Shin, Sally MJ, Paul Rozin, and Gal Zauberman, “Moral Surgery: Negative Revelations Differentially Alter Retrospective Evaluations of Creators and Creations,” Forthcoming at Journal of Experimental Social Psychology

When individuals learn of a moral transgression committed by a creator they once admired, they face psychological tension: How do they reconcile their positive prior evaluations with this newly acquired negative information? While prior research has primarily examined how such revelations alter current attitudes, we investigate how they trigger the differential reconstruction of past attitudes. Across five preregistered experiments, we reveal a robust pattern showing that individuals reconstruct their prior evaluations by separating the art from the artist. Specifically, although the revelation of a transgression prompts a significant downward revision in individuals’ recalled evaluations of the creator, their recalled evaluations of the creation remain relatively resistant to the negative information. This differential reconstruction occurs for both experimentally induced and pre-existing positive evaluations, across multiple creative domains, and for varying levels of transgression severity. Furthermore, through targeted stress-tests, we demonstrate that this observed phenomenon reflects a genuine psychological process rather than a methodological artifact. The effect is robust across incentive-compatible, multi-wave longitudinal, and separate measurement designs that rule out potential alternative explanations such as low cognitive effort, temporal proximity, and demand characteristics.

Shin, Sally MJ, and Alexander G. Fulmer, “Single Farm-to-Table: Preference for Foods with a Single Source of Origin,” Revising for 3rd round review at Journal of Marketing Research

While food and beverage companies frequently communicate origin-related messages about their products, their effectiveness remains largely unknown. The current research investigates the impact of a product’s origin numerosity (i.e., whether its core element is sourced from one versus multiple location(s)) on consumer preference. Nine experiments reveal that consumers favor products sourced from a single origin over otherwise identical products sourced from multiple origins. This preference extends beyond categories with strong marketplace norms favoring single origin (e.g., wine and whiskey) to those where such norms are weaker (e.g., hot sauce, cranberry sauce, and peanut butter). The single origin preference persists across purchase intentions and incentive-compatible choices, using both controlled, minimal stimuli and richer marketplace-inspired operationalizations. This effect is driven by perceptions of essential purity, such that single origin products are perceived as having greater preservation of essential properties. Consequently, information that enhances the perceived purity of a multi-origin product attenuates the preference for single origin numerosity, offering theory-driven and actionable solutions to benefit producers of multi-origin products. Finally, this preference reverses when consumers evaluate bundle packs of products such that multi-origin bundles are preferred to single origin bundles. These findings offer theoretically rich and practically relevant insights into origin-related marketing communications.

Shin, Sally MJ, Nirajana Mishra, and Ravi Dhar, “From Lemons to LemonAIde: How Low Brand Trust Increases Consumer Receptivity to Artificial Intelligence,” Revising for 3rd round review at Journal of Consumer Research

As brands increasingly develop consumer-facing artificial intelligence (AI), understanding how key brand characteristics influence preferences for AI- relative to human-based systems becomes important. Across ten experiments (and five supplemental studies), we identify a novel and counterintuitive role of brand trust: consumers exhibit a greater relative preference for product recommendations from a brand’s AI system when trust in that brand is low (vs. high). This effect is robust across product categories and trust signals, and generalizes beyond preference for individual agents to preferences for firm-level system implementations and recommendation compliance in incentive-compatible settings. We propose that this effect is driven by Differential Essence Transfer: the belief that AI-based systems are weaker embodiments of the brand’s essence (i.e., its (un)trustworthiness) than human-based systems. We find support for this mechanism via moderation, such that consumers who hold stronger beliefs in Differential Essence Transfer amplify the impact of trust on AI preference. We discuss and rule out potential alternative explanations including perceived outsourcing, brand warmth, and brand liking, empirically isolating trust as the key driver of our effect. These findings offer theoretically novel contributions and actionable insights into how managers of brands with different reputational standing can align their deployment of consumer-facing AI-driven services.