Optimistic Predictions Under Uncertainty: Active Information Search Both Supports and Constrains Motivated Bias

Across three studies, we examined how outcome preferences shape predictions when prediction-relevant information must be actively acquired through search versus passively received. Participants in our studies predicted the color of randomly selected squares from grids that either were fully visible or required self-directed search to uncover. Consistent with prior work on confirmation bias and selective exposure, Study 1 found that outcome preferences biased search behavior: participants preferentially sampled evidence associated with preferred outcomes and adjusted how much information they gathered based on early evidence. Surprisingly, however, these search biases did not translate into stronger wishful thinking compared with a passive search context, in which all information was immediately available. At first glance, this appears inconsistent with theories of motivated reasoning (e.g., Kunda, 1990), which suggest that flexibility in search should amplify the impact of directional motives. Closer inspection, however, revealed a countervailing process: the magnitude of desirability bias in predictions decreased as participants searched more extensively. In other words, while outcome preferences shaped both the direction and extent of search, prolonged search also attenuated wishful thinking, producing comparable overall bias across active and passive search conditions. Follow-up Studies 2 and 3 suggest that mechanisms related to effort justification and selective stopping of search could account for this pattern of results. Together, these findings highlight a more nuanced view of motivated cognition, in which outcome preferences, search strategies, and effort jointly determine the expression of motivated bias in predictions.