Chapter 20 Exercises: The Outrage Machine: Anger as Engagement

Instructions

These exercises are designed to develop critical understanding of outrage dynamics in social media through empirical observation, research engagement, analysis, creative work, and structured discussion. Exercises are labeled by type: [Reflection], [Research], [Analysis], [Creative], and [Group Discussion].

[Reflection] Your Outrage History Think about the last five times you shared or commented on content because it made you angry. For each instance: What made the content outrage-inducing? Did you verify its accuracy before sharing? Did your response add to or reduce the emotional temperature of your network? Reflect on what your own outrage behavior reveals about how the outrage cycle operates through individual users like you.
[Research] Brady et al. (2017) Primary Source Find and read the original Brady et al. (2017) paper "Emotion shapes the diffusion of moralized content in social networks" in the Proceedings of the National Academy of Sciences. Write a 600-word summary for a general audience explaining the study's methodology, key findings, and limitations. What are the study's most important implications for how we design and use social media?
[Analysis] The Moral-Emotional Word Audit Choose three politically contested social media posts — one that generated very high engagement (many shares/comments), one medium engagement, and one low engagement. Highlight all moral-emotional language in each post (words that combine emotional charge with moral framing). Test Brady et al.'s hypothesis: Do the higher-engagement posts contain more moral-emotional language? Write up your findings.
[Research] Kramer et al. and the Ethics of Emotional Contagion Research Research the controversy surrounding the Kramer et al. (2014) Facebook emotional contagion study. Find: (a) the original paper, (b) at least two major critical responses to the study's ethics, and (c) Facebook's public defense of the research. Write a 500-word analysis: Was the research ethical? What ethical frameworks might lead to different conclusions about this?
[Reflection] The Anger-to-Action Inventory For one week, keep a record of every time you feel anger as a result of social media content. For each instance, note: the platform, the type of content, what action (if any) you took (share, comment, leave, nothing), and how you felt an hour later. At the end of the week, analyze your data. What patterns do you see? Did your anger-driven actions feel worthwhile in retrospect?
[Analysis] Mapping the Outrage Cycle Identify a specific outrage cycle that played out publicly on social media in the past six months. Trace the five stages described in the chapter (Provocation, Primary Reaction, Counter-Reaction, Algorithmic Amplification, Escalation or Decay) in your chosen case. Write a 600-word case analysis with specific examples at each stage.
[Group Discussion] The Creator Dilemma In groups of 3-4, discuss: You are a content creator who has built an audience of 50,000 subscribers discussing political topics. You have discovered through experience that content containing strong moral-emotional language performs significantly better than measured, nuanced analysis. You need the revenue. What do you do? What ethical considerations apply? What would you actually do if in this situation?
[Research] YouTube Radicalization: Chaslot and AlgoTransparency Research Guillaume Chaslot's AlgoTransparency project (algotransparency.org or related coverage). What methodology does the project use? What findings has it produced? How has YouTube responded to Chaslot's claims? Write a 500-word analysis of the project and its implications for understanding algorithmic radicalization.
[Analysis] The "Ratio" in Practice Spend 30 minutes on Twitter/X specifically searching for examples of heavily "ratioed" posts — posts where replies substantially exceed likes. Collect five examples. For each: What made the post controversial? What do the replies look like? Does the topic fit the moral-emotional language patterns described by Brady et al.? Write a brief analysis of each example.
[Creative] Write Two Versions Choose a genuinely important social issue that you care about. Write two versions of a social media post about this issue: (a) a version that maximizes moral-emotional language and outrage activation, following the patterns documented in Brady et al., and (b) a version that presents the same information without moral-emotional amplification. Which version are you more likely to share? Which version do you think does more good? What does the comparison reveal?
[Research] Moral Foundations Theory Read Jonathan Haidt and Jesse Graham's foundational work on moral foundations theory, particularly the 2007 paper "When Morality Opposes Justice: Conservatives Have Moral Intuitions that Liberals may not Recognize." List the six moral foundations and research what types of social media content most effectively activate each one. Write a 400-word analysis of how the moral foundations map onto different political communities' outrage content.
[Group Discussion] Intent vs. Effect in Corporate Ethics The chapter argues that outrage amplification was initially unintended but that once documented, continuing it becomes a choice. Discuss as a group: Does intent matter in corporate ethics? If a pharmaceutical company accidentally develops a drug with serious side effects, then learns of those effects but continues selling it, is the post-knowledge period morally different from the pre-knowledge period? How does this framework apply to social media platforms and outrage amplification?
[Analysis] The Facebook Files Research the Facebook Files reporting by the Wall Street Journal (September-October 2021) and the subsequent congressional testimony by Frances Haugen. Identify three specific internal Facebook research findings about outrage amplification or divisive content that were documented in the reporting. For each finding, analyze: What did the research show? What response (if any) did Facebook take? What response did the research justify that Facebook did not take?
[Creative] Redesign the News Feed You have been asked to redesign Facebook's News Feed algorithm to minimize outrage amplification while maintaining the platform's core value proposition (connecting people with content they find valuable and meaningful). Write a detailed specification for your redesigned algorithm, including: what signals you would use to rank content, what signals you would de-emphasize or remove, how you would measure success, and what tradeoffs you anticipate.
[Reflection] The Emotion After the Outrage Think about a time you were significantly activated by outrage content on social media — you shared it, argued in the comments, or spent significant time in an outrage thread. How did you feel immediately after the engagement? How did you feel the next day? Was the issue that caused your outrage actually more or less resolved as a result of your social media engagement? Reflect on the relationship between outrage engagement and actual social change.
[Research] Cross-Platform Outrage Comparison Research outrage dynamics on three different platforms: Facebook, Twitter/X, and YouTube. For each platform, find academic or journalistic documentation of how outrage content performs relative to other content types. Write a comparative analysis: Which platform's design features make outrage amplification most severe? Which platform has made the most effective modifications to reduce outrage amplification?
[Analysis] Emotional Contagion in Your Feed Without changing your behavior, observe your own emotional state before and immediately after social media sessions for one week. Rate your emotional state on a simple scale (1=very negative, 10=very positive) before and after each session. At the end of the week, analyze: Do social media sessions reliably shift your emotional state? In what direction? Does the pattern suggest emotional contagion effects consistent with Kramer et al.'s findings?
[Group Discussion] The Democratic Stakes The chapter suggests that outrage amplification may harm democratic societies by intensifying polarization and reducing support for democratic norms. Discuss: How seriously should we take this claim? What evidence would you need to believe that social media is genuinely damaging democracy (rather than merely reflecting pre-existing polarization)? If the evidence were conclusive, what responses would be appropriate — and by whom?
[Creative] The Outrage Antidote Campaign Design a public education campaign aimed at reducing the impact of social media outrage dynamics on political discourse. Your campaign should: accurately explain the outrage cycle in accessible terms, provide specific behavioral strategies individuals can use to reduce their participation in outrage dynamics, and propose platform-level changes that could support individual behavior change. Create a campaign brief, a sample social media post, and a one-page explainer.
[Research] Ribeiro et al. (2019) and the Alternative Influence Network Find and read Ribeiro et al.'s 2019 paper on YouTube's recommendation system and radicalization (available on arXiv or through academic databases). Write a 500-word summary of the paper's methodology, key findings, and the debate that followed. How did YouTube respond? What subsequent research has confirmed or challenged the findings?
[Analysis] Anger vs. Outrage: Is There a Productive Version? The chapter focuses primarily on how anger is exploited by social media algorithms for commercial purposes. But anger has also been a driver of legitimate social movements — civil rights activism, labor organizing, environmental advocacy. Analyze three historical or contemporary social movements: Was anger a productive force in each? How did it manifest differently from the algorithmically amplified outrage described in the chapter? What distinguishes productive moral anger from counterproductive outrage amplification?
[Reflection] Your Outrage Threshold Think about what types of content reliably make you angry online. Try to identify: What are your specific moral foundations (drawing on Haidt's framework)? What types of moral violations most reliably trigger your outrage? How has your outrage threshold changed over the years you have been using social media? Do you think you are more or less easily outraged than you were before you began using social media intensively?
[Group Discussion] Platform Self-Regulation vs. Government Regulation YouTube reduced "borderline content" recommendations and Facebook made various algorithm changes in response to internal and external pressure on outrage amplification. Discuss: Is platform self-regulation sufficient to address the outrage amplification problem? What are its structural limitations? What would government regulation of recommendation algorithms look like? What First Amendment or free speech concerns arise? Which approach does your group find most promising?
[Research] The Huszár et al. Twitter Study Research the study by Huszár and colleagues at Twitter (published around 2022) that examined whether Twitter's recommendation algorithm amplified political right or left content differentially. Find the original study and critical responses to it. Write a 400-word summary: What did the study find? How was it received? What are its limitations? What does it tell us about the relationship between algorithmic optimization and political content?
[Creative] Write the Internal Memo You are a researcher on Facebook's integrity team in 2018. You have just completed research showing that the "meaningful social interactions" algorithm change has amplified outrage content rather than reducing it, as the research predicted. Write the internal memo to your leadership team explaining your findings, their implications, and the response options available to the company. Be specific about what you found, why it happened, and what you recommend.
[Analysis] The Content Creator Study Research three to five prominent political content creators on YouTube or podcasting who cover the same general political space. Compare their rhetorical approaches: Which uses more moral-emotional language? Which generates more engagement? Has any of them changed their approach over time toward or away from outrage content? What might explain the patterns you find?
[Reflection] The Haidt Babel Thesis Read Jonathan Haidt's 2022 Atlantic article "Why the Past 10 Years of American Life Have Been Uniquely Stupid." Write a one-page reflection: Do you find Haidt's argument convincing? What evidence does he provide, and what evidence do you find most and least persuasive? What counter-arguments can you identify? How does Haidt's thesis connect to your own experience of social media and political discourse?
[Group Discussion] Optimizing for What? The chapter suggests that alternatives to engagement optimization — like optimizing for user-reported satisfaction, content diversity, or well-being — could reduce outrage amplification. In groups, discuss: What would you want a social media algorithm to optimize for? How would you measure your chosen outcome? What would be the likely unintended consequences of optimizing for your chosen metric instead of engagement? Which outcome metric does your group think would produce the best platform experience?
[Research] Outrage and Misinformation Research the relationship between outrage content and misinformation. Find at least two studies examining whether high-outrage content is more or less likely to be accurate than low-outrage content. Write a 500-word analysis: Does outrage amplification correlate with misinformation amplification? What are the implications for the quality of information in algorithmically curated environments?
[Creative] The Day After Outrage Write a short story (1,000-1,500 words) from the perspective of someone who has spent an entire day immersed in a social media outrage cycle — following and contributing to a rapidly spreading controversy. The story should capture the phenomenology of the experience (what it feels like from the inside) and end with a quiet moment of reflection the following morning. The story should make the reader feel something — not lecture them.
[Analysis] The Outrage Premium: A Financial Analysis Research the advertising economics of high-engagement content. If outrage content generates 20-30% more engagement than non-outrage content (per Brady et al. and related research), what is the approximate financial value of the outrage premium for a major platform? Using publicly available data on Facebook's or YouTube's advertising revenues and content volumes, estimate what portion of platform revenue might be directly attributable to outrage amplification. Reflect on what this calculation implies for platform incentives.
[Research] International Comparisons Research outrage amplification dynamics in countries other than the United States, particularly in contexts where social media has been implicated in ethnic violence or political crisis (Myanmar, Ethiopia, India). Find academic or journalistic documentation of specific cases. Write a 600-word analysis: Do the outrage dynamics described in this chapter operate in the same way across different political and cultural contexts? What do international cases reveal that US-centric analysis might miss?
[Reflection] After the Outrage: Outcomes Think about a specific political or social controversy that activated significant outrage on social media in the past two years. What was the eventual outcome of the controversy? Did social media outrage contribute to positive change, negative change, or no change? Write a reflection on the relationship between social media outrage cycles and real-world outcomes: Is outrage on social media a meaningful political force, an emotional safety valve, or something else?
[Group Discussion] The Disgust Emoji Case In the Velocity Media sidebar, the team debates adding a disgust emoji reaction. The ethics lead argues against it on the grounds that it will train the algorithm to amplify disgust-inducing content. The product lead argues for it on user expression grounds. In groups of 4-6, conduct a formal product review of the disgust emoji proposal. Each person should argue from a specific stakeholder perspective: product, ethics, user advocacy, advertiser relations, legal, and CEO. What does your group decide?
[Research] Solutions Research Research at least four specific technical, policy, or behavioral interventions that researchers or platform practitioners have proposed to reduce outrage amplification on social media. For each: What is the proposed intervention? What evidence supports its effectiveness? What are its limitations or potential unintended consequences? Write a 700-word analysis evaluating which intervention you find most promising and why.