Chapter 24: Quiz

Facebook's News Feed: A Decade of Optimization Against Users

Instructions: Select the best answer for each question. Each question has one correct answer.

1. When Facebook launched News Feed in September 2006, approximately what percentage of Facebook's total user base signed the protest petition within 48 hours?

A) 1 percent
B) 5 percent
C) 10 percent
D) 25 percent

2. In his 2006 open letter responding to the News Feed protest, Mark Zuckerberg characterized the launch mistake primarily as:

A) A fundamental violation of user privacy that required the feature to be removed
B) A communication and process failure, not a substantive problem with the feature itself
C) An accidental deployment that engineers had not fully tested
D) A deliberate decision that had been overruled by the board of directors

3. According to Albert Hirschman's framework of "exit, voice, and loyalty," the 2006 News Feed protestors primarily exercised which option — and why was it ineffective?

A) Exit — but they lacked alternative platforms that offered equivalent social functionality
B) Voice — but they lacked credible exit because of network lock-in effects
C) Loyalty — but passive acceptance was misread by Facebook as endorsement
D) Exit — but the cost of leaving was subsidized by Facebook's free service model

4. The Like button, launched in February 2009, introduced which psychological reinforcement mechanism into the act of posting on Facebook?

A) Fixed interval reinforcement — rewards arriving at predictable, scheduled times
B) Fixed ratio reinforcement — rewards arriving after a set number of actions
C) Variable ratio reinforcement — rewards arriving at unpredictable intervals
D) Continuous reinforcement — rewards arriving after every action

5. In the EdgeRank algorithm, the "affinity" factor (Ue) measured:

A) How many total friends a user had
B) How recently a piece of content had been posted
C) The closeness of the relationship between the viewing user and content creator, based on past interactions
D) The overall popularity of a piece of content across the entire platform

6. In EdgeRank's weight (We) factor, which type of user action was typically assigned the highest weight?

A) Clicking the Like button
B) Leaving a comment
C) Sharing content to one's own feed
D) Viewing a video for more than 10 seconds

7. The removal of the chronological feed option after 2012 was significant primarily because:

A) It reduced the amount of content users saw, improving their mental health
B) It eliminated users' ability to audit what the algorithm was and was not showing them
C) It forced content creators to post more frequently to maintain visibility
D) It reduced Facebook's server costs by limiting the amount of data processed

8. The 2014 emotional contagion experiment (Kramer et al.) involved manipulation of how many Facebook users' News Feeds?

A) 68,900
B) 689,003
C) 6,890,030
D) 68,900,300

9. Facebook's primary legal defense of the emotional contagion experiment's ethical adequacy was that:

A) The experiment caused no measurable psychological harm to participants
B) The research had been reviewed and approved by a university IRB committee
C) Users had consented to research participation through the platform's Data Use Policy
D) The experiment was not technically "research" because it had no control group

10. The emotional contagion experiment found that users exposed to more negative content in their feeds subsequently:

A) Spent less time on Facebook and logged off more frequently
B) Posted more negatively in their own subsequent status updates
C) Unfriended more people, reducing the platform's social graph density
D) Reported higher levels of satisfaction in follow-up surveys

11. The "pivot to video" strategy beginning in 2014 was driven by which combination of factors?

A) User demand for video content and cheaper production costs for Facebook
B) Higher revenue per impression from video advertising and internal data showing increased time-on-platform from video
C) Regulatory pressure to diversify content formats and competition from YouTube
D) Academic research showing video consumption was more beneficial for user wellbeing than text

12. A 2016 Wall Street Journal investigation found that Facebook had overstated average video viewing time by approximately what percentage range?

A) 10–20 percent
B) 30–40 percent
C) 60–80 percent
D) 100–120 percent

13. The BuzzFeed News analysis published in November 2016 found that in the final three months of the US election campaign, the top 20 fake news stories generated more Facebook engagement than:

A) The top 20 stories from television news networks
B) All content from verified political campaign accounts combined
C) The top 20 stories from 19 major news outlets combined
D) All organic posts from users without verified accounts

14. In January 2018, Mark Zuckerberg announced the "meaningful social interactions" (MSI) algorithm change, framing it as a response to research showing that which behavior was associated with negative wellbeing?

A) Excessive posting frequency
B) Passive consumption of social media content
C) Following accounts with more than 100,000 followers
D) Comparing one's post metrics to friends' post performance

15. After the MSI algorithm change was implemented in 2018, what did internal research find had actually increased in Facebook feeds?

A) Posts from close friends and family members
B) Charitable and civic content from community organizations
C) Politically divisive and emotionally provocative content from Pages
D) Long-form articles from established news publishers

16. According to internal documents disclosed in the Facebook Papers, the "angry" emoji reaction was weighted how many times more heavily than the "Like" reaction in Facebook's content amplification algorithm?

A) Twice as heavily
B) Three times as heavily
C) Five times as heavily
D) Ten times as heavily

17. The "Teen Mental Health Deep Dive" slide deck, prepared for Facebook's leadership and disclosed through the Haugen whistleblower release, found that what percentage of teen girls reported Instagram made them feel worse about their bodies when they already felt bad?

A) 12 percent
B) 22 percent
C) 32 percent
D) 42 percent

18. Frances Haugen disclosed internal Facebook documents to which regulatory body in addition to the Wall Street Journal?

A) The Federal Trade Commission (FTC)
B) The US Securities and Exchange Commission (SEC)
C) The Federal Communications Commission (FCC)
D) The Department of Justice (DOJ)

19. The chapter identifies a recurring pattern in Facebook's integrity vs. growth debates. Which of the following best describes that pattern?

A) Integrity researchers identified harms, modeled interventions, found interventions would reduce engagement, and interventions were not implemented
B) Product teams identified harms, proposed interventions, and implemented them after executive approval
C) Integrity researchers identified harms but consistently failed to produce credible models of intervention costs
D) Interventions were implemented and later reversed when they were found to reduce revenue below projections

20. The chapter's "structural argument" about Facebook's choices holds that the harms produced by the News Feed were:

A) The result of deliberate, malicious choices by specific executives who knew they were causing harm
B) Unavoidable under any conceivable business model for a social media platform
C) Predictable outputs of an engagement-optimization business model that any similar company would face
D) Primarily the result of technical limitations that prevented Facebook from measuring wellbeing directly

21. By 2022–2023, Facebook's pivot to Reels and AI recommendations represented what fundamental change to the nature of the News Feed?

A) A return to the chronological feed model that users had originally preferred
B) A shift from social-graph-based curation to entertainment-content-based algorithmic matching
C) A shift from advertising-supported revenue to subscription-based revenue
D) A shift from passive content delivery to interactive content creation tools

22. The chapter argues that the harms produced by engagement-optimization systems operating at the scale of billions of users are often:

A) Proportionally equivalent to the harms produced at smaller scales, simply multiplied
B) Emergent properties of scale that do not exist in smaller deployments of the same algorithm
C) Less severe at larger scales because the algorithm receives more data and becomes more accurate
D) Primarily experienced by the smallest and least-engaged segment of the user population

Chapter 24: Quiz

Facebook's News Feed: A Decade of Optimization Against Users

Answer Key