exclusion of these participants in Experiment 3, so we are hesitant to change our reported analysis in hindsight.