MétaCan
Menu
Back to cohort
Record W4404620139 · doi:10.1111/deci.12658

Explanation seeking and anomalous recommendation adherence in human‐to‐human versus human‐to‐artificial intelligence interactions

2024· article· en· W4404620139 on OpenAlex

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

VenueDecision Sciences · 2024
Typearticle
Languageen
FieldDecision Sciences
TopicDecision-Making and Behavioral Economics
Canadian institutionsSaint Mary's UniversityQueen's University
Fundersnot available
KeywordsLoss aversionSet (abstract data type)Sunk costsPreferenceComputer scienceRecommender systemRisk aversion (psychology)Bayesian inferenceHuman intelligencePsychologyArtificial intelligenceBayesian probabilityMachine learningMicroeconomicsEconomicsExpected utility hypothesis

Abstract

fetched live from OpenAlex

Abstract The use of artificial intelligence (AI) in operational decision‐making is growing, but individuals can display algorithm aversion, preventing adherence to AI system recommendations—even when the system outperforms human decision‐makers. Understanding why such algorithm aversion occurs and how to reduce it is important to ensure AI is fully leveraged. While the ability to seek an explanation from an AI may be a promising approach to mitigate this aversion, there is conflicting evidence on their benefits. Based on several behavioral theories, including Bayesian choice, loss aversion, and sunk cost avoidance, we hypothesize that if a recommendation is perceived as an anomalous loss, it will decrease recommendation adherence; however, the effect will be mediated by explanations and differ depending on whether the advisor providing the recommendation and explanation is a human or an AI. We conducted a survey‐based lab experiment set in the online rental market space and found that presenting a recommendation as a loss anomaly significantly reduces adherence compared to presenting it as a gain, however, this negative effect can be dampened if the advisor is an AI. We find explanation‐seeking has a limited impact on adherence, even after considering the influence of the advisor; we discuss the managerial and theoretical implications of these findings.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.006
metaresearch head score (Gemma)0.003
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesScholarly communication, Insufficient payload (model declined to judge)
Consensus categoriesInsufficient payload (model declined to judge)
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Other design · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.981
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0060.003
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0030.004
Science and technology studies0.0010.000
Scholarly communication0.0030.002
Open science0.0010.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0010.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.428
GPT teacher head0.537
Teacher spread0.109 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it