CoChat

Content Moderator Mental Health and Associations with Coping Styles: Replication and Extension of Previous Studies

Open access
Spence, Ruth, DeMarco, Jeffrey · 2025 · Behavioral Sciences

The DOI matches its Crossref record; 49 of the 51 references it cites were found in a registry or an index; 6 of the 43 sources its claims cite were checked for support.

Abstract

There is an increasing evidence base that demonstrates the psychological toll of content moderation on the employees that perform this crucial task. Nevertheless, content moderators (CMs) can be based worldwide and have varying working conditions. Therefore, there is a need for studies to be replicated to ensure that the results are robust. The current study used a large sample of commercial CMs employed by an international company to replicate the results from two previous studies that relied on an anonymous online survey. The results pertaining to mental health, wellbeing, and the effectiveness of wellbeing services for this population were mostly replicated. Over a quarter of CMs demonstrated moderate to severe psychological distress, and a quarter were experiencing low wellbeing. Further, the results suggest the potential utility for interventions that increase problem-focussed problem solving, as well as a need for the efficacy of wellbeing services to be evaluated more broadly.

Overview

CoChat Summary

## TL;DR

This study replicated prior findings in 160 commercial content moderators from an international company and found substantial mental health burden: over a quarter had moderate-to-severe psychological distress, and 28.8% had low wellbeing. Exposure frequency and coping style mattered, while wellbeing service use was only clearly linked to lower secondary trauma, not to distress or wellbeing.

## Key findings

  • Among 160 moderators, 24.4% scored in the moderate-to-severe range on the CORE-10, and 33.75% scored 13 or greater, suggesting symptoms associated with clinical depression.
  • 28.8% scored in the low wellbeing range on the SWEMWBS, and 13.5% scored in the high wellbeing range.
  • Daily exposure to distressing content was associated with significantly greater psychological distress and secondary trauma than weekly/monthly or less frequent exposure: psychological distress, F(2, 131) = 6.01, p = 0.003; secondary trauma, F(2, 131) = 8.72, p < 0.001.
  • Problem-focussed coping was associated with lower psychological distress, lower secondary trauma, and greater wellbeing; avoidant coping showed the opposite pattern. Emotion-focussed coping was associated with lower distress and trauma and greater wellbeing in the reported MGLM results.
  • Using wellbeing services was associated with significantly lower secondary trauma, but not with psychological distress or wellbeing; concerns about confidentiality were associated with higher distress and trauma and lower wellbeing.

## Methods

  • Cross-sectional anonymous online survey of 160 content moderators employed by an international company in the entertainment industry; the company employs approximately 200 moderators, so the response rate was 80%.
  • Measures included demographic/work questions, the CORE-10, SWEMWBS, Secondary and Vicarious Trauma Scale, and Brief-COPE.
  • Main analyses were descriptive statistics and multiple multivariate general linear models (MGLMs), adjusting for sex, location, duration in role, and age where relevant.

## Limitations

  • The authors state the study was cross-sectional, so it cannot show how experiences and symptoms change over time; they call for longitudinal research.
  • The authors note they are replicating their own work with data they cannot publicly share, so other researchers cannot yet independently reproduce the findings with this dataset.
  • Evident from the text: the sample comes from a single international company, so generalizability beyond that company is limited, even though the authors argue the 80% response rate supports representativeness within the company.

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

Integrity signals

Registry

Crossref · 10.3390/bs15040487

Cites

49 of 49 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Mean scores on psychological distress, secondary trauma, and wellbeing by frequency of exposure.
Figure 1 Mean scores on psychological distress, secondary trauma, and wellbeing by frequency of exposure.CC BY · PubMed Central
Figure 2 Mean scores on psychological distress, secondary trauma, and wellbeing by wellbeing service use, concerns about confidentiality, and if services understand the pressure of the role.
Figure 2 Mean scores on psychological distress, secondary trauma, and wellbeing by wellbeing service use, concerns about confidentiality, and if services understand the pressure of the role.CC BY · PubMed Central
Figure 3 Associations between psychological distress predicted values and different coping strategies.
Figure 3 Associations between psychological distress predicted values and different coping strategies.CC BY · PubMed Central
Figure 4 Associations between secondary trauma predicted values and different coping strategies.
Figure 4 Associations between secondary trauma predicted values and different coping strategies.CC BY · PubMed Central
Figure 5 Associations between wellbeing predicted values and different coping strategies.
Figure 5 Associations between wellbeing predicted values and different coping strategies.CC BY · PubMed Central

Verification

Claims the paper makes

6 Valid · 0 Not valid · 29 Not checked · 90 Own results and interpretations

66 more claims

Create a free account to see all 125 claims and to verify the 4 references still open.

See all claims — free account
  • Probably not a substantial claim. 38
  • It may state two separate claims as one. 30

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(2, 131) = 6.01reported p = .003 · recomputed p = .003

    “F(2, 131) = 6.01, p = .003”

  • ConsistentF(2, 131) = 8.72reported p < .001 · recomputed p < .001

    “F(2, 131) = 8.72, p < .001”

  • ConsistentF(2, 131) = 1.19reported n.s. · recomputed p = .307

    “F(2, 131) = 1.19, n.s.”

  • ConsistentF(1, 152) = 3.37reported n.s. · recomputed p = .068

    “F(1, 152) = 3.37, n.s.”

  • ConsistentF(1, 152) = 1.29reported n.s. · recomputed p = .258

    “F(1, 152) = 1.29, n.s.”

5 of 21 · the rest aren’t loaded

16 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

Research works printed but not linked: 4

  1. Trafimow, D. (in press). Replication and theory development in the social and psychological sciences. Research handbook on the replication crisis, Edward Elgar.
  2. Hofstede, G. (2001). Culture’s consequences: Comparing values, behaviors, institutions, and organizations across nations, Sage Publications. [2nd ed.].
  3. Pompili, 2017, The impact of COVID-19 on unemployment: A long-term risk factor for suicide?, Journal of Affective Disorders, 268, 157
  4. Steiger, M., Bharucha, T. J., Venkatagiri, S., Riedl, M. J., and Lease, M. (, January May). The psychological well-being of content moderators: The emotional labor of commercial moderation and avenues for improving support. CHI Conference on Human Factors in Computing Systems (CHI’21), Yokohama, Japan.

Cited by

See all 7

Metadata

Authors
Spence, R., & DeMarco, J.
Affiliations
Middlesex University London (GB)
Corresponding author
DeMarco, Jeffrey
Venue
Behavioral Sciences
Publisher
MDPI AG
Published
2025-04-08
Volume
15
Issue
4
Pages
487
ISSN
2076-328X (ISSN-L)
OpenAlex
W4409242339
Cited by
6
References
51
Citations by year
2026: 5 · 2025: 1
FWCI
6.60
Type
journal article
Language
en
Open access
Locations
  • Behavioral Sciences (publishedVersion, cc-by)
  • Preprints.org (acceptedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
Middlesex University
Topics
  • Social Sciences › Psychology › Social Psychology › Stress and Burnout Research
  • Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Workplace Health and Well-being
  • Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Employment and Welfare Studies
Keywords
mental health, psychological distress, well-being, coping styles, problem-focused coping, well-being services, replication, online content moderation

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.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

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).

Cited sources are checked with a language model’s help (gpt-6-sol).

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.