Reliance on Numerical Precision: Compatibility between Accuracy versus Efficiency Goals and Numerical Precision Level Influence Attribute Weighting in Two‐Stage Decisions

Abstract

This research examines how the weighting of an attribute is jointly affected by attribute precision and decision stage. Building on prior work suggesting (a) that less (more) precise numerical values are easier to process (more accurate), (b) that decision‐makers' motivation to be efficient (accurate) is greater when creating a consideration set (making a final choice), and (c) that decision‐makers tend to overweight information that is compatible with their goals, we hypothesize that when creating a consideration set (making a choice) participants tend to assign greater weight to less (more) precise attributes. Five studies (two of them reported in the Appendix S1) offer triangulating evidence for these predictions. Overall, this work contributes to research on numerical cognition, efficiency versus accuracy trade‐offs, attribute weighting, and two‐stage decisions.

To be precise (imprecise) in utilitarian (hedonic) contexts: Examining the influence of numerical precision on consumer reactions to artificial intelligence‐based recommendations

“…Xie and Kronrod (2012) suggested that when presented with precise numbers, consumers who are low in advertising skepticism are likely to perceive the advertised company as more competent. Pena‐Marin and Yan (2021) investigated how attribute precision and decision stage affect the weighting of an attribute and found that when creating a consideration set (making a decision), participants tend to assign greater weight to less (more) precise attributes since they generally overestimate the relevance of data that fits their goals. The dominant findings in prior numerical cognition literature provide a basic understanding of how appropriate numerical precision in combination with decision context or product attributes can result in favorable evaluations, providing us a better understanding of the application of numerical forms in AI recommendation systems.…”

Comparing estimates for decision-making: Numerical processing and preferences for underestimates versus overestimates.

“…Probing their feelings, they may focus on fluency, their subjective experience of ease or difficulty in the mental task (Alter & Oppenheimer, 2008; Schwarz, 2004). Fluency has been shown to influence judgments of numerical information (Coulter & Roggeveen, 2014; King & Janiszewski, 2011; Pena-Marin & Yan, 2020). For instance, Biswas et al (2013) found that the ease of calculating the difference between two prices (original and sale price) improved sale prices evaluation.…”

A Multistage, Multiprocess Analysis of Consumer Judgment: A Selective Review and Conceptual Framework

“…The use of round precise numbers may depend on the goal at hand. Pena‐Marin and Yan () found that people are more influenced by attributes described in round numbers they construct a consideration set but prefer attribute descriptions in precise numbers at the decision stage.…”