Mackenzie J. Siesel¹ (), Tianyu Zhang¹, Andrew B. Leber¹; ¹The Ohio State University

In daily life, people can choose among multiple visual search strategies to find a target of interest (e.g., car in a parking lot). Lab-based studies have revealed that people’s chosen strategies are surprisingly suboptimal. While several factors may be responsible, lacking explicit knowledge of the optimal strategy is a key factor predicting suboptimality. Here we examine how explicit knowledge and motivational factors interact to jointly predict strategy. Specifically, will individuals persist in using optimal strategy instructions even when the behavioral benefits are reduced? We further assessed the frequency with which people monitor the utility of their strategy: if they knowingly abandon the optimal strategy when it is perceived to be less useful, how quickly will they reinstate it if the behavioral benefits return? We used the Adaptive Choice Visual Search (Irons & Leber, 2018) to assess strategy usage. In this task, two targets are presented in both a red and blue subset, but only one must be found. The optimal strategy in this task is to search through the less numerous color subset. For the first and third phases of the experiment, we used the standard paradigm in which the subsets were presented in a 2:1 ratio on all trials. In the critical second phase, to disincentivize the optimal strategy (i.e., reduce its behavioral benefits), we mixed color subset ratios such that 30% of trials had the typical 2:1 ratio and 70% had a 1.1:1 ratio. Even with explicit strategy instruction, optimality was significantly decreased in the mixed ratio blocks. However, optimality rapidly returned to its Phase-1 level when the standard ratio was restored in Phase 3. These findings show that explicit strategy knowledge is sometimes insufficient to elicit the optimal strategy, but this choice likely reflects the judicious – and frequent – reevaluation of the strategy’s expected behavioral benefits.

Acknowledgements: This work is supported by NSF BCS-2021038 to ABL.