Abstract
This study examined whether Gestalt-based manipulations can reduce individual's tendency to follow negative recency (NR) or positive recency (PR) by inducing their perception that events are not related. Two types of manipulation - grouping and closure - were introduced through a series of Coin Prediction Tasks. In the Grouping-based Coin Prediction Task (G-CPT), participants were more likely to exhibit NR or PR tendencies when past outcomes and predictions were presented within the same trial compared to when they were presented in separate trials. In the Closure-based Coin Prediction Task (C-CPT), previous results ending with a streak of a dominant event elicited more NR or PR responses, whereas previous outcomes ending with a non-dominant event reduced these fallacies. Overall, this study provides evidence that Gestalt-based manipulations can mitigate cognitive fallacies. Our findings emphasize the importance of Gestalt framing in probabilistic decision-making contexts. Limitations and directions for future research are also discussed.