Task-Dependent Algorithm Aversion
The DOI matches its Crossref record.
Abstract
Research suggests that consumers are averse to relying on algorithms to perform tasks that are typically done by humans, despite the fact that algorithms often perform better. The authors explore when and why this is true in a wide variety of domains. They find that algorithms are trusted and relied on less for tasks that seem subjective (vs. objective) in nature. However, they show that perceived task objectivity is malleable and that increasing a task’s perceived objectivity increases trust in and use of algorithms for that task. Consumers mistakenly believe that algorithms lack the abilities required to perform subjective tasks. Increasing algorithms’ perceived affective human-likeness is therefore effective at increasing the use of algorithms for subjective tasks. These findings are supported by the results of four online lab studies with over 1,400 participants and two online field studies with over 56,000 participants. The results provide insights into when and why consumers are likely to use algorithms and how marketers can increase their use when they outperform humans.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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?
In the citing paper
“Including humans in algorithmic decision-making processes costs time [20,29,68]”
From Disentangling Fairness Perceptions in Algorithmic Decision-Making: the Effects of Explanations, Human Oversight, and Contestability (2023) · cites this paper as [20]
In the citing paper
“Later studies have demonstrated that the issue of algorithm aversion is especially acute in subjective, creative, or morally charged areas (Castelo et al., 2019;Bigman and Gray, 2018).”
From The Impact of Generative AI on Personalized Content Marketing in E-Commerce (2025) · cites this paper as Castelo et al., 2019
In the citing paper
“Castelo et al. (2019) suggest a single framework: consumers perceive AI as more competent but less warm compared to humans.”
From The Impact of Generative AI on Personalized Content Marketing in E-Commerce (2025) · cites this paper as Castelo et al. (2019)
In the citing paper
“People are generally more willing to rely on al gorithms for objective tasks than for subjective ones (Castelo, Bos, & Lehmann, 2019; Logg et al., 2019)”
From Algorithm aversion in academic libraries: User resistance to AI-enabled discovery and search services (2026) · cites this paper as Castelo, Bos, & Lehmann, 2019
In the citing paper
“When AI-based algorithm behavior turns out to be human-like, users tend to trust and follow it, regardless of the exact task (and level of reliability) (Castelo et al., 2019).”
From Why do users trust algorithms? A review and conceptualization of initial trust and trust over time (2022) · cites this paper as Castelo et al., 2019
1 more claim about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
No references could be resolved yet.
Research works printed but not linked: 51
- Armstrong J. Scott, 1980, Technology Review, 82, 16
- 10.1287/mnsc.15.5.215 · DOI · Paper page
- 10.1073/pnas.1517780113 · DOI · Paper page
- 10.1207/S15324834BASP2502_3 · DOI · Paper page
- 10.1287/mnsc.36.8.887 · DOI · Paper page
- 10.1023/A:1008047504898 · DOI
- 10.1109/HICSS.2003.1174902 · DOI · Paper page
- 10.1287/mnsc.35.8.982 · DOI · Paper page
- 10.1126/science.2648573 · DOI · Paper page
- 10.1111/j.1468-2389.2011.00548.x · DOI · Paper page
- 10.1287/mnsc.2016.2643 · DOI · Paper page
- Dietvorst Berkeley J., 2014, Journal of Experimental Psychology: General, 143, 1
- 10.1126/science.1121629 · DOI · Paper page
- 10.1037/a0024354 · DOI · Paper page
- 10.1002/bdm.741 · DOI · Paper page
- 10.2307/258191 · doi:10.2307/258191 · Paper page
- 10.1007/s12369-016-0338-y · DOI · Paper page
- 10.1111/j.1756-8765.2008.01006.x · DOI · Paper page
- Gray Heather M., 2007, Science, 619, 10
- 10.1016/j.cognition.2012.06.007 · DOI · Paper page
- 10.1177/1088868310367640 · DOI · Paper page
- 10.1037/1040-3590.12.1.19 · DOI · Paper page
- 10.1207/s15327957pspr1003_4 · DOI · Paper page
- 10.1037/0022-3514.89.6.937 · DOI · Paper page
- 10.1521/soco.2008.26.2.248 · DOI · Paper page
- 10.1080/09298215.2014.881888 · DOI
- 10.1037/a0020215 · DOI · Paper page
- 10.1016/S0148-2963(03)00140-1 · DOI · Paper page
- 10.1109/FG.2013.6553780 · DOI
- 10.2307/25148760 · DOI · Paper page
- 10.1016/j.jsis.2008.01.001 · DOI · Paper page
- Logg Jennifer M. (2017), “Theory of Machine: When Do People Rely on Algorithms?” Working Paper No. 17-086, Harvard Business School, Harvard University. · DOI
- 10.1016/j.obhdp.2018.12.005 · DOI · Paper page
- 10.1093/jcr/ucz013 · DOI · Paper page
- 10.1111/j.1467-9280.2007.01858.x · DOI · Paper page
- 10.1073/pnas.1710966114 · DOI · Paper page
- 10.1109/CVPRW.2013.130 · DOI · Paper page
- Meehl Paul E., 1954, Clinical vs. Statistical Prediction: A Theoretical Analysis and a Review of the Evidence
- O’Neil Cathy, 2016, Weapons of Math Destruction
- 10.1002/bdm.637 · DOI · Paper page
- 10.1287/isre.1040.0015 · DOI · Paper page
- 10.1016/j.jesp.2017.01.006 · DOI · Paper page
- 10.1037/a0026966 · DOI · Paper page
- 10.1002/bdm.542 · DOI · Paper page
- 10.1086/208642 · DOI · Paper page
- 10.1177/0272989X12453501 · DOI · Paper page
- 10.1111/j.0272-4332.2004.00433.x · DOI · Paper page
- 10.1006/obhd.1995.1040 · DOI · Paper page
- 10.1037/a0015145 · DOI · Paper page
- 10.1016/S1057-7408(07)70013-X · DOI · Paper page
- 10.1002/bdm.2118 · DOI · Paper page
Cited by
Items 1,001–1,007 of 1,007
Kim, K., Yoo, J., Jeong, S.-H., & Ihm, J. (2026) · International Journal of Business Communication · doi:10.1177/23294884261498410
Guo, K., & Wang, L. (2026) · PeerJ Computer Science · doi:10.7717/peerj-cs.4143
Andrianto, T., Susanto, E., Septyandi, C. B., & Suryana, M. (2026) · ABAC Journal · doi:10.1108/abac-06-2026-0219
The effects of AI identity disclosure on user responses: A meta-analysis
Chen, H., Sun, M., Chen, H., & Liu, B. (2026) · Information & Management · doi:10.1016/j.im.2026.104411 · Cited by 1
Peng, X., Wang, X., Zhang, J., Liang, Y., Jiang, H., & Yang, J. (2026) · International Journal of Industrial Ergonomics · doi:10.1016/j.ergon.2026.104042
Chen, X., Ji, L., Zhuang, M., & Tang, C. (2026) · Information & Management · doi:10.1016/j.im.2026.104464
Diwya, K. G. S., Alberta, N. A. J., Prasetya, K. A., & Gandhi, I. (2026) · Journal of Risk and Financial Management · doi:10.3390/jrfm19110861
Metadata
- Authors
- Castelo, N., Bos, M. W., & Lehmann, D. R.
- Affiliations
- University of Alberta (CA)
- Snap (United States) (US)
- Columbia University (US)
- Venue
- Journal of Marketing Research
- Publisher
- SAGE Publications
- Published
- 2019-07-15
- Volume
- 56
- Issue
- 5
- Pages
- 809-825
- ISSN
- 0022-2437
- 1547-7193
- OpenAlex
- W2961720344
- Cited by
- 1,500
- References
- 0
- Citations by year
- 2026: 480 · 2025: 311 · 2024: 251 · 2023: 199 · 2022: 138 · 2021: 90 · 2020: 30 · 2019: 1
- FWCI
- 81.29
- Type
- book chapter
- Language
- en
- Open access
- closed
- Licence
- journals.sagepub.com/page/policies/text-and-data-mining-license (tdm)
- Locations
- Springer eBooks (publishedVersion)
- Journal of Marketing Research (publishedVersion)
- Topics
- Social Sciences › Business, Management and Accounting › Marketing › AI in Service Interactions
- Social Sciences › Decision Sciences › Information Systems and Management › Technology Adoption and User Behaviour
- Social Sciences › Human Factors and Ergonomics › Human-Automation Interaction and Safety
- Keywords
- algorithm aversion
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.