On Human Factors in Machine Fairness: Essays in Behavioral Economics
Authors: AC Wömmel
Published: 2025
Publication: 2025 - ediss.sub.uni-hamburg.de
Human behavior can undermine fairness interventions in AI, amplify inequalities, and reinforce socioeconomic disparities, emphasizing the need to integrate behavioral mechanisms into AI governance.
Methods: The dissertation utilized three empirical approaches: (i) deliberation experiments with UK participants using NLP analysis, (ii) online hiring experiments testing algorithmic fairness interventions, and (iii) panel data analysis of German households measuring digital skills and confidence levels.
Key Findings: Public attitudes towards AI, the adoption of fairness interventions in algorithmic tools, and socioeconomic disparities in digital skills and confidence.
Limitations: Potential variability across cultural and geographic contexts, limited scope for longitudinal behavioral analysis, and risk of biases in self-reported data in household panels.
Institution: University of Hamburg
Research Area: Machine Fairness, Behavioral Economics, Human-Machine Interaction
Discipline: Behavioral Economics, Artificial Intelligence, Human-Machine Interaction
Citations: 1