Psychological Benefits and Costs of Diversifying Algorithmic Recourse

2026-05-12

AI Summary

This study investigates how algorithmically generated diverse counterfactual explanations balance users’ psychological benefits against cognitive costs. Through a between-subjects experiment (N=750), the authors systematically manipulated both the diversity and size of recommendation sets, integrating psychometric scales with counterfactual generation techniques. They demonstrate for the first time that small-scale diversity significantly enhances users’ intention to act without increasing cognitive load, whereas large-scale diversity—despite improving recommendation variety—substantially heightens cognitive burden and degrades user experience. These findings underscore the necessity of aligning algorithmic design with human cognitive constraints and provide empirical grounding for developing counterfactual recommendation methods that jointly optimize effectiveness and usability.

Abstract

Algorithmic recourse provides counterfactual action plans that help people overturn unfavorable AI decisions. While diverse recourse sets may improve transparency and motivation, they may also impose cognitive load and negative emotions by increasing counterfactual reasoning demands. To examine this trade-off, we conducted a between-subjects controlled experiment (N=750) that manipulated recourse-set diversity and size, and evaluated these effects on psychological benefits and costs. Results show that diversification enhances psychological benefits (e.g., willingness to act) for small sets without incurring additional psychological costs, whereas for large sets, it makes cognitive load more salient. These findings suggest that naively diversifying recourse can burden decision subjects, underscoring the need for new diversification methods that incorporate human cognition and psychology to mitigate such costs.

Problem

• algorithmic recourse
• diversity
• cognitive load
• counterfactual reasoning
• psychological costs

Innovation

• algorithmic recourse
• counterfactual explanations
• cognitive load
• human-centered AI
• diversification