My Model is Unfair, Do People Even Care? Visual Design Affects Trust and Perceived Bias in Machine Learning

Gaba, Aimen; Kaufman, Zhanna; Cheung, Jason; Shvakel, Marie; Hall, Kyle Wm.; Brun, Yuriy; Bearfield, Cindy Xiong

doi:10.1109/tvcg.2023.3327192

Cited by 2 publications

(3 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Behavioral trust is the action resulting from attitudinal trust, reflecting genuine preferences and decisions. Inspired by a variant of the trust game where participants choose a game partner [109], and the two-alternative forced choice task used for comparing trust in two models [53], we measure behavioral trust by asking participants to choose among different forecast websites. We frame each display (e.g., dotplot) as one forecast website, and represent each using an icon (e.g.,…”

Section: Methodsmentioning

confidence: 99%

“…Trust. Many research communities, such as AI/ML [22,60], economics [26,84,96], political science [29,95], and visualization [35,47,48,53], have attended to trust through different lenses. As a fundamental aspect of social structures, trust determines how people interact with others and utilize new techniques.…”

Section: Related Workmentioning

confidence: 99%

“…Given the non-conclusive effects of probability correction and calibration on behavioral trust, we conduct a small-scale exploratory analysis based on participants' choices in the last forecast/election cycle. Prior work identified gender [53] and partisan [113] differences in trust. In the present work, we find that the partisan divide is more consistent than any gender difference.…”

Section: Non-preregistered Exploratory Analysismentioning

confidence: 99%

See 2 more Smart Citations

In Dice We Trust: Uncertainty Displays for Maintaining Trust in Election Forecasts Over Time

Yang,

Mortenson,

Nisbet

et al. 2024

Preprint

View full text Add to dashboard Cite

Trust in high-profile election forecasts influences the public’s confidence in democratic processes and electoral integrity. Yet, maintaining trust after unexpected outcomes like the 2016 U.S. presidential election is a significant challenge. Our work confronts this challenge through three experiments that gauge trust in election forecasts. We generate simulated U.S. presidential election forecasts, vary win probabilities and outcomes, and present them to participants in a professional-looking website interface. In this website interface, we explore (1) four different uncertainty displays, (2) a technique for subjective probability correction, and (3) visual calibration that depicts an outcome with its forecast distribution. Our quantitative results suggest that text summaries and quantile dotplots engender the highest trust over time, with observable partisan differences. The probability correction and calibration show small-to-null effects on average. Complemented by our qualitative results, we provide design recommendations for conveying U.S. presidential election forecasts and discuss long-term trust in uncertainty communication. We provide preregistration, code, data, model files, and videos at https://doi.org/10.17605/OSF.IO/923E7.

show abstract

Section: Methodsmentioning

confidence: 99%