Evaluating the learning of stimulus-control associations through incidental memory of reinforcement events

A spatial version of the Stroop task for examining proactive and reactive control independently from non-conflict processes

Conflict-induced control refers to humans' ability to regulate attention in the processing of target information (e.g., the color of a word in the color-word Stroop task) based on experience with conflict created by distracting information (e.g., an incongruent color word), and to do so either in a proactive (preparatory) or a reactive (stimulus-driven) fashion. Interest in conflict-induced control has grown recently, as has the awareness that effects attributed to those processes might be affected by conflict-unrelated processes (e.g., the learning of stimulus-response associations). This awareness has resulted in the recommendation to move away from traditional interference paradigms with small stimulus/response sets and towards paradigms with larger sets (at least four targets, distractors, and responses), paradigms that allow better control of non-conflict processes. Using larger sets, however, is not always feasible. Doing so in the Stroop task, for example, would require either multiple arbitrary responses that are difficult for participants to learn (e.g., manual responses to colors) or non-arbitrary responses that can be difficult for researchers to collect (e.g., vocal responses in online experiments). Here, we present a spatial version of the Stroop task that solves many of those problems. In this task, participants respond to one of six directions indicated by an arrow, each requiring a specific, non-arbitrary manual response, while ignoring the location where the arrow is displayed. We illustrate the usefulness of this task by showing the results of two experiments in which evidence for proactive and reactive control was obtained while controlling for the impact of non-conflict processes.

Cognitive control controls the effect of irrelevant stimulus-response learning

Research has established that two cognitive processes, cognitive control and irrelevant stimulus-response (S-R) learning, may underlie the proportion congruency effect, which refers to the findings that the size of interference effects (e.g., the Stroop, Simon, or Eriksen flanker effect) reduces with increasing the proportion of incongruent trials. Further studies have begun to investigate the interaction between these two cognitive processes, which not only provide more plausible accounts for empirical data, but also advance theories. The present study set out to investigate whether cognitive control can modulate the effect of irrelevant S-R learning. In two experiments, we combined a color-letter contingency task, in which we manipulated the contingencies (low vs. high) of irrelevant S-R associations, with a color-Chinese character Stroop task, in which we manipulated the ratio of neutral to incongruent trials (mostly neutral (MN) versus mostly incongruent (MI)). Experiment 1 showed a proportion neutral effect (the Stroop effect was smaller in the MI than in the MN condition), suggesting changes in control demand. Critically, the contingency effect (faster responses in the high- than in the low-contingency condition) reduced in the MI than in the MN condition. Experiment 2 (preregistered) increased the number of Chinese characters to exclude a familiarity account for the proportion neutral effect, which replicated the findings of Experiment 1. These results suggest that cognitive control induced in the Stroop task transferred to the contingency task and modulated the contingency effect. Thus, this study provides clear evidence that cognitive control can modulate the effect of irrelevant S-R learning.

Principles of cognitive control over task focus and task switching

Adaptive behaviour requires the ability to focus on a task and protect it from distraction (cognitive stability) and to rapidly switch tasks when circumstances change (cognitive flexibility). Burgeoning research literatures have aimed to understand how people achieve task focus and task switch readiness. In this Perspective, I integrate these literatures to derive a cognitive architecture and functional rules underlying the regulation of cognitive stability and flexibility. I propose that task focus and task switch readiness are supported by independent mechanisms. However, I also suggest that the strategic regulation of both mechanisms is governed by shared learning principles: an incremental, online learner that nudges control up or down based on the recent history of task demands (a recency heuristic) and episodic reinstatement when the current context matches a past experience (a recognition heuristic). Finally, I discuss algorithmic and neural implementations of these processes, as well as clinical implications.

Minimal impact of consolidation on learned switch-readiness

Adaptive behavior is characterized by our ability to create, maintain, and update (or switch) rules by which we categorize and respond to stimuli across changing contexts (cognitive flexibility). Recent research suggests that people can link the control process of task-switching to contextual cues through associative learning, whereby the behavioral cost of switching is reduced for contexts that require frequent switching. One example is the listwide proportion switch (LWPS) effect, denoting smaller switch costs in blocks of trials where switching is more frequent. However, the conditions that govern such learned cognitive flexibility are poorly understood. One major unanswered question is whether this type of learning benefits from memory consolidation effects. To address this question, we manipulated whether task-sets and/or specific task stimuli were more frequently linked with task-switching (vs. repeating), and ran participants over two experimental sessions, separated by a 24-hr delay. We expected that consolidation would facilitate learned cognitive flexibility, resulting in a greater reduction of switch costs with increasing task-switch likelihood on Session 2 compared with Session 1. Across two experiments, we observed robust LWPS effects in both sessions. However, we found little evidence for effects of consolidation on learned cognitive flexibility: The magnitude of the LWPS effect did not change from Session 1 to 2. Altogether our results suggest that people reliably and quickly acquire task-set and stimulus-based switch associations, but this form of control learning-unlike many instances of reward-based learning-does not benefit from long-term memory consolidation. Possible reasons for these findings are discussed. (PsycInfo Database Record (c) 2022 APA, all rights reserved).