AI-Assisted Decision-making: a Cognitive Modeling Approach to Infer Latent Reliance Strategies
Open accessAbstract
Abstract AI assistance is readily available to humans in a variety of decision-making applications. In order to fully understand the efficacy of such joint decision-making, it is important to first understand the human’s reliance on AI. However, there is a disconnect between how joint decision-making is studied and how it is practiced in the real world. More often than not, researchers ask humans to provide independent decisions before they are shown AI assistance. This is done to make explicit the influence of AI assistance on the human’s decision. We develop a cognitive model that allows us to infer the latent reliance strategy of humans on AI assistance without asking the human to make an independent decision. We validate the model’s predictions through two behavioral experiments. The first experiment follows a concurrent paradigm where humans are shown AI assistance alongside the decision problem. The second experiment follows a sequential paradigm where humans provide an independent judgment on a decision problem before AI assistance is made available. The model’s predicted reliance strategies closely track the strategies employed by humans in the two experimental paradigms. Our model provides a principled way to infer reliance on AI-assistance and may be used to expand the scope of investigation on human-AI collaboration.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The major disadvantage of measuring appropriateness by the percentage of reliance on correct AI advice and non-reliance on incorrect AI advice is that it cannot distinguish correct discrimination from mere overlap of human and AI decisions.
“The major disadvantage of this measurement is that we cannot know whether the reliance on correct AI advice stems from a correct discrimination or simply an overlap of the human and the AI's decision, i.e. instances where the AI advisor just confirms the human [70].”
From Appropriate Reliance on AI Advice: Conceptualization and the Effect of Explanations (2023) · cites this paper as [70]
In the section on assessing reliance, the paper says that agreement between a human’s final decision and the AI’s prediction may reflect either reliance on the AI or a decision the human would have made independently. This supports the claimed ambiguity, but the passage discusses agreement in the concurrent paradigm, not specifically a percentage measure combining reliance on correct advice with non-reliance on incorrect advice. No page number is available in our copy. (LLM, conf=0.83)
Experimental — accuracy not yet measuredImproved metacognition helps individuals with management of time, focus, and effort, problem-solving, academic performance, emotional well-being, and overall decision-making.
“Studies have shown that improved metacognition helps individuals with management of time, focus, and effort [212], problem-solving [67], academic performance [45,52,106,126,180,214], emotional well-being [200], and overall decision-making [206].”
From The Metacognitive Demands and Opportunities of Generative AI (2024) · cites this paper as [180]
The discussion under “Classifier C” links participants’ understanding of their own expertise and confidence to effective AI-reliance strategies in a simple image-classification task. That supports a narrow connection between self-assessment and decision-making, but the paper does not show that improving metacognition improves overall decision-making. Citation [206] is attached only to the decision-making part of the citing sentence. (LLM, conf=0.82)
Experimental — accuracy not yet measuredIn other work, a latent construct has been derived to measure reliance behaviour.
“In other work, a latent construct has been derived to measure reliance behaviour [70].”
From Appropriate Reliance on AI Advice: Conceptualization and the Effect of Explanations (2023) · cites this paper as [70]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
The citation lists haven’t been fetched yet
References 36 · Cited by 53
Research works printed but not linked: 15
- Becker, F., Skirzyński, J., van Opheusden, B., & Lieder, F. (2022). Boosting human decision-making with AI-generated decision aids. arXiv preprint arXiv:2203.02776 · arXiv
- Castelo, N., Bos, M. W., & Lehmann, D. R. (2019). Task-dependent algorithm aversion. Journal of Marketing Research, 56, 809-825.
- Cummings, M. L. (2017). Automation bias in intelligent time critical decision support systems. In Decision Making in Aviation (pp. 289-294). Routledge.
- Fox, J.-P. (2010). Bayesian item response modeling: Theory and appli- cations. Springer Science & Business Media.
- Geirhos, R., Medina Temme, C., Rauber, J., Schütt, H., Bethge, M., & Wichmann, F. (2019). Generalisation in humans and deep neural networks. In Thirty-second Annual Conference on Neural Information Processing Systems (NeurIPS 2018) (pp. 7549-7561).
- Kleinberg, J., Lakkaraju, H., Leskovec, J., Ludwig, J., & Mullaina- than, S. (2018). Human decisions and machine predictions. The Quarterly Journal of Economics, 133, 237-293.
- Kumar, A., Patel, T., Benjamin, A. S., & Steyvers, M. (2021). Explaining algorithm aversion with metacognitive bandits. In Proceedings of the Annual Meeting of the Cognitive Science Society. volume 43.
- Lu, Z., & Yin, M. (2021). Human reliance on machine learning mod- els when performance feedback is limited: Heuristics and risks. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems (pp. 1-16).
- Park, J. S., Barber, R., Kirlik, A., & Karahalios, K. (2019). A slow algorithm improves users' assessments of the algorithm's accu- racy. Proceedings of the ACM on Human-Computer Interaction, 3, 1-15.
- Phillips, P. J., Yates, A. N., Hu, Y., Hahn, C. A., Noyes, E., Jackson, K., Cavazos, J. G., Jeckeln, G., Ranjan, R., Sankaranarayanan, S., et al. (2018). Face recognition accuracy of forensic examiners, superrecognizers, and face recognition algorithms. Proceedings of the National Academy of Sciences, 115, 6171-6176.
- Rastogi, C., Zhang, Y., Wei, D., Varshney, K. R., Dhurandhar, A., & Tomsett, R. (2022). Deciding fast and slow: The role of cognitive biases in AI-assisted decision-making. Proceedings of the ACM on Human-Computer Interaction, 6, 1-22.
- Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., et al. (2015). ImageNet large scale visual recognition challenge. International Journal of Computer Vision, 115, 211-252.
- Schaffer, J., O'Donovan, J., Michaelis, J., Raglin, A., & Höllerer, T. (2019). I can do better than your AI: Expertise and explanations. In Proceedings of the 24th International Conference on Intelligent User Interfaces (pp. 240-251).
- Simonyan, K., & Zisserman, A. (2014). Very deep convolutional net- works for large-scale image recognition. arXiv preprint arXiv: 1409. 1556
- Steyvers, M., Tejeda, H., Kerrigan, G., & Smyth, P. (2022). Bayes- ian modeling of human-AI complementarity. Proceedings of the National Academy of Sciences, 119, e2111547119.
- Authors
- Tejeda, H., Kumar, A., Smyth, P., & Steyvers, M.
- Affiliations
- University of California, Irvine (US)
- Corresponding author
- Tejeda, Heliodoro
- Venue
- Computational Brain & Behavior
- Published
- 2022-10-19
- Volume
- 5
- Issue
- 4
- Pages
- 491-508
- ISSN
- 2522-0861 (ISSN-L)
- 2522-087X
- OpenAlex
- W4306809523
- Cited by
- 53
- References
- 36
- Citations by year
- 2026: 12 · 2025: 22 · 2024: 15 · 2023: 4
- FWCI
- 9.68
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Locations
- Computational Brain & Behavior (publishedVersion, cc-by)
- Funders
- National Science Foundation (1900644, 1927245)
- Topics
- Social Sciences › Psychology › Social Psychology › Human-Automation Interaction and Safety
- Social Sciences › Decision Sciences › General Decision Sciences › Decision-Making and Behavioral Economics
- Social Sciences › Psychology › Social Psychology › Social Robot Interaction and HRI
- Keywords
- AI-assisted decision-making, AI reliance, human-AI collaboration, cognitive modeling, shared decision-making, human-AI interaction
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
