Cited by 384Child and Animal Learning DevelopmentPsychology of Moral and Emotional JudgmentEmotions and Moral Behavior

The DOI matches its Crossref record; 65 of the 65 references it cites were found in a registry or an index; none of the 51 sources its claims cite is checked for support yet.

The accidental transgressor: Morally-relevant theory of mind

Open access

Abstract

To test young children’s false belief theory of mind in a morally relevant context, two experiments were conducted. In Experiment 1, children (N = 162) at 3.5, 5.5, and 7.5 years of age were administered 3 tasks: prototypic moral transgression task, false belief theory of mind task (ToM), and an “accidental transgressor” task, which measured a morally relevant false belief theory of mind (MoToM). Children who did not pass false belief ToM were more likely to attribute negative intentions to an accidental transgressor than children who passed false belief ToM, and to use moral reasons when blaming the accidental transgressor. In Experiment 2, children (N = 46) who did not pass false belief ToM viewed it as more acceptable to punish the accidental transgressor than did participants who passed false belief ToM. Findings are discussed in light of research on the emergence of moral judgment and theory of mind.

Overview

AI snapshot

Written from the paper’s full text.

TL;DR

In two experiments, children’s false-belief theory of mind was related to morally relevant judgments: children who failed false-belief ToM attributed more negative intentions to an accidental transgressor and endorsed more punishment, while those who passed distinguished intentions from outcomes. Embedding false belief in a moral context made the transgressor false-contents question harder than a prototypic false-belief task.

Key findings

  • Children who did not pass false-belief ToM were more likely to attribute negative intentions to the accidental transgressor and to use moral reasons when blaming him/her. In the MoToM task, children who passed false belief judged the actor’s intentions as more all right (M = 2.72) than they evaluated the act (M = 1.71), whereas children who failed did not differ (M = 1.56 vs. M = 1.50), F(1,150) = 10.63, p < .01.
  • The morally relevant false-contents task was harder than the prototypic false-contents task: more participants failed the MoToM false-contents task and succeeded on the prototypic task, χ²(1, N = 37) = 11.92, p < .01. Location-change false belief was equally difficult in both contexts.
  • In Experiment 2, children who passed MoToM false belief viewed punishment as less acceptable (M = .73) than children who did not pass (M = 1.61), F(1,31) = 7.862, p < .01. Children who passed also differentiated punishment acceptability between the moral transgression (M = 1.91) and MoToM (M = .69) scenarios, while children who failed did not.
  • All participants viewed the prototypic moral transgression as wrong (M = 1.26), and older participants viewed it as more wrong than younger groups, F(2,154) = 7.35, p < .01.

Methods

  • Experiment 1 included N = 162 children at 3.5, 5.5, and 7.5 years of age in a within-participants design with fixed task order. Tasks were a prototypic moral transgression (pushing someone off a swing), prototypic false-belief ToM tasks (false contents and location change), and a morally relevant theory-of-mind task (MoToM) in which a classroom helper throws away a paper bag containing another child’s cupcake.
  • Experiment 2 included N = 46 children in the same three age groups and repeated the tasks with an added punishment-acceptability question.
  • Analyses used ANOVAs with repeated measures, chi-square tests, and age as a covariate where noted; justifications were coded into harm, negligence, and lack of negative intent categories, with inter-rater reliability Cohen’s κ = .85.

Limitations

  • The paper states that the verbal protocol did not allow assessment of nonlinguistic participants and that nonverbal or spontaneous-response paradigms are needed. It also notes that the oldest participants still evaluated the accidental transgression as wrong, suggesting further work with older children and adults is warranted.
  • The paper notes a possible fixed-order carry-over effect, in which participants may have anticipated positive intentions after receiving a scenario with explicit positive intentions.
  • Evident from the text: Experiment 2 had a small sample (N = 46), and both experiments drew from suburbs of a large Mid-Atlantic city in preschools and elementary schools serving a middle- to low-income population, limiting generalizability.

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

Integrity signals

Registry

Crossref · 10.1016/j.cognition.2011.01.006

Cites

65 of 65 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

0 of 51 cited sources checked · 0 supported · 0 flagged

  • It may state two separate claims as one. 51
  • Probably not a substantial claim. 20
  • The quote only partly states this claim. 1
64 of 64
  • The paper relies on this without support.

    Children's responses to the hypothetical vignettes reflect their moral judgments in daily peer interactions.

    “Most situations confronted by children in their daily lives involve the simultaneous activation of these forms of knowledge. Children’s theory of mind competence and moral understanding are called upon when making decisions on the playground and among peers in social situations.”
  • The paper relies on this without support.

    Without an understanding of false belief, children should have difficulty with intent considerations.

    “Without an understanding of false belief, children should have difficulty with intent considerations.”
2 of 2

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(2, 154) = 7.35reported p < .01 · recomputed p < .001

    “F (2,154) = 7.35, p < .01”

  • ConsistentF(2, 141) = 40.63reported p < .001 · recomputed p < .001

    “F (2,141) = 40.63, p < .001”

  • ConsistentF(2, 157) = 32.02reported p < .001 · recomputed p < .001

    “F (2,157) = 32.02, p < .001”

  • ConsistentF(2, 157) = 45.97reported p < .001 · recomputed p < .001

    “F (2,157) = 45.97, p < .001”

  • ConsistentF(2, 154) = 34.81reported p < .001 · recomputed p < .001

    “F (2,154) = 34.81, p < .001”

  • ConsistentF(2, 159) = 37.35reported p < .001 · recomputed p < .001

    “F (2,159) = 37.35, p < .001”

  • Consistentχ²(3) = 39.16reported p < .001 · recomputed p < .001

    “χ2 (3, N = 155) = 39.16, p < .001”

  • Consistentχ²(3) = 88.73reported p < .001 · recomputed p < .001

    “χ2 (3, N = 160) = 88.73, p < .001”

  • Consistentχ²(1) = 11.92reported p < .01 · recomputed p < .001

    “χ2 (1, N = 37) = 11.92, p < .01”

  • ConsistentF(1, 150) = 35.63reported p < .001 · recomputed p < .001

    “F (1,150) = 35.63, p < .001”

  • ConsistentF(2, 150) = 4.05reported p < .05 · recomputed p = .019

    “F (2,150) = 4.05, p < .05”

  • ConsistentF(1, 150) = 10.63reported p < .01 · recomputed p = .001

    “F (1,150) = 10.63, p < .01”

  • ConsistentF(1, 152) = 7.99reported p < .01 · recomputed p = .005

    “F (1,152) = 7.99, p < .01”

  • ConsistentF(2, 318) = 62.84reported p < .001 · recomputed p < .001

    “F (2, 318) = 62.84, p < .001”

  • ConsistentF(4, 318) = 26.21reported p < .001 · recomputed p < .001

    “F (4,318) = 26.21, p < .001”

  • ConsistentF(2, 318) = 15.54reported p < .001 · recomputed p < .001

    “F (2, 318) = 15.54, p < .001”

  • ConsistentF(4, 318) = 3.96reported p < .01 · recomputed p = .004

    “F (4,318) = 3.96, p < .01”

  • ConsistentF(1, 150) = 4.56reported p < .05 · recomputed p = .034

    “F (1,150) = 4.56, p < .05”

  • ConsistentF(2, 150) = 4.16reported p < .05 · recomputed p = .017

    “F (2,150) = 4.16, p < .05”

  • ConsistentF(1, 138) = 9.27reported p < .01 · recomputed p = .003

    “F (1,138) = 9.27, p < .01”

  • ConsistentF(2, 314) = 104.84reported p < .001 · recomputed p < .001

    “F (2, 314) = 104.84, p < .001”

  • ConsistentF(2, 316) = 248.81reported p < .001 · recomputed p < .001

    “F (2,316) = 248.81, p < .001”

  • ConsistentF(1, 31) = 7.862reported p < .01 · recomputed p = .009

    “F (1, 31) = 7.862, p < .01”

  • ConsistentF(1, 27) = 4.75reported p < .05 · recomputed p = .038

    “F (1,27) = 4.75, p < .05”

  • ConsistentF(1, 34) = 5.91reported p < .05 · recomputed p = .020

    “F (1,34) = 5.91, p < .05”

  • ConsistentF(1, 43) = 8.75reported p < .01 · recomputed p = .005

    “F (1,43) = 8.75, p < .01”

  • ConsistentF(1, 43) = 8.3reported p < .01 · recomputed p = .006

    “F (1, 43) = 8.30, p < .01”

  • ConsistentF(1, 29) = 4.619reported p < .05 · recomputed p = .040

    “F (1, 29) = 4.619, p < .05”

28 of 28

Self-check

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

Each contradiction between two passages listed here was confirmed by an independent check.

  • Overall no-negative-intent rate does not match age-group rates

    Major

    The three age-group rates for no-negative-intent justifications average .54, not the reported overall .25. The stated group sizes also do not suggest a weighting that would bring the overall rate near .25.

    Reported 0.25 · recomputed 0.543333

  • Overall harm-justification rate exceeds every age-group rate

    Major

    For justifications of the accidental transgressor's intentions, the reported overall harm rate is .54, but the three reported age-group rates are .49, .23, and .03. An overall rate cannot exceed every group's rate.

2 of 2

Citations

Cites

Flagged first, then by citations

Show all 59 references

Research works printed but not linked: 2

  1. Smetana, 1995, Morality in context: Abstractions, ambiguities, and applications, Vol. 10, 83
  2. Turiel, 1998, The development of morality, 5th ed., Vol. 3, 863

Cited by

Sorted by citations

See all 384

Metadata

Authors
Killen, M., Mulvey, K. L., Richardson, C., Jampol, N., & Woodward, A.
Affiliations
  • University of Maryland, College Park (US)
  • University of Chicago (US)
Corresponding author
Killen, Melanie
Venue
Cognition
Publisher
Elsevier BV
Published
2011-05
Volume
119
Issue
2
Pages
197-215
ISSN
0010-0277 (ISSN-L)
OpenAlex
W2129128322
Cited by
384
References
59
Citations by year
2026: 18 · 2025: 21 · 2024: 27 · 2023: 18 · 2022: 55 · 2021: 31 · 2020: 35 · 2019: 28 · 2018: 31 · 2017: 22 · 2016: 34 · 2015: 16 · 2014: 13 · 2013: 17 · 2012: 11
FWCI
30.52
Type
journal article
Language
en
Open access
Locations
  • Cognition (publishedVersion)
  • PubMed (publishedVersion)
  • Europe PMC (PubMed Central) (acceptedVersion)
  • PubMed Central (acceptedVersion)
Funders
Eunice Kennedy Shriver National Institute of Child Health and Human Development (T32HD007542)
Topics
  • Social Sciences › Psychology › Developmental and Educational Psychology › Child and Animal Learning Development
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Psychology of Moral and Emotional Judgment
  • Social Sciences › Psychology › Social Psychology › Emotions and Moral Behavior
Keywords
theory of mind, false belief understanding, moral transgression, moral judgment, intention attribution, blame, punishment, moral reasoning development, children
MeSH terms
Accidents, Child, Child, Preschool, Concept Formation, Female, Humans, Judgment, Male, Morals, Punishment, Reproducibility of Results, Crime Victims, Theory of Mind

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 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (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