AI Error Difficulty Modulates the Effectiveness of Explainability in Decision Support Systems
ACM Transactions on Computer-Human Interaction · 31.08.2026 · 33 (4) · 1-34
DOI: 10.1145/3817603 ↗Kurz gefasst
Fachartikel in der Zeitschrift ACM Transactions on Computer-Human Interaction, online erschienen am 08.08.2026. Verfasst von Tobias Rieger und 3 weiteren Autor:innen. Die Arbeit nennt 156 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
In AI decision support, explainability (e.g., disclosing system weaknesses) should help users spot erroneous recommendations, but its efficacy may depend on task difficulty. We ran three experiments in a simulated medical visual detection task. In Experiment 1, we manipulated error difficulty (easy vs. difficult) and explainability (non-XAI vs. XAI). Experiment 2 added a virtually impossible error difficulty. Across both, explainability consistently reduced reliance on incorrect recommendations for difficult errors, showed no benefit for easy errors, and showed a small benefit for impossible errors. Experiment 3 varied error and task difficulty within-subjects and extended these patterns; as task difficulty rose, participants behaved less rationally, exhibiting both under- and overreliance. Notably, these behavioral benefits were generally not accompanied by reduced trust in the AI system. Our findings suggest that disclosing system weaknesses enhances detection of AI errors but is most effective for tasks of moderate difficulty where AI recommendations are still verifiable.
Zitieren
APA
Rieger, T., Schindler, H., Koch, K., & Onnasch, L. (2026). AI Error Difficulty Modulates the Effectiveness of Explainability in Decision Support Systems. ACM Transactions on Computer-Human Interaction, 33(4), 1-34. https://doi.org/10.1145/3817603
Vancouver
Rieger T, Schindler H, Koch K, Onnasch L. AI Error Difficulty Modulates the Effectiveness of Explainability in Decision Support Systems. ACM Transactions on Computer-Human Interaction. 2026;33(4):1-34. doi:10.1145/3817603
BibTeX
@article{Rieger2026,
author = {Rieger, Tobias and Schindler, Hanna and Koch, Katharina and Onnasch, Linda},
title = {AI Error Difficulty Modulates the Effectiveness of Explainability in Decision Support Systems},
journal = {ACM Transactions on Computer-Human Interaction},
year = {2026},
volume = {33},
number = {4},
pages = {1--34},
doi = {10.1145/3817603}
}
EUVIMED