CoChat

Toward Quantifying Trust Dynamics: How People Adjust Their Trust After Moment-to-Moment Interaction With Automation

Open access
Yang, X. Jessie, Schemanske, Christopher, Searle, Christine · 2023 · Human Factors: The Journal of the Human Factors and Ergonomics Society

The DOI matches its Crossref record; 47 of the 48 references it cites were found in a registry or an index; none of the 43 sources its claims cite is checked for support yet.

Abstract

OBJECTIVE: We examine how human operators adjust their trust in automation as a result of their moment-to-moment interaction with automation. BACKGROUND: Most existing studies measured trust by administering questionnaires at the end of an experiment. Only a limited number of studies viewed trust as a dynamic variable that can strengthen or decay over time. METHOD: Seventy-five participants took part in an aided memory recognition task. In the task, participants viewed a series of images and later on performed 40 trials of the recognition task to identify a target image when it was presented with a distractor. In each trial, participants performed the initial recognition by themselves, received a recommendation from an automated decision aid, and performed the final recognition. After each trial, participants reported their trust on a visual analog scale. RESULTS: Outcome bias and contrast effect significantly influence human operators' trust adjustments. An automation failure leads to a larger trust decrement if the final outcome is undesirable, and a marginally larger trust decrement if the human operator succeeds the task by him/herself. An automation success engenders a greater trust increment if the human operator fails the task. Additionally, automation failures have a larger effect on trust adjustment than automation successes. CONCLUSION: Human operators adjust their trust in automation as a result of their moment-to-moment interaction with automation. Their trust adjustments are significantly influenced by decision-making heuristics/biases. APPLICATION: Understanding the trust adjustment process enables accurate prediction of the operators' moment-to-moment trust in automation and informs the design of trust-aware adaptive automation.

Overview

CoChat Summary

## TL;DR

This experiment tracked 75 adults’ trust in an automated decision aid trial-by-trial during a 40-trial aided memory recognition task. Trust rose after automation successes and fell after failures, but the size of the adjustment was also shaped by outcome bias and contrast effect.

## Key findings

  • In one-sample t-tests with Bonferroni adjustment, correct recommendations increased trust for all reported success patterns: Pattern 2, *t*(1,71)=7.81, *p*<.001, *d*=0.92; Pattern 3, *t*(1,70)=12.6, *p*<.001, *d*=1.48; Pattern 7, *t*(1,72)=9.94, *p*<.001, *d*=1.17.
  • Wrong recommendations decreased trust for all reported failure patterns: Pattern 0, *t*(1,67)=−8.29, *p*<.001, *d*=1.00; Pattern 4, *t*(1,62)=−8.72, *p*<.001, *d*=1.10; Pattern 5, *t*(1,62)=−6.74, *p*<.001, *d*=0.85.
  • Automation failures had a larger impact on trust than automation successes: Pattern 0 vs. 2, *t*(1,65)=5.34, *p*<.001, *d*=0.66; Pattern 5 vs. 7, *t*(1,60)=4.81, *p*<.001, *d*=0.62.
  • Heuristic effects were observed: an automation failure produced a larger trust decrement when the final outcome was undesirable, *t*(1,52)=−2.63, *p*=.01, *d*=0.36; and an automation success produced a greater trust increment when the human operator initially failed the task, *t*(1,69)=4.40, *p*<.001, *d*=0.53.

## Methods

  • Mixed design with 75 adults (mean age 23.4 years, SD 4.1); between-subjects automation reliability had 3 levels (70%, 80%, 90%), and within-subjects performance pattern had 8 possible patterns across 40 two-alternative forced-choice trials.
  • Participants completed a simulated memory recognition task: viewed 150 images, did an interpolated memory task, then completed 40 trials where they made an initial recognition, received an automated recommendation, made a final recognition, and rated trust after each trial on a visual analog scale.
  • Main analyses included planned comparisons, ANCOVAs, one-sample and paired-samples t-tests with Bonferroni adjustments, and linear regressions; trust propensity was tested as a covariate and then removed when nonsignificant.

## Limitations

  • The paper states that it used a one-item trust scale, which may not capture all subdimensions of trust as well as multidimensional measures; the authors suggest developing a succinct multi-scale trust measure for repeated interactions.
  • Patterns 1 and 6 were excluded from data analysis because they occurred too rarely; the paper says further research should purposely induce these patterns.
  • Evident from the text: the task was “fairly simple,” and the authors note that the outcome-bias finding should be explored further in more cognitively demanding tasks such as medical decision making.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1177/00187208211034716

Cites

47 of 47 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 The static snapshot view of trust versus the dynamic view of trust. If taking a snapshot at time t, both agents have the same trust level, but their trust dynamics differ.
Figure 1 The static snapshot view of trust versus the dynamic view of trust. If taking a snapshot at time t, both agents have the same trust level, but their trust dynamics differ.CC BY-NC · PubMed Central
Figure 2 Flowchart of the aided memory recognition task.
Figure 2 Flowchart of the aided memory recognition task.CC BY-NC · PubMed Central
Figure 3 Illustration of target pictures a, b, c, ….
Figure 3 Illustration of target pictures a, b, c, ….CC BY-NC · PubMed Central
Figure 4 Illustration of target–distractor pairs.
Figure 4 Illustration of target–distractor pairs.CC BY-NC · PubMed Central
Figure 5 Illustration of the two-alternative forced-choice (2AFC) test.
Figure 5 Illustration of the two-alternative forced-choice (2AFC) test.CC BY-NC · PubMed Central
Figure 6 The automated aid’s recommendation interface.
Figure 6 The automated aid’s recommendation interface.CC BY-NC · PubMed Central
Figure 7 Illustration of the visual analog scale for measuring trusti after each trial.
Figure 7 Illustration of the visual analog scale for measuring trusti after each trial.CC BY-NC · PubMed Central
Figure 8 Mean and standard error (SE) values of trust increment in Patterns 2, 3, and 7 (***p < .001).
Figure 8 Mean and standard error (SE) values of trust increment in Patterns 2, 3, and 7 (***p < .001).CC BY-NC · PubMed Central
Figure 9 Mean and standard error (SE) values of trust decrement in Patterns 0, 4, and 5 (***p < .001).
Figure 9 Mean and standard error (SE) values of trust decrement in Patterns 0, 4, and 5 (***p < .001).CC BY-NC · PubMed Central
Figure 10 Comparing the magnitude of trust adjustment between (a) Patterns 0 and 2 and between (b) Patterns 5 and 7 (***p < .001).
Figure 10 Comparing the magnitude of trust adjustment between (a) Patterns 0 and 2 and between (b) Patterns 5 and 7 (***p < .001).CC BY-NC · PubMed Central
Figure 11 (a) Comparing trust increment between Patterns 2 and 3. (b) Comparing trust decrement between Patterns 4 and 5 (**p < .01).
Figure 11 (a) Comparing trust increment between Patterns 2 and 3. (b) Comparing trust decrement between Patterns 4 and 5 (**p < .01).CC BY-NC · PubMed Central
Figure 12 (a) Comparing trust increment between Patterns 3 and 7 (***p < .001). (b) Comparing trust decrement between Patterns 0 and 4 (†p < .1).
Figure 12 (a) Comparing trust increment between Patterns 3 and 7 (***p < .001). (b) Comparing trust decrement between Patterns 0 and 4 (†p < .1).CC BY-NC · PubMed Central

Verification

Claims the paper makes

0 Valid · 0 Not valid · 25 Not checked · 40 Own results and interpretations

36 more claims

Create a free account to see all 65 claims and to verify the 1 reference still open.

See all claims — free account
  • It may state two separate claims as one. 15
  • Probably not a substantial claim. 15
  • The quote only partly states this claim. 1

Create a free account to see these claims.

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?

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • ConsistentF(1, 73) = 0.73reported p = .4 · recomputed p = .396

    “F(1,73)=.73, p = .40”

  • Consistentt(72) = 4.91reported p < .001 · recomputed p < .001

    “t(72)=4.91, p < .001”

  • Consistentt(72) = 2.72reported p < .01 · recomputed p = .008

    “t(72)=2.72, p < .01”

  • Consistentt(72) = 3.6reported p < .001 · recomputed p < .001

    “t(72)=3.60, p < .001”

  • Consistentt(72) = 2.74reported p < .01 · recomputed p = .008

    “t(72)=2.74, p < .01”

5 of 7 · the rest aren’t loaded

2 more recomputed tests

Create a free account to see every test the Stats check recomputed.

See all tests — free account

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Self-check results appear here once an independent check has confirmed them.

Citations

Cites

Show all 47 references

Research works printed but not linked: 2

  1. Gino F., 2009, Social Decision Making: Social Dilemmas, Social Values, and Ethical Judgments, 241
  2. Yang X. J., Human-automation interaction: Transportation

Cited by

See all 132

Metadata

Authors
Yang, X. J., Schemanske, C., & Searle, C.
Affiliations
University of Michigan (US)
Corresponding author
Yang, X. Jessie
Venue
Human Factors: The Journal of the Human Factors and Ergonomics Society
Published
2021-08-29
Volume
65
Issue
5
Pages
862-878
ISSN
  • 0018-7208 (ISSN-L)
  • 1547-8181
OpenAlex
W3180190266
Cited by
132
References
48
Citations by year
2026: 46 · 2025: 29 · 2024: 32 · 2023: 16 · 2022: 9
FWCI
9.90
Type
journal article
Language
en
Open access
Locations
  • Human Factors The Journal of the Human Factors and Ergonomics Society (publishedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • arXiv (Cornell University) (cc-by)
  • University of Michigan Library
Funders
National Science Foundation (2045009)
Topics
  • Social Sciences › Human Factors and Ergonomics › Human-Automation Interaction and Safety
  • Social Sciences › Psychology › Social Psychology › Social Robot Interaction and HRI
  • Physical Sciences › Engineering › Safety, Risk, Reliability and Quality › Risk and Safety Analysis
Keywords
trust in automation, trust dynamics, human-automation interaction, outcome bias, contrast effects, heuristic decision making, automated decision aids, automation failure
MeSH terms
Heuristics, Automation, Awareness, Humans, Male, Man-Machine Systems, Task Performance and Analysis, Trust

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.

A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

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

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Toward Quantifying Trust Dynamics: How People Adjust Their Trust After Moment-to-Moment Interaction With Automation. https://app.cochat.ai/papers/10.1177/00187208211034716
The paper
Yang, X. J., Schemanske, C., & Searle, C. (2023). Toward Quantifying Trust Dynamics: How People Adjust Their Trust After Moment-to-Moment Interaction With Automation. Human Factors: The Journal of the Human Factors and Ergonomics Society. https://doi.org/10.1177/00187208211034716

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.