Reduce algorithm aversion by educating about the value of consistent weighting or convincing users that models can learn¶

Strategy · All strategies
Evidence · 5 claims (5 for) · 1 study (1 causal), q3 · 0 of 1 report an effect size · 5 claims rest on one study

Description¶

Based on exploratory attribute comparisons in Studies 2–4, the authors recommend that practitioners "attempt to reduce algorithm aversion by either educating people of the importance of providing consistent and appropriate weights, or by convincing them that models can learn or that humans cannot." Participants in their data credited the model with better consistency and weighing and credited humans with learning and improvement, suggesting these perceptions as intervention targets.

Design Implications¶

Context¶

Requirements¶

• Interventions must address the specific perceived attribute gaps: consistent and appropriate weighting by the model, and perceived ability to learn from mistakes

Constraints¶

• The article presents these as preliminary findings for future research to build on, not as tested interventions

Target Learners¶

• adult decision makers considering algorithmic decision aids

Target Learning Goals¶

• willingness to adopt superior algorithmic forecasting tools

Affordances¶

• Algorithm Aversion Theory

Claims¶

• Algorithm aversion persists even among participants who saw the algorithm outperform the human forecaster [+W]
• People lose confidence more quickly in algorithms than in humans after seeing the same kind of mistakes [+W]
• Confidence in the algorithm's forecasts mediates the effect of seeing it perform on choosing it [+W]
• A belief that the algorithm is less likely to be perfect does not explain experience-driven algorithm aversion [+W]
• Seeing an algorithm perform makes people significantly less likely to choose it over a human forecaster [+W]

Related Strategies¶

Examples¶

-

Key Sources¶

• Berkeley J. Dietvorst, Joseph P. Simmons, and Cade Massey. (2014). Algorithm Aversion: People Erroneously Avoid Algorithms After Seeing Them Err. Journal of Experimental Psychology: General. https://doi.org/10.1037/xge0000033