Understanding algorithm aversion: When is advice from automation discounted?
Abstract
Forecasting advice from human advisors is often utilized more than advice from automation. There is little understanding of why “algorithm aversion” occurs, or specific conditions that may exaggerate it. This paper first reviews literature from two fields—interpersonal advice and human–automation trust—that can inform our understanding of the underlying causes of the phenomenon. Then, an experiment is conducted to search for these underlying causes. We do not replicate the finding that human advice is generally utilized more than automated advice. However, after receiving bad advice, utilization of automated advice decreased significantly more than advice from humans. We also find that decision makers describe themselves as having much more in common with human than automated advisors despite there being no interpersonal relationship in our study. Results are discussed in relation to other findings from the forecasting and human–automation trust fields and provide a new perspective on what causes and exaggerates algorithm aversion.
Integrity signals
Registry
Cites
Statistics
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?
Earlier studies have also shown tendencies toward underuse of algorithms.
“Earlier studies have also shown tendencies toward underuse of algorithms.25,26”
From Appropriate trust in artificial intelligence for the optical diagnosis of colorectal polyps: the role of human/artificial intelligence interaction (2024) · cites this paper as [26]
Cites
Trust and TAM in Online Shopping: An Integrated Model1
Gefen, D., Karahanna, E., & Straub, D. W. (2003) · MIS Quarterly · doi:10.2307/30036519 · Cited by 8,635 · see citing works
Computer-Mediated Communication: Impersonal, Interpersonal, and Hyperpersonal Interaction
WALTHER, J. B. (1996) · Communication Research · doi:10.1177/009365096023001001 · Cited by 4,798 · see citing works
Mass Communication and Para-Social Interaction
Horton, D., & Richard Wohl, R. (1956) · Psychiatry · doi:10.1080/00332747.1956.11023049 · Cited by 4,380 · see citing works
Algorithm aversion: People erroneously avoid algorithms after seeing them err.
Dietvorst, B. J., Simmons, J. P., & Massey, C. (2015) · Journal of Experimental Psychology: General · doi:10.1037/xge0000033 · Cited by 3,232 · see citing works
Dietvorst, B. J., Simmons, J. P., & Massey, C. (2018) · Management Science · doi:10.1287/mnsc.2016.2643 · Cited by 1,230 · see citing works
Bonaccio, S., & Dalal, R. S. (2006) · Organizational Behavior and Human Decision Processes · doi:10.1016/j.obhdp.2006.07.001 · Cited by 1,011 · see citing works
The Perceived Utility of Human and Automated Aids in a Visual Detection Task
Dzindolet, M. T., Pierce, L. G., Beck, H. P., & Dawe, L. A. (2002) · Human Factors: The Journal of the Human Factors and Ergonomics Society · doi:10.1518/0018720024494856 · Cited by 386 · see citing works
Persuasiveness of expert systems
Dijkstra, J. J., Liebrand, W. B. G., & Timminga, E. (1998) · Behaviour & Information Technology · doi:10.1080/014492998119526 · Cited by 161 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Algorithm appreciation: People prefer algorithmic to human judgment
Logg, J. M., Minson, J. A., & Moore, D. A. (2019) · Organizational Behavior and Human Decision Processes · doi:10.1016/j.obhdp.2018.12.005 · Cited by 2,061 · see citing works
A systematic review of algorithm aversion in augmented decision making
Burton, J. W., Stein, M., & Jensen, T. B. (2020) · Journal of Behavioral Decision Making · doi:10.1002/bdm.2155 · Cited by 814 · see citing works
“However, conflicting results exist in which competitive cash rewards did not lead to algorithm aversion (e.g., Prahl & Van Swol, 2017), as well as experiments where algorithm aversion persists in the absence of competitive economic incentivization (e.g., Önkal et al., 2009).”
BackgroundMethodInfluentialForecasting: theory and practice
Petropoulos, F., Apiletti, D., Assimakopoulos, V., Babai, M. Z., Barrow, D. K., Taieb, S. B., Bergmeir, C., Bessa, R., Bijak, J., Boylan, J., et al. (2020) · International Journal of Forecasting · doi:10.1016/j.ijforecast.2021.11.001 · Cited by 630 · see citing works
What influences algorithmic decision-making? A systematic literature review on algorithm aversion
Mahmud, H., Islam, A. K. M. N., Ahmed, S. I., & Smolander, K. (2022) · Technological Forecasting and Social Change · doi:10.1016/j.techfore.2021.121390 · Cited by 601 · see citing works
Algorithmic management in a work context
Jarrahi, M. H., Newlands, G., Lee, M. K., Wolf, C. T., Kinder, E., & Sutherland, W. (2021) · Big Data & Society · doi:10.1177/20539517211020332 · Cited by 532 · see citing works
“Algorithm aversion refers to people’s hesitance to use algorithmic results, usually after observing imperfect performance by algorithms; this can even occur in situations where the algorithm may outperform humans in certain accuracy measures (Dietvorst et al., 2016; Prahl and Van Swol, 2017).”
BackgroundTrust and reliance on AI — An experimental study on the extent and costs of overreliance on AI
Klingbeil, A., Grützner, C., & Schreck, P. (2024) · Computers in Human Behavior · doi:10.1016/j.chb.2024.108352 · Cited by 450 · see citing works
Yam, K. C., Tang, P. M., Jackson, J. C., Su, R., & Gray, K. J. (2022) · Journal of Applied Psychology · doi:10.1037/apl0001045 · Cited by 352 · see citing works
Trusting Automation: Designing for Responsivity and Resilience
Chiou, E. K., & Lee, J. D. (2023) · Human Factors: The Journal of the Human Factors and Ergonomics Society · doi:10.1177/00187208211009995 · Cited by 301 · see citing works
“Algorithm aversion is the opposite trust frame that leads people to overweight poor performance (Dietvorst et al., 2014; Prahl & Van Swol, 2017).”
BackgroundHegner, S. M., Beldad, A. D., & Brunswick, G. J. (2019) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2019.1572353 · Cited by 249 · see citing works
“Research on algorithm aversion shows that people in general trust their own ability more than an algorithmic forecast (Dietvorst, Simmons, & Massey, 2015, 2016; Logg, 2017; Logg, Minson, & Moore, 2018; Prahl & Van Swol, 2017).”
BackgroundKeding, C. (2021) · Management Review Quarterly · doi:10.1007/s11301-020-00181-x · Cited by 231 · see citing works
“…that first sets of empirical work have emerged that show ambiguous effects on management practices with respect to algorithmic biases (D’Acunto et al. 2019; Lambrecht and Tucker 2019), algorithm aversion (Dietvorst et al. 2018; Prahl and Van Swol 2017) and algorithm appreciation (Logg et al. 2019).”
BackgroundInfluentialHumans as Creativity Gatekeepers: Are We Biased Against AI Creativity?
Magni, F., Park, J., & Chao, M. M. (2024) · Journal of Business and Psychology · doi:10.1007/s10869-023-09910-x · Cited by 201 · see citing works
“…the fear and anxiety of being replaced by generative AI (Schmelzer, 2019) and AI’s limitations in terms of autonomy, motivation, and emotions (H.-K. Lee, 2022) often engender a negative assessment of AI-made creative artifacts (Castelo et al., 2019; Longoni et al., 2022; Prahl & Van Swol, 2017).”
When Algorithms Fail: Consumers’ Responses to Brand Harm Crises Caused by Algorithm Errors
Srinivasan, R., & Sarial‐Abi, G. (2021) · Journal of Marketing · doi:10.1177/0022242921997082 · Cited by 199 · see citing works
“…ability to deliver promised benefits to consumers is compromised or, even worse, the brand causes physical harm to consumers (Dutta and Pullig 2011; Pullig, Netemeyer, and Biswas 2006), causing consumers to respond negatively to the brand (Ahluwalia, Burnkrant, and Unnava 2000; Lei, Dawar, and…”
BackgroundThe ethical use of artificial intelligence in human resource management: a decision-making framework
Bankins, S. (2021) · Ethics and Information Technology · doi:10.1007/s10676-021-09619-6 · Cited by 181 · see citing works
“While many individual characteristics are relevant to understanding technology ‘utilization’, I focus here on two human biases that could particularly undermine the ethical use of AI: algorithm aversion and automation bias (Bahner et al., 2008; Prahl & Van Swol, 2017).”
BackgroundRapid Trust Calibration through Interpretable and Uncertainty-Aware AI
Tomsett, R., Preece, A., Braines, D., Cerutti, F., Chakraborty, S., Srivastava, M., Pearson, G., & Kaplan, L. (2020) · Patterns · doi:10.1016/j.patter.2020.100049 · Cited by 186 · see citing works
“64 Other experiments have produced con-flicting results, suggesting that only expert forecasters are susceptible to algorithm aversion while lay users are more likely to trust algorithmic advice.”
BackgroundFrom Man vs. Machine to Man + Machine: The Art and AI of Stock Analyses
CAO, S. S., Jiang, W., Wang, J., & Yang, B. (2021) · National Bureau of Economic Research · doi:10.3386/w28800 · Cited by 43 · see citing works
Watch Me Improve—Algorithm Aversion and Demonstrating the Ability to Learn
Berger, B., Adam, M., Rühr, A., & Benlian, A. (2020) · Business & Information Systems Engineering · doi:10.1007/s12599-020-00678-5 · Cited by 184 · see citing works
Why do users trust algorithms? A review and conceptualization of initial trust and trust over time
Cabiddu, F., Moi, L., Patriotta, G., & Allen, D. G. (2022) · European Management Journal · doi:10.1016/j.emj.2022.06.001 · Cited by 138 · see citing works
“AI-based algorithms are increasingly being deployed in contemporary work settings to support managerial and organizational decisions (Burton et al., 2020; Glikson & Woolley, 2020; Kellogg et al., 2019; Lepri et al., 2017; Prahl & Van Swol, 2017).”
Paluch, S., Tuzovic, S., Holz, H. F., Kies, A., & Jörling, M. (2021) · Journal of service management · doi:10.1108/josm-11-2020-0406 · Cited by 195 · see citing works
Inkpen, K., Chappidi, S., Mallari, K., Nushi, B., Ramesh, D., Michelucci, P., Mandava, V., Vepřek, L. H., & Quinn, G. (2023) · ACM Transactions on Computer-Human Interaction · doi:10.1145/3534561 · Cited by 132 · see citing works
“People want to understand where a recommendation is coming from, and ind that they have more in common with human-based recommendations as opposed to algorithms [12, 47, 60].”
BackgroundArıkan, E., Altinigne, N., Kuzgun, E., & Okan, M. (2023) · Journal of Retailing and Consumer Services · doi:10.1016/j.jretconser.2022.103175 · Cited by 120 · see citing works
Algorithmic Transference: People Overgeneralize Failures of AI in the Government
Longoni, C., Cian, L., & Kyung, E. J. (2023) · Journal of Marketing Research · doi:10.1177/00222437221110139 · Cited by 121 · see citing works
“One dimension of consumer response has assessed reliance on an algorithmic versus a human advisor before and after learning of the advisor failure rate (Berger et al. 2021; Dietvorst, Simmons, and Massey 2015; Prahl and Van Swol 2017; Srinivasan and Sarial-Abi 2021).”
BackgroundHou, Y. T.-Y., & Jung, M. F. (2021) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3479864 · Cited by 108 · see citing works
“Van Swol (2017) [24] Steven, a person experienced in operating room management issues.”
BackgroundChugunova, M., & Sele, D. (2022) · Journal of Behavioral and Experimental Economics · doi:10.1016/j.socec.2022.101897 · Cited by 118 · see citing works
Jain, R. M., Garg, N., & Khera, S. N. (2022) · Frontiers in Psychology · doi:10.3389/fpsyg.2022.893691 · Cited by 155 · see citing works
“Early literature on algorithmic support suggests that people avoided inputs from algorithms even when they had information about superior performance (Dietvorst et al., 2015; Prahl and Van Swol, 2017).”
BackgroundLeesakul, N., Oostveen, A., Eimontaite, I., Wilson, M. L., & Hyde, R. (2022) · Sustainability · doi:10.3390/su14063311 · Cited by 131 · see citing works
“When researchers tested individuals’ tolerance for mistakes made by artificial intelligence software, they found that people demand a much higher success rate from robots than from humans [68].”
BackgroundLourenco, C. J., Dellaert, B. G. C., & Donkers, B. (2020) · Journal of Interactive Marketing · doi:10.1016/j.intmar.2019.10.003 · Cited by 96 · see citing works
Reducing algorithm aversion through experience
Filiz, I., Judek, J. R., Lorenz, M., & Spiwoks, M. (2021) · Journal of Behavioral and Experimental Finance · doi:10.1016/j.jbef.2021.100524 · Cited by 88 · see citing works
The Effects of Explanations in Automated Essay Scoring Systems on Student Trust and Motivation
Conijn, R., Kahr, P. K., & Snijders, C. C. P. (2023) · Journal of Learning Analytics · doi:10.18608/jla.2023.7801 · Cited by 100 · see citing works
“However, even in cases where systems show high accuracy, we see that humans tend to have low trust (Prahl & van Swol, 2017).”
BackgroundVirtual vs. human influencer: Effects on users’ perceptions and brand outcomes
Cicco, R. D., Iacobucci, S., Cannito, L., Onesti, G., Ceccato, I., & Palumbo, R. (2024) · Technology in Society · doi:10.1016/j.techsoc.2024.102488 · Cited by 96 · see citing works
Mahmud, H., Islam, A. K. M. N., & Mitra, R. K. (2023) · Technological Forecasting and Social Change · doi:10.1016/j.techfore.2023.122641 · Cited by 98 · see citing works
Effective human–AI work design for collaborative decision-making
Jain, R., Garg, N., & Khera, S. N. (2023) · Kybernetes · doi:10.1108/k-04-2022-0548 · Cited by 85 · see citing works
Predictive maintenance for industry 5.0: behavioural inquiries from a work system perspective
van Oudenhoven, B., de Calseyde, P. P. F. M. V., Basten, R., & Demerouti, E. (2022) · International Journal of Production Research · doi:10.1080/00207543.2022.2154403 · Cited by 95 · see citing works
“However, humans falsely expect that systems are perfectly accurate (Prahl and Van Swol 2017; Dzindolet et al. 2002), a criterion that is absent for advice given by humans (Alvarado-Valencia and Bar-rero 2014).”
BackgroundBiased Humans, (Un)Biased Algorithms?
Pethig, F., & Kroenung, J. (2022) · Journal of Business Ethics · doi:10.1007/s10551-022-05071-8 · Cited by 97 · see citing works
“This effect is more pronounced (a) after witnessing algorithm failure (Dietvorst et al., 2015; Prahl & Van Swol, 2017), (b) for subjective than objective tasks (Castelo et al., 2019), (c) in moral domains (Bigman & Gray, 2018), and (d) among experts than non-experts (Logg et al., 2019).”
BackgroundHow do citizens perceive the use of Artificial Intelligence in public sector decisions?
Haesevoets, T., Verschuere, B., Van Severen, R., & Roets, A. (2024) · Government Information Quarterly · doi:10.1016/j.giq.2023.101906 · Cited by 74 · see citing works
Job candidates’ reactions to AI-enabled job application processes
van Esch, P., Black, J. S., & Arli, D. (2021) · AI and Ethics · doi:10.1007/s43681-020-00025-0 · Cited by 72 · see citing works
“Prahl and Van Swol [72] did not find people preferred human advice over automated advice, but they did find that people utilized automated advice significantly less than human advice after the advice was demonstrated to be bad.”
BackgroundA meta-analysis of the weight of advice in decision-making
Bailey, P. E., Leon, T., Ebner, N. C., Moustafa, A. A., & Weidemann, G. (2023) · Current Psychology · doi:10.1007/s12144-022-03573-2 · Cited by 82 · see citing works
Liu, P., Du, Y., & Xu, Z. (2019) · Accident Analysis & Prevention · doi:10.1016/j.aap.2019.02.012 · Cited by 65 · see citing works
The Next Generation of Medical Decision Support: A Roadmap Toward Transparent Expert Companions
Bruckert, S., Finzel, B., & Schmid, U. (2020) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2020.507973 · Cited by 66 · see citing works
“Explanations can provide a valuable basis for providing transparency and comprehensibility regarding systems’ decisions and therefore can lead to increased trust of ML users (Pu et al., 2011; Prahl and Swol, 2017; Miller, 2019).”
BackgroundValues and Ethics in Information Systems
Spiekermann, S., Krasnova, H., Hinz, O. F., Baumann, A., Benlian, A., Gimpel, H., Heimbach, I., Köster, A., Maedche, A., Niehaves, B., et al. (2022) · Business & Information Systems Engineering · doi:10.1007/s12599-021-00734-8 · Cited by 74 · see citing works
“These mistakes, however, may severely diminish trust towards automation, leading to changes in individual attitudes and perceptions in the short and long term (e.g., Dietvorst et al. 2015; Prahl and Van Swol 2017).”
BackgroundDo Humans Trust Advice More if it Comes from AI?: An Analysis of Human-AI Interactions
Vodrahalli, K., Daneshjou, R., Gerstenberg, T., & Zou, J. (2022) · Proceedings of the 2022 AAAI/ACM Conference on AI, Ethics, and Society · doi:10.1145/3514094.3534150 · Cited by 63 · see citing works
“Previous work has attempted to understand the effects of peer vs. expert advice [18], human vs. algorithmic advice [6, 18, 21, 22], task difficulty [10], and advice confidence [26, 27] on human utilization of advice.”
BackgroundMethodInfluentialWhen Algorithms Err : Differential Impact of Early vs. Late Errors on Users’ Reliance on Algorithms
Kim, A., Yang, M., & Zhang, J. (2020) · ACM Transactions on Computer-Human Interaction · doi:10.1145/3557889 · Cited by 40 · see citing works
“In one set of studies, researchers have found that people tend to be less tolerant of errors made by algorithms than those by humans, and easily avoid using algorithms in decision-making after seeing the algorithms make mistakes [15, 25, 46].”
BackgroundMirror, Mirror on the Wall: Algorithmic Assessments, Transparency, and Self-Fulfilling Prophecies
Bauer, K., & Gill, A. (2023) · Information Systems Research · doi:10.1287/isre.2023.1217 · Cited by 66 · see citing works
“Whether human decision makers underuse algorithmic advice depends on their task experience (Prahl and Van Swol 2017), unmet expectations about the algorithm’s characteristics (Castelo et al. 2019), the perceived loss of decision autonomy (Scherer et al. 2015), and the presence of a…”
BackgroundChoung, H., Seberger, J. S., & David, P. (2023) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2023.2266244 · Cited by 67 · see citing works
“These studies indicate that people tend to discount algorithmic decisions relative to human decisions, particularly when an algorithm is observed to have made an error (Dietvorst et al., 2015; Prahl & Van Swol, 2017).”
Wang, G., Guo, Y., Zhang, W., Xie, S., & Chen, Q. (2023) · Government Information Quarterly · doi:10.1016/j.giq.2023.101803 · Cited by 71 · see citing works
To err is human, not algorithmic – Robust reactions to erring algorithms
Renier, L. A., Mast, M. S., & Bekbergenova, A. (2021) · Computers in Human Behavior · doi:10.1016/j.chb.2021.106879 · Cited by 62 · see citing works
Shaikh, S. J., & Cruz, I. (2022) · AI & Society · doi:10.1007/s00146-021-01335-5 · Cited by 34 · see citing works
“…(e.g., advisor credibility and trust), algorithm-related (e.g., error-prone technology and lack of explainable AI), and/or task-related (e.g., logic problems) variables (Dietvorst et al. 2015; Dijkstra 2010; Prahl and Van Swol; 2017; Logg et al. 2019; Yeomans et al. 2019) variables.”
BackgroundUncalibrated Models Can Improve Human-AI Collaboration
Vodrahalli, K., Gerstenberg, T., & Zou, J. (2022) · neural information processing systems · doi:10.52202/068431-0289 · Cited by 19 · see citing works
“Previous work has studied the effect of peer vs. expert advice [15], automated vs. human advice [22, 5, 15, 21], task difficulty [7], and advice confidence [25, 26].”
BackgroundMethodImpact of Algorithmic Decision Making on Human Behavior: Evidence from Ultimatum Bargaining
Erlei, A., Nekdem, F., Meub, L., Anand, A., & Gadiraju, U. (2020) · Proceedings of the AAAI Conference on Human Computation and Crowdsourcing · doi:10.1609/hcomp.v8i1.7462 · Cited by 50 · see citing works
Artificial intelligence in talent acquisition: exploring organisational and operational dimensions
Paramita, D., Okwir, S., & Nuur, C. (2024) · International journal of organizational analysis · doi:10.1108/ijoa-09-2023-3992 · Cited by 99 · see citing works
Using judgment to select and adjust forecasts from statistical models
Baets, S. D., & Harvey, N. (2020) · European Journal of Operational Research · doi:10.1016/j.ejor.2020.01.028 · Cited by 54 · see citing works
“Prahl and Van Swol (2017) work supports this: they found that people’s use of advice dropped significantly more after they received bad advice from an algorithm than bad advice from a human judge.”
BackgroundThe extent of algorithm aversion in decision-making situations with varying gravity
Filiz, I., Judek, J. R., Lorenz, M., & Spiwoks, M. (2023) · PLoS ONE · doi:10.1371/journal.pone.0278751 · Cited by 52 · see citing works
“Prahl and Van Swol [28] provide feedback after every individual decision, enabling an assessment of the success of the algo-rithm.”
BackgroundInfluentialWhen Will Workers Follow an Algorithm?: A Field Experiment with a Retail BusinessPreprint
Kawaguchi, K. (2019) · SSRN Electronic Journal · doi:10.2139/ssrn.3157132 · Cited by 47 · see citing works
Cabitza, F., Campagner, A., Angius, R., Natali, C., & Reverberi, C. (2023) · Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3544548.3581095 · Cited by 51 · see citing works
“measure advice utilization [73], is proposed by Prahl and Van Swol in [79].”
BackgroundEffect of Explanation Conceptualisations on Trust in AI-assisted Credibility Assessment
Pareek, S., van Berkel, N., Velloso, E., & Goncalves, J. (2024) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3686922 · Cited by 42 · see citing works
“On the other hand, undue scepticism or distrust towards automated systems, known as algorithm aversion [52, 87], may lead users to solely rely on their own judgements, disregarding automated advice.”
BackgroundPrahl, A., & Swol, L. M. V. (2021) · Human-Machine Communication · doi:10.30658/hmc.2.11 · Cited by 44 · see citing works
Self-citationOvercoming Algorithm Aversion: A Comparison between Process and Outcome Control
Cheng, L., & Chouldechova, A. (2023) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3544548.3581253 · Cited by 24 · see citing works
“The degree of aversion has been found to depend on various factors including the characteristics of the tasks [9], the decision makers [42], and the interactions between humans and algorithms [50].”
BackgroundMachine Thinking, Fast and Slow
Bonnefon, J., & Rahwan, I. (2020) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2020.09.007 · Cited by 45 · see citing works
“It is not at all clear that the admission of past mistakes might increase trust in the machine – insofar as we know that people lose trust in machines that make mistakes faster than they lose trust in humans who make mistakes [38,39].”
BackgroundStability in the inefficient use of forecasting systems: A case study in a supply chain company
Fildes, R., & Goodwin, P. (2020) · International Journal of Forecasting · doi:10.1016/j.ijforecast.2020.11.004 · Cited by 36 · see citing works
Kämmer, J. E., Choshen-Hillel, S., Müller-Trede, J., Black, S. L., & Weibler, J. (2023) · Decision · doi:10.1037/dec0000199 · Cited by 35 · see citing works
“The work on algorithmic advice (Logg etal., 2019; Prahl & Van Swol, 2017), coaching (de Haan, 2019), and mentoring (Son & Kim, 2018) that we have previously referred to is among these exclusions.”
Advancing Dermatological Diagnostics: Interpretable AI for Enhanced Skin Lesion Classification
Metta, C., Beretta, A., Guidotti, R., Yin, Y., Gallinari, P., Rinzivillo, S., & Giannotti, F. (2024) · Diagnostics · doi:10.3390/diagnostics14070753 · Cited by 34 · see citing works
“This observation aligns with prior studies in the field of algorithm aversion [50,51,53].”
MethodResultFinkel, M., & Krämer, N. (2022) · International Journal of Social Robotics · doi:10.1007/s12369-022-00879-w · Cited by 33 · see citing works
“At first glance, algorithm appreciation seems opposed to the concept of algorithm aversion, but while algorithm appreciation refers to people’s presumed superiority of algorithms, algorithm aversion occurs when people are witnessing an algorithm failing at giving a correct output [7,31].”
BackgroundAlgorithm aversion? On the influence of advice accuracy on trust in algorithmic advice
Daschner, S., & Obermaier, R. (2022) · Journal of Decision System · doi:10.1080/12460125.2022.2070951 · Cited by 34 · see citing works
“Empirical evidence comes from Prahl and Van Swol (2017) who documented algorithm aversion after providing bad advice.”
BackgroundMethodResultInfluentialSchaap, G., Bosse, T., & Vettehen, P. H. (2023) · AI & Society · doi:10.1007/s00146-023-01649-6 · Cited by 37 · see citing works
“…of humans and algorithms, only finding a decreased use of algorithms when the algorithm makes mistakes or gives bad advice (cf. Castelo et al. 2019; Dietvorst et al. 2015; Prahl and Van Swol 2017), or when the human decision maker is suddenly replaced by an algorithm (Prahl and Van Swol 2021).”
InfluentialAlgorithm Aversion as an Obstacle in the Establishment of Robo Advisors
Filiz, I., Judek, J. R., Lorenz, M., & Spiwoks, M. (2022) · Journal of Risk and Financial Management · doi:10.3390/jrfm15080353 · Cited by 31 · see citing works
When Self-Humanization Leads to Algorithm Aversion
Heßler, P. O., Pfeiffer, J., & Hafenbrädl, S. (2022) · Business & Information Systems Engineering · doi:10.1007/s12599-022-00754-y · Cited by 33 · see citing works
“Of course, predicting the future perfectly is inherently difficult; thus, even extremely well-crafted algorithms will err from time to time (Diet-vorst et al. 2015; Prahl and van Swol 2017).”
BackgroundInfluentialUnderstanding Choice Independence and Error Types in Human-AI Collaboration
Erlei, A., Sharma, A., & Gadiraju, U. (2024) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613904.3641946 · Cited by 26 · see citing works
“This remarkably resilient pattern has been replicated in many contexts [16, 21, 22, 29, 31, 51, 59, 83, 85, 89, 90], although humans have also exhibited preferences for algorithms in task domains that are perceived as "objective" [21, 59, 69].”
BackgroundLe, H., Shen, Y., Li, Z., Xia, M., Tang, L., Li, X., Jia, J., Wang, Q., Gašević, D., & Fan, Y. (2025) · British Journal of Educational Technology · doi:10.1111/bjet.13614 · Cited by 38 · see citing works
Cao, B., Li, Z., & Jiang, L. (2024) · Telematics and Informatics · doi:10.1016/j.tele.2024.102189 · Cited by 32 · see citing works
Task-specific algorithm advice acceptance: A review and directions for future research
Kaufmann, E., Chacon, A., Kausel, E. E., Herrera, N., & Reyes, T. H. (2023) · Data and Information Management · doi:10.1016/j.dim.2023.100040 · Cited by 28 · see citing works
“Computer says no”: Algorithmic decision support and organisational responsibility
Adensamer, A., Gsenger, R., & Klausner, L. D. (2021) · Journal of Responsible Technology · doi:10.1016/j.jrt.2021.100014 · Cited by 28 · see citing works
Factors influencing trust in algorithmic decision-making: an indirect scenario-based experiment
Marmolejo‐Ramos, F., Marrone, R. L., Korolkiewicz, M. W., Gabriel, F. C., Siemens, G., Joksimovíc, S., Yamada, Y., Mori, Y., Rahwan, T., Sahakyan, M., et al. (2025) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2024.1465605 · Cited by 41 · see citing works
Advice from artificial intelligence: a review and practical implications
Baines, J. I., Dalal, R. S., Ponce, L. P., & Tsai, H.-C. (2024) · Frontiers in Psychology · doi:10.3389/fpsyg.2024.1390182 · Cited by 28 · see citing works
“Research has also shown that decision-makers place more weight on AI errors than human errors (Dietvorst et al., 2015; Prahl and Van Swol, 2017; Gaube et al., 2023).”
BackgroundPublic preferences for governing AI technology: Comparative evidence
Ehret, S. (2022) · Journal of European Public Policy · doi:10.1080/13501763.2022.2094988 · Cited by 28 · see citing works
“There is a general tendency of individuals to distrust algorithms (Burton et al., 2020; Dietvorst et al., 2015; Logg et al., 2019), conditional on the expectations placed on algorithms in their domains (Castelo et al., 2019; Logg et al.,2019; Prahl & Van Swol, 2017).”
BackgroundTursunbayeva, A., Fernández, V., Gallardo-Gallardo, E., & Moschera, L. (2025) · European Management Journal · doi:10.1016/j.emj.2025.03.002 · Cited by 34 · see citing works
Snijders, C. C. P., Conijn, R., de Fouw, E., & van Berlo, K. (2022) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2022.2097601 · Cited by 27 · see citing works
Saragih, M., & Morrison, B. W. (2021) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2021.1990518 · Cited by 25 · see citing works
“Meanwhile, humans are expected to be imperfect and are therefore forgiven more readily for committing errors (Prahl & Van Swol, 2017).”
BackgroundResultInfluentialvan der Zander, Q. E. W., Roumans, R., Kusters, C. H. J., Dehghani, N., Masclee, A. A. M., de With, P. H. N., van der Sommen, F., Snijders, C. C. P., & Schoon, E. J. (2024) · Gastrointestinal Endoscopy · doi:10.1016/j.gie.2024.06.029 · Cited by 27 · see citing works
Myers, S., & Everett, J. A. C. (2024) · Cognition · doi:10.1016/j.cognition.2024.106028 · Cited by 31 · see citing works
“Such results build on previous work on algorithm aversion in the non-moral domain (e.g. Dietvorst et al., 2015; ¨ et al., 2009; Prahl & Van Swol, 2017; Promberger & Baron, 2006) and evidence of people being averse to machines making indirectly morally relevant decisions in transport, parole and…”
Liu, W., & Wang, Y. (2024) · International Journal of Human-Computer Interaction · doi:10.1080/10447318.2024.2313921 · Cited by 27 · see citing works
“Our results support the proposition that people generally prefer human input over algorithms in the area of product recommendations, as shown in various studies (Dawes et al., 1989; Kleinmuntz, 1990; Kleinmuntz & Schkade, 1993; Prahl & Van Swol, 2017; Morewedge, 2022).”
Advice Utilization From Predictive Analytics Tools: The Trend is Your Friend
Fehrenbacher, D., Ghio, A., & Weisner, M. (2022) · The European Accounting Review · doi:10.1080/09638180.2022.2138934 · Cited by 26 · see citing works
“…both, circumstances under which decision-makers prefer algorithmic advice over human advice (i.e., algorithm appreciation – see Logg et al., 2019), and circumstances where the opposite is true (i.e., algorithm aversion – e.g., Chen et al., 2022; Dietvorst et al., 2018; Prahl & Van Swol, 2017).”
BackgroundA longitudinal approach for understanding algorithm use
Chacon, A., Kausel, E. E., & Reyes, T. H. (2022) · Journal of Behavioral Decision Making · doi:10.1002/bdm.2275 · Cited by 28 · see citing works
Technology-Augmented Choice: How Digital Innovations Are Transforming Consumer Decision Processes
Melumad, S., Hadi, R., Hildebrand, C., & Ward, A. F. (2020) · Customer Needs and Solutions · doi:10.1007/s40547-020-00107-4 · Cited by 36 · see citing works
Jung, M., & Seiter, M. (2021) · Journal of Management Control · doi:10.1007/s00187-021-00326-3 · Cited by 25 · see citing works
Downen, T., Kim, S., & Lee, L. S. (2023) · International Journal of Accounting Information Systems · doi:10.1016/j.accinf.2023.100664 · Cited by 30 · see citing works
The AI‐augmented crowd: How human crowdvoters adopt AI (or not)
Freisinger, E., Unfried, M., & Schneider, S. (2023) · Journal of Product Innovation Management · doi:10.1111/jpim.12708 · Cited by 30 · see citing works
Hybrid forecasting of geopolitical events†
Benjamin, D. M., Morstatter, F., Abbas, A. E., Abeliuk, A., Atanasov, P., Bennett, S. T., Beger, A., Birari, S., Budescu, D., Catasta, M., et al. (2023) · The AI Magazine · doi:10.1002/aaai.12085 · Cited by 24 · see citing works
Human preferences toward algorithmic advice in a word association task
Bogert, E., Lauharatanahirun, N., & Schecter, A. (2022) · Scientific Reports · doi:10.1038/s41598-022-18638-2 · Cited by 23 · see citing works
“Conversely, aversion is typically strongest in subjective or open-ended tasks 10,11,14,39 .”
BackgroundLaw and tech collide: foreseeability, reasonableness and advanced driver assistance systems
Leiman, T. (2020) · Policy and Society · doi:10.1080/14494035.2020.1787696 · Cited by 23 · see citing works
“(Prahl & Van Swol, 2017, p. 693) Could this mean that unless all ADAS features were shown to be 100% safe 100% of the time, reliance on those features would be perceived by a reasonable person as an unacceptable risk.”
BackgroundWeber, R. O., Johs, A. J., Goel, P., & Marques-Silva, J. P. (2024) · AI Magazine · doi:10.1002/aaai.12184 · Cited by 16 · see citing works
Liao, W., Oh, Y. J., Feng, B., & Zhang, J. (2023) · Communication Research · doi:10.1177/00936502221138427 · Cited by 25 · see citing works
“…has touched effects of artificial agents’ identity (e.g., C. Edwards et al., 2016; A. Edwards et al., 2019; Dietvorst et al., 2015; Ho et al., 2018; Prahl & Swol, 2017, 2017; Spence et al., 2014; also see Fox et al., 2015) but have not yet attended fully to the social categorization and…”
BackgroundInfluentialThe effects of generative AI’s human-like competencies on clinical decision-making
Spatscheck, N., Schaschek, M., & Winkelmann, A. (2024) · Journal of Decision System · doi:10.1080/12460125.2024.2430731 · Cited by 27 · see citing works
Bursell, M., & Roumbanis, L. (2024) · Big Data & Society · doi:10.1177/20539517231221758 · Cited by 19 · see citing works
“Alternatively, a common response among people who have witnessed or heard about algorithms making mistakes or delivering low quality output is that they develop algo-rithm aversion (Dietvorst et al., 2015; Prahl and Van Swol, 2017; Yeomans et al., 2019).”
Mendel, T., Nov, O., & Wiesenfeld, B. M. (2024) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3686925 · Cited by 17 · see citing works
“Trust in AI is more fragile than trust in humans [84].”
Grgic-Hlaca, N., Castelluccia, C., & Gummadi, K. P. (2022) · AAAI Conference on Human Computation & Crowdsourcing · doi:10.48550/arxiv.2209.03821 · Cited by 20 · see citing works
Fenneman, A., Sickmann, J., Pitz, T., & Sanfey, A. G. (2021) · PLoS ONE · doi:10.1371/journal.pone.0247084 · Cited by 27 · see citing works
“Previous research suggests that experimental participants typically describe themselves as having more in common with human rather than algorithmic advisors [4] and report a higher degree of trust in algorithms with more human-like features [21–23].”
BackgroundThe Trust Recovery Journey. The Effect of Timing of Errors on the Willingness to Follow AI Advice.
Kahr, P. K., Rooks, G., Snijders, C., & Willemsen, M. C. (2024) · Proceedings of the 29th International Conference on Intelligent User Interfaces · doi:10.1145/3640543.3645167 · Cited by 21 · see citing works
“Just like people were found to value losses more than comparable gains [31], algorithmic failure seems to weigh heavier than algorithmic correctness [53] and slows down the recovery of trust [43].”
BackgroundRoumbanis, L. (2025) · Futures · doi:10.1016/j.futures.2025.103538 · Cited by 22 · see citing works
After opening the black box: Meta-dehumanization matters in algorithm recommendation aversion
Chen, G., Dang, J., & Liu, L. (2024) · Computers in Human Behavior · doi:10.1016/j.chb.2024.108411 · Cited by 23 · see citing works
Beyond Recommendations: From Backward to Forward AI Support of Pilots' Decision-Making Process
Zhang, Z. T., Feger, S. S., Dullenkopf, L., Liao, R., Süsslin, L. E., Liu, Y., & Butz, A. M. (2024) · Proceedings of the ACM on Human-Computer Interaction · doi:10.1145/3687024 · Cited by 16 · see citing works
Maasland, C., & Weißmüller, K. S. (2022) · Frontiers in Psychology · doi:10.3389/fpsyg.2022.779028 · Cited by 19 · see citing works
“This is in line with prior research on human–human vis-à-vis human–machine trust (Madhavan and Wiegmann, 2007; Prahl and van Swol, 2017, 2021; Filiz et al., 2021).”
Influential
- Authors
- Prahl, A., & Van Swol, L.
- Affiliations
- University of Wisconsin–Madison (US)
- Department of Communication Arts University of Wisconsin–Madison Wisconsin USA
- Corresponding author
- Prahl, Andrew
- ORCID
- Prahl, Andrew
- Venue
- Journal of Forecasting
- Publisher
- Wiley
- Published
- 2017-03-24
- Volume
- 36
- Issue
- 6
- Pages
- 691-702
- ISSN
- 0277-6693 (ISSN-L)
- 1099-131X
- OpenAlex
- W2600700815
- Cited by
- 308
- References
- 0
- Citations by year
- 2026: 45 · 2025: 56 · 2024: 59 · 2023: 39 · 2022: 47 · 2021: 31 · 2020: 19 · 2019: 10 · 2018: 2
- FWCI
- 12.17
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Journal of Forecasting (publishedVersion)
- Topics
- Social Sciences › Psychology › Social Psychology › Human-Automation Interaction and Safety
- Social Sciences › Decision Sciences › Management Science and Operations Research › Forecasting Techniques and Applications
- Social Sciences › Decision Sciences › General Decision Sciences › Decision-Making and Behavioral Economics
- Keywords
- algorithm aversion, trust in automation, decision maker attitude
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
