Pith. sign in

REVIEW 3 major objections 5 minor 129 references

Reply, Delete, or Ignore? Examining How Content Creators Perceive and Select Comment Moderation Strategies

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Content creators choose whether to delete, reply to, or ignore hateful comments based on expected emotional safety and impression management benefits; their beliefs about recommendation algorithms do not predict these choices.

desk verdict A solid within-subjects survey of creators' moderation choices whose headline null on algorithmic incentives needs a stronger measurement argument before we trust it. read the letter →

arxiv 2608.04951 v1 pith:6DRYDW66 submitted 2026-08-05 cs.HC

Yunhee Shim , Shagun Jhaver This is my paper · ORCID
classification cs.HC
keywords contentmoderationcreatorshatespeechcommentstrategiesemotionalsafetyimpressionmanagementalgorithmicincentivesmiddle-levelgovernance
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks how content creators decide what to do with hateful comments on their posts—delete them, reply to them, or ignore them—and what they think each strategy buys them. Surveying 584 creators, the authors find that deletion is seen as the most protective for emotional safety, that both deleting and replying look better for managing the creator's public image than ignoring, and that replying is believed to give the biggest algorithmic boost. The decisive finding is asymmetrical: expected safety and image gains significantly predict which strategy a creator says they would adopt, but perceived algorithmic benefits do not predict adoption of any of the three strategies. The authors argue that creators act as audience-facing middle-level governance actors, weighing self-protection against how their actions will read to viewers, and that platform design should support that role rather than offload safety entirely onto creators.

What carries the argument

The machinery is a within-subjects survey design that decomposes each moderation strategy into three perceived-benefit composites. Emotional safety is a six-item composite adapted from a harm-severity framework; impression management is a five-item composite derived from an established impression-management scale; algorithmic incentives are a four-item composite constructed from rank/time and follower/non-follower dimensions of algorithmic visibility. Repeated-measures ANOVAs compare the three strategies on each composite, and regression models test which composites predict self-reported adoption likelihood for each strategy while controlling for effort, gender, and follower count. The critical load-bearing piece is the algorithmic-incentive scale: it asks creators to speculate about counterfactual effects on recommendation systems, and the null result for this scale is the paper's headline.

What would settle it

A field experiment in which creators are randomly shown credible information that replying to hate comments increases algorithmic reach (or that deleting suppresses it), and their actual reply and delete rates over the following weeks are compared with a control group; if the treated group changes its moderation behavior, the paper's null claim that algorithm beliefs do not drive adoption would be overturned.

Watch

Extended reading notes

Core claim

The paper's central claim is that creators make comment-moderation decisions through a three-part evaluative lens—emotional safety, impression management, and algorithmic incentives—and that the three strategies (delete, reply, ignore) score differently on each lens. Deleting is rated highest for safety, with ignoring next and replying last; deleting and replying are both rated above ignoring for impression management with no reliable difference between them; and replying is rated highest for algorithmic benefits, followed by ignoring and deleting. In the adoption regressions, perceived safety and impression-management benefits are consistent positive predictors of willingness to adopt each strategy, while the perceived algorithmic-incentive composite is not a significant predictor for any strategy. The authors interpret this as evidence that creators' moderation practices are safety- and audience-driven actions, distinct from the visibility-seeking behaviors documented in earlier work on algorithmic labor.

Load-bearing premise

The load-bearing premise is that the survey's new four-item measure of perceived algorithmic benefits actually captures what creators believe about recommendation systems; if that scale is flawed, the headline null result—that algorithm beliefs don't predict moderation choices—could be an artifact of measurement rather than a real absence of effect.

Editorial extensions

If this is right

  • Because creators treat deletion as the safety-first option, platforms that add protections beyond deletion—repeat-offender restrictions, review queues, or filters based on removed content—could strengthen the protective promise and increase willingness to delete.
  • Since replying and deleting both help creators look like responsible comment-space stewards, lightweight norm-signaling tools that mark a comment as unacceptable without composing a full reply may capture much of replying's image benefit at lower effort and escalation risk.
  • Ignoring is not passive in creators' eyes: it scores above replying for safety and carries modest image benefits, so platforms should not assume visible unmoderated comments mean neglect; intermediate options like hiding comments from the creator's view could combine non-engagement with reduced exposure.
  • Algorithmic-incentive beliefs differ by strategy but do not predict adoption, so interventions aimed at changing creators' beliefs about recommendation algorithms are unlikely to shift their moderation behavior.
  • Perceived effort is not a barrier: replying was rated most effortful and was still adopted when image or safety gains were expected, suggesting creators accept governance labor when it signals their values.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: the null for algorithmic incentives may be a measurement artifact rather than a true absence of effect—the four-item scale asks creators to predict changes in opaque ranking systems, so a behavioral measure of algorithm beliefs might reveal an association the survey missed.
  • Editorial inference: if safety and image are the true drivers, then platform changes that make deletion backlash-proof (such as anonymized removals or auto-suggested filters) could push creators toward deletion even when they suspect it costs algorithmic reach.
  • Editorial inference: the scenario used mild identity-belittling comments without swear words; with severe hate speech, different audience sizes, or direct monetization tied to engagement, the ranking of strategies and the role of algorithmic incentives could shift, suggesting the findings are bounded to everyday, moderately toxic comment environments.
  • Editorial inference: a natural extension is to test the same three-dimensional framework against observed moderation behavior in logged platform data rather than self-reported adoption likelihood, which would clarify whether the stated intentions match creators' actions.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. This paper reports a within-subjects survey of 584 content creators recruited through Prolific. Participants read a vignette about receiving about 10 identity-based hateful comments on a post and rated three moderation strategies (deleting, replying, ignoring) on perceived emotional safety, impression management benefits, and algorithmic incentives, as well as their likelihood of adopting each strategy. The authors find that deleting is perceived as most beneficial for safety; deleting and replying are seen as better than ignoring for impression management; replying is seen as most algorithmically beneficial; and perceived safety and impression management benefits significantly predict adoption likelihood, while perceived algorithmic benefits do not.

Significance. If the findings hold, the paper provides a valuable empirical contribution to HCI and CSCW scholarship on creator-led moderation: it shows that creators evaluate moderation strategies along multiple, strategy-specific dimensions and that their adoption choices are driven primarily by safety and audience-image concerns rather than by beliefs about recommendation algorithms. The study has clear strengths: a relatively large sample of actual content creators, a preregistered-style analysis (though preregistration is not mentioned), appropriate repeated-measures ANOVA with Greenhouse-Geisser corrections and Tukey-adjusted post-hoc tests, honest reporting of disconfirmed hypotheses (H1c, H2a), and regression models with sensible controls (gender, follower size, effort). The central null result, however, rests on a self-developed four-item scale whose construct validity is not established, and one safety composite item appears to be scored ambiguously; these measurement questions are load-bearing for the paper's main claims.

major comments (3)
  1. [Section 3.3 and Tables 3-5] The headline null result—that perceived algorithmic benefits do not predict adoption—is only interpretable if the four-item algorithmic incentive composite validly measures strategy-specific algorithmic incentives. The manuscript reports only Cronbach's alpha (0.90-0.94), which establishes internal consistency but not construct or discriminant validity. The items ask creators to make speculative counterfactual forecasts about an opaque system (e.g., 'increase how long recommendation algorithms promote my content to people who do not currently follow me'), so ratings may capture generic engagement folk theories or acquiescence rather than strategy-specific algorithm incentives. Moreover, the same section reports a separate general-belief measure (alpha = .88) that is never used in the adoption regressions; this leaves open the possibility that the strategy-specific scale measures a correlated but distinct construct. Please report evidence for convergent/discriminant validity (e.g., correlations with the general-belief measure), or otherwise justify the interpretation of the null. For the replying model in Table 4, the coefficient is .075 with p = .085, so 'does not predict' rests on a conventional threshold; please complement the null tests with equivalence testing or a clear evidential statement.
  2. [Section 3.3 and Tables 2-5] The six-item emotional safety composite includes item (5), 'have consequences for the safety of other users on the platform beyond myself,' which measures consequences for others rather than the creator's own emotional safety, and item (6), 'result in an escalation of the conflict,' which is negatively valenced. The paper states that the six items 'were averaged to create a composite emotional safety score' but does not state whether item (6) was reverse-coded before averaging. If it was not, higher escalation scores paradoxically increase the safety composite, which would bias the safety comparisons in Table 2 and the adoption regressions in Tables 3-5. Please clarify the scoring, and if necessary re-run the analyses with the escalation item properly reverse-coded or removed, and report whether the H1a-H1c results and the adoption regressions are robust to that change.
  3. [Section 4.2 and Tables 3-5] The regression models enter perceived safety, impression management, and algorithmic incentives simultaneously but do not report collinearity diagnostics or zero-order correlations among these predictors. This is important because the strategy-specific algorithm ratings are likely correlated with the general belief about negative comments and algorithm promotion (Section 3.3), which is not included as a control. Please report variance inflation factors and the correlation matrix, and consider a robustness model that includes the general-belief measure as a control; otherwise the reported beta coefficients and the null result for algorithmic incentives may be unstable.
minor comments (5)
  1. [Introduction] There is a typo, 'visibilty' in the first sentence of the Introduction; it should be 'visibility.'
  2. [Table 8] In the content creation topic table, 'Health and fitness' reports 1 respondent (0.2%) and 'Others' reports 42 respondents (0.8%); the 'Others' percentage appears to be a typo, as 42/584 is approximately 7.2%.
  3. [Table 8] In the age categories, '60-54' should presumably read '60-64'.
  4. [Section 5.5] The text refers to effort results as being in Section 4.1, but the effort repeated-measures ANOVA is reported in Section 4.2.1; the cross-reference should be corrected.
  5. [Section 3.1] The paper reports excluding 120 responses that failed attention checks but does not specify how many attention checks were used or the threshold for failure; a brief description would help readers assess data quality.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is an empirical survey whose claims are statistical outcomes, not derivations from their own inputs.

full rationale

This paper makes no first-principles derivation; its central claims are empirical survey results obtained from a within-subjects experiment with 584 creators. The headline finding—that perceived algorithmic benefits do not predict adoption—is an estimated regression coefficient (e.g., β=.001, p=.977 in Table 3; β=.075, p=.085 in Table 4), an empirical outcome that could in principle have been significant, so it is not forced by construction. The algorithmic incentive composite is self-developed, but that is a measurement-validity concern rather than circularity: the four items ask creators to rate expected strategy-specific effects on recommendation visibility, and the regression treats those ratings as predictors; there is no equation in the paper that defines the predictor as the outcome. The only self-citation is to the authors' prior work, Shim and Jhaver [107], used in Section 2.4 to motivate why creators may perceive toxic comments as algorithmically favored. That citation is not load-bearing because the paper independently measures creators' general belief that negative comments increase algorithmic promotion (Section 4.1.1: M=4.58, t(583)=11.81, p<.001), which corroborates the premise without relying on the cited result. Hypotheses H3a–H3c are ordinary literature-derived predictions tested on new data, not restatements of the data used to fit them. The potential reverse-coding ambiguity of the emotional-safety item 'result in an escalation of the conflict' and the absence of equivalence testing for the null are methodological limitations, not circular reductions. Therefore, no circular step can be exhibited, and the appropriate score is 0.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

Because this is an empirical survey, the ledger records design choices rather than fitted constants. The hand-chosen scenario (100 comments, 10 hateful), the unit-weighted composites, and the unspecified attention-check exclusion rule shape all reported numbers. The domain assumptions that carry the argument are: self-reported likelihood stands in for real moderation behavior; the three selected strategies are a fair comparative set; Prolific self-identification reaches the creator population; and the adapted and new scales measure the intended constructs. These are standard for survey research but are exactly where a replication would fail if any of them is wrong.

free parameters (3)
  • Hypothetical scenario parameters = 100 comments; 10 inappropriate; subtle identity-based belittlement
    Hand-chosen stimulus values in Section 3.2, intended for ecological validity; they define the evaluation context that all measured perceptions refer to.
  • Attention-check exclusion rule = 120 of 755 responses (15.9%) excluded
    Section 3.1 excludes responses that failed attention checks but never specifies the number, content, or passing threshold of those checks, so the rule's effect on the analytic sample cannot be audited.
  • Unit-weighted composite scoring = Averages of 4-6 Likert items per construct
    Section 3.3 scores each dimension as the mean of its items; equal weighting is a modeling choice that affects the regression coefficients reported in Tables 3-5.
assumptions (5)
  • domain assumption Self-reported adoption likelihood in a hypothetical scenario tracks real moderation behavior
    The dependent variable is a single Likert item, 'How likely are you to (reply to/delete/ignore) these comments?' (Section 3.3); the paper's language about 'adoption' assumes stated likelihood transfers to actual behavior under real harassment and algorithmic uncertainty.
  • domain assumption The three strategies are comparable comment-level actions fully under creators' control
    Section 2.1 selects delete, reply, and ignore because their effects are immediate, comment-level, and creator-controlled; reporting, blocking, muting, and filtering are excluded, narrowing the governance space the findings describe.
  • domain assumption Prolific prescreening plus self-report screening identifies the target creator population
    Section 3.1 recruits through Prolific's 'Influencer' and 'Video content creator' filters and two self-report items; sample validity rests on participants' self-identification and on distribution of the survey to a panel rather than a platform-native creator sample.
  • domain assumption Adapted and newly written scales measure the intended constructs
    Section 3.3 adapts Bolino and Turnley's impression management items and Scheuerman et al.'s harm dimensions one-facet-per-item, and writes a new four-item algorithm scale with no external validation; construct validity is presumed from internal consistency (Cronbach's alpha).
  • standard math Repeated-measures ANOVA and OLS regression assumptions hold
    Sphericity was tested and corrected (Mauchly's W=.959, Greenhouse-Geisser), but normality, homoscedasticity, and multicollinearity for the regressions in Section 4.2 are not reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Reply, Delete, or Ignore? Examining How Content Creators Perceive and Select Comment Moderation Strategies." pith.science (2026). https://pith.science/paper/6DRYDW66

@misc{pith2026260804951,
  author       = {Pith},
  title        = {Pith review of: Reply, Delete, or Ignore? Examining How Content Creators Perceive and Select Comment Moderation Strategies},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6DRYDW66}},
  note         = {Machine review of arXiv:2608.04951}
}
read the original abstract

Content creators on social media sites occupy highly visible positions on their channels. As a result, creators, especially those with large followings, experience disproportionate levels of online harm. To address such harm, they enact a range of moderation strategies, which in turn shape the visibility of content that their audiences encounter. This paper examines how content creators perceive three moderation strategies to address hateful comments (deleting, replying to, or simply ignoring) and how they decide which strategy to deploy. While creator moderation is usually examined through the lens of safety, creators' regulation decisions may also be shaped by concerns about how their actions appear to audiences and how they are rewarded or penalized by platforms' recommendation algorithms. Conducting a survey of 584 content creators, we found that in their view, (1) deleting is the most beneficial for achieving safety, (2) both deleting and replying produce better impression management benefits than ignoring, and (3) replying is perceived to yield the highest algorithmic benefits. Crucially, while expectations of emotional safety and impression management benefits significantly predicted creators' willingness to adopt each comment moderation strategy, perceived algorithmic benefits did not. By unpacking how creators evaluate these trade-offs, this study contributes to HCI research on understanding creator-led, middle-level governance. We conclude with design implications for supporting creators as crucial governance actors without burdening them with sole responsibility for online safety.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

129 extracted references · 72 canonical work pages

  1. [1]

    Crystal Abidin. 2016. Visibility labour: Engaging with Influencers’ fashion brands and# OOTD advertorial campaigns on Instagram.Media International Australia161, 1 (2016), 86–100

    2016

  2. [2]

    Nazanin Andalibi, Pinar Ozturk, and Andrea Forte. 2017. Sensitive self-disclosures, responses, and social support on Instagram: The case of# depression. InProceedings of the 2017 ACM conference on computer supported cooperative work and social computing. 1485–1500

    2017

  3. [3]

    Shubham Atreja, Libby Hemphill, and Paul Resnick. 2023. Remove, reduce, inform: what actions do people want Social Media platforms to take on potentially misleading content?Proceedings of the ACM on Human-Computer Interaction7, CSCW2 (2023), 1–33

    2023

  4. [4]

    Babak Bahador. 2021. Countering hate speech. InThe Routledge companion to media disinformation and populism. Routledge, 507–518

    2021

  5. [5]

    Fabienne Baider. 2023. Accountability issues, online covert hate speech, and the efficacy of counter-speech.Politics and Governance11, 2 (2023), 249–260

    2023

  6. [6]

    2014.Accounts, excuses, and apologies: Image repair theory and research

    William L Benoit. 2014.Accounts, excuses, and apologies: Image repair theory and research. Suny Press

    2014

  7. [7]

    Thales Costa Bertaglia, Adrien Dubois, and Catalina Goanta. 2021. Clout chasing for the sake of content monetization: Gaming algorithmic architectures with self-moderation strategies.Morals & Machines1, 1 (2021), 22–29

    2021

  8. [8]

    Sophie Bishop. 2019. Managing visibility on YouTube through algorithmic gossip.New media & society21, 11-12 (2019), 2589–2606

    2019

Show all 129 references
  1. [9]

    Lindsay Blackwell, Jill Dimond, Sarita Schoenebeck, and Cliff Lampe. 2017. Classification and its consequences for online harassment: Design insights from heartmob.Proceedings of the ACM on human-computer interaction1, CSCW (2017), 1–19

    2017

  2. [10]

    Lindsay Blackwell, Nicole Ellison, Natasha Elliott-Deflo, and Raz Schwartz. 2019. Harassment in social VR: Implications for design. In2019 IEEE conference on virtual reality and 3D user interfaces (VR). IEEE, 854–855

    2019

  3. [11]

    Mark C Bolino and William H Turnley. 1999. Measuring impression management in organizations: A scale development based on the Jones and Pittman taxonomy.Organizational Research Methods2, 2 (1999), 187–206

    1999

  4. [12]

    Catherine Buerger. 2022. Why they do it: Counterspeech theories of change.A vailable at SSRN 4245211(2022)

    2022

  5. [13]

    Jie Cai, Aashka Patel, Azadeh Naderi, and Donghee Yvette Wohn. 2024. Content Moderation Justice and Fairness on Social Media: Comparisons Across Different Contexts and Platforms. InExtended Abstracts of the CHI Conference on Human Factors in Computing Systems. 1–9

    2024

  6. [14]

    Goran Calic, Rene Arseneault, and Maryam Ghasemaghaei. 2023. The dark side of Machiavellian rhetoric: Signaling in reward-based crowdfunding performance.Journal of Business Ethics182, 3 (2023), 875–896

    2023

  7. [15]

    Colin Campbell and Justine Rapp Farrell. 2020. More than meets the eye: The functional components underlying influencer marketing.Business horizons63, 4 (2020), 469–479

    2020

  8. [16]

    Despoina Chatzakou, Nicolas Kourtellis, Jeremy Blackburn, Emiliano De Cristofaro, Gianluca Stringhini, and Athena Vakali. 2017. Measuring# GamerGate: A tale of hate, sexism, and bullying. InProceedings of the 26th international conference on world wide web companion. 1285–1290

    2017

  9. [17]

    Yoonseo Choi, Eun Jeong Kang, Min Kyung Lee, and Juho Kim. 2023. Creator-friendly algorithms: Behaviors, challenges, and design opportunities in algorithmic platforms. InProceedings of the 2023 CHI Conference on Human Factors in Computing Systems. 1–22

    2023

  10. [18]

    Shih-Wei Chou and Guan-Ying Lu. 2022. Content creation intention in digital participation based on identity management on Twitch.Behaviour & Information Technology41, 12 (2022), 2578–2595

    2022

  11. [19]

    Danielle Keats Citron. 2014. Addressing cyber harassment: An overview of hate crimes in cyberspace.Case W. Res. JL Tech. & Internet6 (2014), 1

    2014

  12. [20]

    Frances Corry. 2021. Screenshot, save, share, shame: Making sense of new media through screenshots and public shame.First Monday(2021)

    2021

  13. [21]

    I can defend myself

    Amanda C Cote. 2017. “I can defend myself” women’s strategies for coping with harassment while gaming online. Games and culture12, 2 (2017), 136–155

    2017

  14. [22]

    Kelley Cotter. 2019. Playing the visibility game: How digital influencers and algorithms negotiate influence on Instagram.New media & society21, 4 (2019), 895–913. J. ACM, Vol. 37, No. 4, Article 111. Publication date: August 2027. Reply, Delete, or Ignore? Examining How Conte...

    2019

  15. [23]

    Kelley Cotter. 2024. Practical knowledge of algorithms: The case of BreadTube.New Media & Society26, 4 (2024), 2131–2150

    2024

  16. [24]

    Kate Crawford and Tarleton Gillespie. 2016. What is a flag for? Social media reporting tools and the vocabulary of complaint.New Media & Society18, 3 (2016), 410–428

    2016

  17. [25]

    Anna Davidovic, Catherine Talbot, Catherine Hamilton-Giachritsis, and Adam Joinson. 2023. To intervene or not to intervene: Young adults’ views on when and how to intervene in online harassment.Journal of Computer-Mediated Communication28, 5 (2023), zmad027

    2023

  18. [26]

    Fred D Davis. 1989. Perceived usefulness, perceived ease of use, and user acceptance of information technology.MIS quarterly13, 3 (1989), 319–340

    1989

  19. [27]

    David C DeAndrea and Megan A Vendemia. 2019. The influence of self-generated and third-party claims online: Perceived self-interest as an explanatory mechanism.Journal of Computer-Mediated Communication24, 5 (2019), 223–239

    2019

  20. [28]

    Michael Ann DeVito. 2022. How transfeminine TikTok creators navigate the algorithmic trap of visibility via folk theorization.Proceedings of the ACM on Human-Computer Interaction6, CSCW2 (2022), 1–31

    2022

  21. [29]

    Brooke Erin Duffy and Colten Meisner. 2023. Platform governance at the margins: Social media creators’ experiences with algorithmic (in) visibility.Media, Culture & Society45, 2 (2023), 285–304

    2023

  22. [30]

    Brooke Erin Duffy, Annika Pinch, Shruti Sannon, and Megan Sawey. 2021. The nested precarities of creative labor on social media.Social media+ society7, 2 (2021), 20563051211021368

    2021

  23. [31]

    Guillermo Echauri. 2026. Algorithmic-driven exposure: hyper-saturation and ever-expansion in the content creator ecosystem.Celebrity Studies(2026), 1–15

    2026

  24. [32]

    Stine Eckert. 2018. Fighting for recognition: Online abuse of women bloggers in Germany, Switzerland, the United Kingdom, and the United States.New Media & Society20, 4 (2018), 1282–1302

    2018

  25. [33]

    Nicole Ellison, Rebecca Heino, and Jennifer Gibbs. 2006. Managing impressions online: Self-presentation processes in the online dating environment.Journal of computer-mediated communication11, 2 (2006), 415–441

    2006

  26. [34]

    Motahhare Eslami, Karrie Karahalios, Christian Sandvig, Kristen Vaccaro, Aimee Rickman, Kevin Hamilton, and Alex Kirlik. 2016. First I" like" it, then I hide it: Folk Theories of Social Feeds. InProceedings of the 2016 cHI conference on human factors in computing systems. 2371–2382

    2016

  27. [35]

    Megan French and Jeff Hancock. 2017. What’s the folk theory? Reasoning about cyber-social systems.Reasoning About Cyber-Social Systems (February 2, 2017)(2017)

    2017

  28. [36]

    Joshua Garland, Keyan Ghazi-Zahedi, Jean-Gabriel Young, Laurent Hébert-Dufresne, and Mirta Galesic. 2022. Impact and dynamics of hate and counter speech online.EPJ data science11, 1 (2022), 3

    2022

  29. [37]

    R Stuart Geiger. 2016. Bot-based collective blocklists in Twitter: the counterpublic moderation of harassment in a networked public space.Information, Communication & Society19, 6 (2016), 787–803

    2016

  30. [38]

    Susan A Gelman and Cristine H Legare. 2011. Concepts and folk theories.Annual review of anthropology40, 1 (2011), 379–398

    2011

  31. [39]

    2018.Custodians of the Internet: Platforms, content moderation, and the hidden decisions that shape social media

    Tarleton Gillespie. 2018.Custodians of the Internet: Platforms, content moderation, and the hidden decisions that shape social media. Yale University Press

    2018

  32. [40]

    Rosalie Gillett, Zahra Stardust, and Jean Burgess. 2022. Safety for whom? Investigating how platforms frame and perform safety and harm interventions.Social Media+ Society8, 4 (2022), 20563051221144315

    2022

  33. [41]

    Angela Glotfelter. 2019. Algorithmic circulation: how content creators navigate the effects of algorithms on their work.Computers and composition54 (2019), 102521

    2019

  34. [42]

    Erving Goffman et al. 2002. The presentation of self in everyday life. 1959.Garden City, NY259 (2002), 2002

    2002

  35. [43]

    Michael H Goldhaber. 1997. The attention economy and the net.First monday(1997)

    1997

  36. [44]

    Eric Goldman. 2021. Content moderation remedies.Mich. Tech. L. Rev.28 (2021), 1

    2021

  37. [45]

    James Grimmelmann. 2015. The virtues of moderation.Yale JL & Tech.17 (2015), 42

    2015

  38. [46]

    Lea Gröber, Waleed Arshad, Angelica Goetzen, Elissa M Redmiles, Maryam Mustafa, Katharina Krombholz, et al

  39. [47]

    Andrew M Guess, Neil Malhotra, Jennifer Pan, Pablo Barberá, Hunt Allcott, Taylor Brown, Adriana Crespo-Tenorio, Drew Dimmery, Deen Freelon, Matthew Gentzkow, et al. 2023. How do social media feed algorithms affect attitudes and behavior in an election campaign?Science381, 6656...

    2023

  40. [48]

    Michael Haenlein, Ertan Anadol, Tyler Farnsworth, Harry Hugo, Jess Hunichen, and Diana Welte. 2020. Navigating the new era of influencer marketing: How to be successful on Instagram, TikTok, & Co.California management review63, 1 (2020), 5–25

    2020

  41. [49]

    2021.How to make money with fake news

    Jo Harper. 2021.How to make money with fake news. https://www.dw.com/en/the-conspiracy-business-how-to- make-money-with-fake-news/a-56660466 Accessed: 2025-09-04. J. ACM, Vol. 37, No. 4, Article 111. Publication date: August 2027. 111:22 Shim and Jhaver

    2021

  42. [50]

    Vulnerable, Victimized, and Objectified

    Sharon Heung, Lucy Jiang, Shiri Azenkot, and Aditya Vashistha. 2024. “Vulnerable, Victimized, and Objectified”: Understanding Ableist Hate and Harassment Experienced by Disabled Content Creators on Social Media. InProceedings of the 2024 CHI Conference on Human Factors in Comp...

    2024

  43. [51]

    Tatjana Hödl and Thomas Myrach. 2023. Content creators between platform control and user autonomy: the role of algorithms and revenue sharing.Business & Information Systems Engineering65, 5 (2023), 497–519

    2023

  44. [52]

    Phoebe Yiqing Huang, Jiaming Deng, Yingchen Yang, and Spencer Williams. 2025. Empowering Creators in the Fight Against Online Hate: A Qualitative Exploration of AI-Mediated Counterspeech Tools.Proceedings of the ACM on Human-Computer Interaction9, 7 (2025), 1–27

    2025

  45. [53]

    Jane Im, Sarita Schoenebeck, Marilyn Iriarte, Gabriel Grill, Daricia Wilkinson, Amna Batool, Rahaf Alharbi, Audrey Funwie, Tergel Gankhuu, Eric Gilbert, et al. 2022. Women’s perspectives on harm and justice after online harassment. Proceedings of the ACM on Human-Computer Inte...

    2022

  46. [54]

    Shagun Jhaver, Quan Ze Chen, Detlef Knauss, and Amy X Zhang. 2022. Designing word filter tools for creator-led comment moderation. InProceedings of the 2022 CHI conference on human factors in computing systems. 1–21

    2022

  47. [55]

    Shagun Jhaver, Seth Frey, and Amy X. Zhang. 2023. Decentralizing Platform Power: A Design Space of Multi-level Governance in Online Social Platforms.Social Media + Society(2023), 10 pages

    2023

  48. [56]

    Shagun Jhaver, Sucheta Ghoshal, Amy Bruckman, and Eric Gilbert. 2018. Online harassment and content moderation: The case of blocklists.ACM Transactions on Computer-Human Interaction (TOCHI)25, 2 (2018), 1–33

    2018

  49. [57]

    Shagun Jhaver, Yoni Karpfen, and Judd Antin. 2018. Algorithmic anxiety and coping strategies of Airbnb hosts. In Proceedings of the 2018 CHI conference on human factors in computing systems. 1–12

    2018

  50. [58]

    Shagun Jhaver, Alice Qian Zhang, Quan Ze Chen, Nikhila Natarajan, Ruotong Wang, and Amy X Zhang. 2023. Personalizing content moderation on social media: User perspectives on moderation choices, interface design, and labor.Proceedings of the ACM on Human-Computer Interaction7, ...

    2023

  51. [59]

    Shagun Jhaver and Amy X Zhang. 2023. Do users want platform moderation or individual control? Examining the role of third-person effects and free speech support in shaping moderation preferences.New Media & Society(2023), 14614448231217993

    2023

  52. [60]

    Nicholas John. 2024. A Classification of Features for Interpersonal Disconnectivity in Digital Media: Block, Unfriend, Unfollow, Mute, Withhold, and Eject.Media and Communication12, Article 8716 (2024)

    2024

  53. [61]

    Prerna Juneja, Deepika Rama Subramanian, and Tanushree Mitra. 2020. Through the Looking Glass: Study of Transparency in Reddit’s Moderation Practices.Proceedings of the ACM on Human-Computer Interaction4, GROUP (2020), 1–35

    2020

  54. [62]

    Cheryl R Kaiser, Nao Hagiwara, Lori W Malahy, and Clara L Wilkins. 2009. Group identification moderates attitudes toward ingroup members who confront discrimination.Journal of Experimental Social Psychology45, 4 (2009), 770–777

    2009

  55. [63]

    Anja Kalch and Teresa K Naab. 2017. Replying, disliking, flagging: How users engage with uncivil and impolite comments on news sites. (2017)

    2017

  56. [64]

    Eun Jeong Kang, Jingruo Chen, and Susan R Fussell. 2025. Understanding Content Creators’ Struggles and Expectations of AI in Direct Messaging. InProceedings of the Extended Abstracts of the CHI Conference on Human Factors in Computing Systems. 1–8

    2025

  57. [65]

    Brian TaeHyuk Keum, Michele J Wong, and India Sanders. 2024. Racial hate at the intersection of online and offline worlds: The joint impact of online and offline racism on the mental health of racially minoritized individuals.Journal of interpersonal violence39, 11-12 (2024), ...

    2024

  58. [66]

    Sara Kiesler, Robert Kraut, Paul Resnick, and Aniket Kittur. 2012. Regulating behavior in online communities.Building successful online communities: Evidence-based social design1 (2012), 4–2

    2012

  59. [67]

    ecological validity

    John F Kihlstrom. 2021. Ecological validity and “ecological validity”.Perspectives on Psychological Science16, 2 (2021), 466–471

    2021

  60. [68]

    Daniel Klug, Yiluo Qin, Morgan Evans, and Geoff Kaufman. 2021. Trick and please. A mixed-method study on user assumptions about the TikTok algorithm. InProceedings of the 13th ACM web science conference 2021. 84–92

    2021

  61. [69]

    Rewarding good creators

    Susanne Kopf. 2020. “Rewarding good creators”: Corporate social media discourse on monetization schemes for content creators.Social Media+ Society6, 4 (2020), 2056305120969877

    2020

  62. [70]

    Brianna L Lane, Ioana A Cionea, Norah E Dunbar, and Caleb T Carr. 2022. Antecedents and Effects of Online Third-Party Information on Offline Impressions.Journal of Media Psychology(2022)

    2022

  63. [71]

    Caroline Lang and Hannah Barton. 2015. Just untag it: Exploring the management of undesirable Facebook photos. Computers in Human Behavior43 (2015), 147–155

    2015

  64. [72]

    Jeehyun Jenny Lee. 2025. Managing unwanted visibility: how transnational Korean women content creators experience and manage harmful algorithmic visibility on global social media.Information, Communication & Society(2025), 1–17. J. ACM, Vol. 37, No. 4, Article 111. Publication...

    2025

  65. [73]

    Laura Leets. 2002. Experiencing hate speech: Perceptions and responses to anti-semitism and antigay speech.Journal of social issues58, 2 (2002), 341–361

    2002

  66. [74]

    Eden Litt. 2012. Knock, knock. Who’s there? The imagined audience.Journal of broadcasting & electronic media56, 3 (2012), 330–345

    2012

  67. [75]

    Eden Litt and Eszter Hargittai. 2016. The imagined audience on social network sites.Social Media+ Society2, 1 (2016), 2056305116633482

    2016

  68. [76]

    Renkai Ma, Xinning Gui, and Yubo Kou. 2023. Multi-platform content creation: the configuration of creator ecology through platform prioritization, content synchronization, and audience management. InProceedings of the 2023 CHI Conference on Human Factors in Computing Systems. 1–19

    2023

  69. [77]

    2010.Reputation Management and Social Media: How people monitor their identity and search for others online

    Mary Madden and Aaron Smith. 2010.Reputation Management and Social Media: How people monitor their identity and search for others online. Technical Report. Pew Research Center’s Internet & American Life Project. https: //www.pewresearch.org/internet/2010/05/26/reputation-manag...

    2010

  70. [78]

    Michela Mariotto, Sara Costa, Noemi Di Brango, Giuseppe Corbelli, Serena Verbena, Benedetta Emanuela Palladino, Antonio Zuffianò, Salvatore Ioverno, et al. 2026. Understanding Online Hate Toward Sexual and Gender Minorities: A Systematic Review.TRAUMA, VIOLENCE & ABUSE(2026), 0–19

    2026

  71. [79]

    Alice E Marwick. 2015. Instafame: Luxury selfies in the attention economy.Public culture27, 1 (2015), 137–160

    2015

  72. [80]

    Alice E Marwick and Danah Boyd. 2011. I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience.New media & society13, 1 (2011), 114–133

    2011

  73. [81]

    Tara Matthews, Elie Bursztein, Patrick Gage Kelley, Lea Kissner, Andreas Kramm, Andrew Oplinger, Andreas Schou, Manya Sleeper, Stephan Somogyi, Dalila Szostak, et al. 2025. Supporting the Digital Safety of At-Risk Users: Lessons Learned from 9+ Years of Research and Training.A...

    2025

  74. [82]

    It can feel uncomfortable to say something, but it’s for the best of the group

    Stephanie Murphy, Doireann Peelo Dennehy, Kellie Morrissey, John McCarthy, and Sarah Foley. 2026. “It can feel uncomfortable to say something, but it’s for the best of the group”: Examining Moderator Practices in Online Pregnancy Loss Communities.Proceedings of the ACM on Huma...

    2026

  75. [83]

    Maria Murumaa-Mengel and Liisi Maria Muuli. 2021. Misogynist content exposé pages on Instagram: Five types of shamings, moderators and audience members.Participations: Journal of Audience and Reception Studies18, 2 (2021), 100–123

    2021

  76. [84]

    Sarah Myers West. 2018. Censored, suspended, shadowbanned: User interpretations of content moderation on social media platforms.New Media & Society20, 11 (2018), 4366–4383

    2018

  77. [85]

    Sangwon Na, Thilo Kunkel, and Jason Doyle. 2020. Exploring athlete brand image development on social media: The role of signalling through source credibility.European Sport Management Quarterly20, 1 (2020), 88–108

    2020

  78. [86]

    Arvind Narayanan. 2023. Understanding social media recommendation algorithms. (2023)

    2023

  79. [87]

    2024.Social Media and Adolescent Health

    National Academies of Sciences, Engineering, and Medicine; Health and Medicine Division; Board on Population Health and Public Health Practice; Committee on the Impact of Social Media on Adolescent Health. 2024.Social Media and Adolescent Health. National Academies Press (US),...

    2024 doi

  80. [88]

    Magdalena Obermaier, Desirée Schmuck, and Muniba Saleem. 2023. I’ll be there for you? Effects of Islamophobic online hate speech and counter speech on Muslim in-group bystanders’ intention to intervene.New Media & Society 25, 9 (2023), 2339–2358

    2023

  81. [89]

    Michael Olsson. 2022. Architectures, algorithms & agency: the information practices of YouTube content creators. (2022)

    2022

  82. [90]

    Jessica A Pater, Moon K Kim, Elizabeth D Mynatt, and Casey Fiesler. 2016. Characterizations of online harassment: Comparing policies across social media platforms. InProceedings of the 2016 ACM International Conference on Supporting Group Work. 369–374

    2016

  83. [91]

    Kaike Ping, Anisha Kumar, Xiaohan Ding, and Eugenia H Rho. 2024. Behind the Counter: Exploring the Motivations and Barriers of Online Counterspeech Writing.ACM Transactions on Computer-Human Interaction(2024)

    2024

  84. [92]

    Jeffrey G Proudfoot, David Wilson, Joseph S Valacich, and Michael D Byrd. 2018. Saving face on Facebook: Privacy concerns, social benefits, and impression management.Behaviour & Information Technology37, 1 (2018), 16–37

    2018

  85. [93]

    Dimas Rahmatulah, Manik Sunuantari, and Tamara Klicek. 2024. Impression Management In Building Personal Branding Marco Randy.INJECT (Interdisciplinary Journal of Communication)9, 2 (2024), 315–328

    2024

  86. [94]

    Rayhan Rashed and Farnaz Jahanbakhsh. 2026. What If Moderation Didn’t Mean Suppression? A Case for Personalized Content Transformation. InProceedings of the 2026 CHI Conference on Human Factors in Computing Systems. 1–22

    2026

  87. [95]

    It’s easier than causing confrontation

    Yasmeen Rashidi, Apu Kapadia, Christena Nippert-Eng, and Norman Makoto Su. 2020. " It’s easier than causing confrontation": sanctioning strategies to maintain social norms and privacy on social media.Proceedings of the ACM on human-computer interaction4, CSCW1 (2020), 1–25. J....

    2020

  88. [96]

    Diana Rieger, Josephine B Schmitt, and Lena Frischlich. 2018. Hate and counter-voices in the Internet: Introduction to the special issue.SCM Studies in Communication and Media7, 4 (2018), 459–472

    2018

  89. [97]

    2014.Behind the screen: The hidden digital labor of commercial content moderation

    Sarah T Roberts. 2014.Behind the screen: The hidden digital labor of commercial content moderation. University of Illinois at Urbana-Champaign

    2014

  90. [98]

    Jian Rui and Michael A Stefanone. 2013. Strategic self-presentation online: A cross-cultural study.Computers in human behavior29, 1 (2013), 110–118

    2013

  91. [99]

    Millions} of people are watching{you

    Patrawat Samermit, Anna Turner, Patrick Gage Kelley, Tara Matthews, Vanessia Wu, Sunny Consolvo, and Kurt Thomas. 2023.{“Millions} of people are watching{you”}: Understanding the{Digital-Safety} Needs and Practices of Creators. In32nd USENIX Security Symposium (USENIX Security...

    2023

  92. [100]

    Julia Sasse and Jens Grossklags. 2023. Breaking the silence: Investigating which types of moderation reduce negative effects of sexist social media content.Proceedings of the ACM on Human-Computer Interaction7, CSCW2 (2023), 1–26

    2023

  93. [101]

    Morgan Klaus Scheuerman, Jialun Aaron Jiang, Casey Fiesler, and Jed R Brubaker. 2021. A framework of severity for harmful content online.Proceedings of the ACM on Human-Computer Interaction5, CSCW2 (2021), 1–33

    2021

  94. [102]

    Brubaker

    Morgan Klaus Scheuerman, Jialun Aaron Jiang, Casey Fiesler, and Jed R. Brubaker. 2021. A Framework of Severity for Harmful Content Online.Proc. ACM Hum.-Comput. Interact.5, CSCW2 (Oct. 2021), 368:1–368:33. https://doi.org/10. 1145/3479512

    2021

  95. [103]

    Carla Schieb and Mike Preuss. 2016. Governing hate speech by means of counterspeech on Facebook. In66th ica annual conference, at fukuoka, japan. 1–23

    2016

  96. [104]

    Sarita Schoenebeck, Oliver L Haimson, and Lisa Nakamura. 2021. Drawing from justice theories to support targets of online harassment.new media & society23, 5 (2021), 1278–1300

    2021

  97. [105]

    Leah Scolere, Urszula Pruchniewska, and Brooke Erin Duffy. 2018. Constructing the platform-specific self-brand: The labor of social media promotion.Social Media+ Society4, 3 (2018), 2056305118784768

    2018

  98. [106]

    Carol F Scott, Gabriela Marcu, Riana Elyse Anderson, Mark W Newman, and Sarita Schoenebeck. 2023. Trauma- informed social media: Towards solutions for reducing and healing online harm. InProceedings of the 2023 CHI Conference on Human Factors in Computing Systems. 1–20

    2023

  99. [107]

    Yunhee Shim and Shagun Jhaver. 2025. The Pin of Shame: Examining Content Creators’ Adoption of Pinning Inappropriate Comments as a Moderation Strategy.arXiv preprint arXiv:2505.14844(2025)

    2025 arXiv

  100. [108]

    Joshua Smallridge, Philip Wagner, and Justin N Crowl. 2016. Understanding cyber-vigilantism: A conceptual frame- work.Journal of Theoretical & Philosophical Criminology8, 1 (2016)

    2016

  101. [109]

    I feel physically safe but not politically safe

    Ananta Soneji, Vaughn Hamilton, Adam Doupé, Allison McDonald, and Elissa M Redmiles. 2024. " I feel physically safe but not politically safe": Understanding the Digital Threats and Safety Practices of{OnlyFans} Creators. In33rd USENIX Security Symposium (USENIX Security 24). 1–18

    2024

  102. [110]

    Haimson, and Danielle Lottridge

    Katta Spiel, Oliver L. Haimson, and Danielle Lottridge. 2019. How to Do Better with Gender on Surveys: A Guide for HCI Researchers. InExtended Abstracts of the 2019 CHI Conference on Human Factors in Computing Systems. 1–8. https://doi.org/10.1145/3290607.3310439

    2019

  103. [111]

    Hanne M Stegeman, Carolina Are, and Thomas Poell. 2024. Strategic invisibility: How creators manage the risks and constraints of online hyper (in) visibility.Social Media+ Society10, 2 (2024), 20563051241244674

    2024

  104. [112]

    Vidi Sukmayadi, Dante Darmawangsa, Suffian Hadi Ayub, and Sarah Annisa Fadhila. 2024. Constructing fame: A phenomenological study of online impression management among Indonesian TikTok celebrities.The Qualitative Report29, 6 (2024), 1727–1741

    2024

  105. [113]

    Shaojing Sun, Zhiyuan Liu, and David Waxman. 2025. A dynamical measure of algorithmically infused visibility. Royal Society Open Science12, 11 (2025)

    2025

  106. [114]

    Wondwesen Tafesse and Mumin Dayan. 2023. Content creators’ participation in the creator economy: Examining the effect of creators’ content sharing frequency on user engagement behavior on digital platforms.Journal of Retailing and Consumer Services73 (2023), 103357

    2023

  107. [115]

    Edson C Tandoc Jr, Darren Lim, and Rich Ling. 2020. Diffusion of disinformation: How social media users respond to fake news and why.Journalism21, 3 (2020), 381–398

    2020

  108. [116]

    Kurt Thomas, Devdatta Akhawe, Michael Bailey, Dan Boneh, Elie Bursztein, Sunny Consolvo, Nicola Dell, Zakir Durumeric, Patrick Gage Kelley, Deepak Kumar, et al. 2021. Sok: Hate, harassment, and the changing landscape of online abuse. In2021 IEEE symposium on security and priva...

    2021

  109. [117]

    It’s common and a part of being a content creator

    Kurt Thomas, Patrick Gage Kelley, Sunny Consolvo, Patrawat Samermit, and Elie Bursztein. 2022. “It’s common and a part of being a content creator”: Understanding How Creators Experience and Cope with Hate and Harassment Online. InProceedings of the 2022 CHI conference on human...

    2022

  110. [118]

    Alison B Tuck and Renee J Thompson. 2024. The social media use scale: Development and validation.Assessment31, 3 (2024), 617–636

    2024

  111. [119]

    Jirassaya Uttarapong, Jie Cai, and Donghee Yvette Wohn. 2021. Harassment experiences of women and LGBTQ live streamers and how they handled negativity. InProceedings of the 2021 ACM international conference on interactive J. ACM, Vol. 37, No. 4, Article 111. Publication date: ...

    2021

  112. [120]

    The algorithm is like a mercurial god

    Roland Verwiebe, Claudia Buder, Sarah Weissmann, Chiara Osorio-Krauter, and Aaron Philipp. 2024. “The algorithm is like a mercurial god”: Exploring content creators’ perception of algorithmic agency on YouTube.New Media & Society(2024), 14614448241307931

    2024

  113. [121]

    Maheshya Weerasinghe, Shaun Macdonald, Cristina Fiani, Joseph O’Hagan, Mathieu Chollet, Mark McGill, and Mohamed Khamis. 2025. Beyond mute and block: adoption and effectiveness of safety tools in social VR, from ubiquitous harassment to social sculpting.IEEE Transactions on Vi...

    2025

  114. [122]

    Richard Ashby Wilson and Molly K Land. 2020. Hate speech on social media: Content moderation in context.Conn. L. Rev.52 (2020), 1029

    2020

  115. [123]

    Donghee Yvette Wohn and Guo Freeman. 2020. Audience Management Practices of Live Streamers on Twitch. In ACM International Conference on Interactive Media Experiences(Cornella, Barcelona, Spain)(IMX ’20). Association for Computing Machinery, New York, NY, USA, 106–116. https:/...

    2020

  116. [124]

    Lucas Wright, Derek Ruths, Kelly P Dillon, Haji Mohammad Saleem, and Susan Benesch. 2017. Vectors for counter- speech on twitter. InProceedings of the first workshop on abusive language online. 57–62

    2017

  117. [125]

    Eva Yiwei Wu, Emily Pedersen, and Niloufar Salehi. 2019. Agent, gatekeeper, drug dealer: How content creators craft algorithmic personas.Proceedings of the ACM on Human-Computer Interaction3, CSCW (2019), 1–27

    2019

  118. [126]

    Jeong-yun Yeom, Ryu-won Kim, and Se-hoon Jeong. 2020. A Meta-analysis of the Effects of User Comments.Journal of Communication Research57, 2 (2020), 5–49. https://s-space.snu.ac.kr/handle/10371/214465 Meta-analysis of domestic comment effect studies; 47 studies analyzed

    2020

  119. [127]

    Alice Qian Zhang, Kaitlin Montague, and Shagun Jhaver. 2023. Cleaning up the streets: Understanding motivations, mental models, and concerns of users flagging social media posts.arXiv preprint arXiv:2309.06688(2023)

    2023 arXiv

  120. [128]

    Marc Ziegele and Pablo B Jost. 2020. Not funny? The effects of factual versus sarcastic journalistic responses to uncivil user comments.Communication research47, 6 (2020), 891–920. J. ACM, Vol. 37, No. 4, Article 111. Publication date: August 2027. 111:26 Shim and Jhaver A SUR...

    2020

  121. [2024]

    I chose to fight, be brave, and to deal with it

    " I chose to fight, be brave, and to deal with it": Threat Experiences and Security Practices of Pakistani Content Creators. In33rd USENIX Security Symposium (USENIX Security 24). 19–36

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.