Chapter 26 Exercises

Exercise 26.1 — Typology Application (Individual, 30 minutes)

Below are ten examples of political content. For each, apply Wardle and Derakhshan's seven-type typology (satire/parody, misleading content, imposter content, fabricated content, false connection, false context, manipulated content) and explain your reasoning. Where the content falls into more than one category, identify the primary type and explain any secondary classification.

Content examples:

1. A Facebook post shares a 2014 photograph of a large crowd at a protest, labeling it as a rally that occurred last week against a congressional candidate.

2. A website mimics the visual design of the Associated Press and publishes a fabricated story claiming a senate candidate was indicted for tax fraud.

3. A video ad for a candidate accurately shows their opponent's voting record on a budget bill, but frames the votes as "cutting Social Security" — a characterization disputed by policy analysts.

4. A satirical website publishes an article headlined "Governor Announces Plan to Ban All Cars," clearly labeled as satire at the bottom of the page; the article is screenshotted and shared without the satirical label.

5. A tweet shows a real interview clip of a candidate discussing immigration, but the audio has been selectively edited to remove a qualifying clause.

For each: (a) Identify the type, (b) Explain why it fits that type, (c) Identify what verification approach would be most appropriate, (d) Estimate the difficulty of correction.

Exercise 26.2 — Fact-Check Methodology (Pair, 45 minutes)

Working with a partner, select a claim from a current political campaign that has been fact-checked by at least one major organization (PolitiFact, FactCheck.org, or Washington Post Fact Checker). Then:

Part A: Replicate the Fact-Check 1. Identify the exact claim, speaker, date, and context 2. Locate all primary sources relevant to the claim 3. Reach your own rating on the claim (using PolitiFact's 6-point scale) 4. Compare your rating to the published fact-check and document any differences

Part B: Evaluate the Fact-Check Design Using the best-practices criteria for correction message design (leads with true information, provides causal narrative, comes from credible source, is timely, explicitly flags the correction), rate the published fact-check on each criterion. What did they do well? What would you change?

Part C: Assess Reach Using CrowdTangle, social media search tools, or simple link tracking, estimate what proportion of people who saw the original claim are likely to have seen the published fact-check. What explains the gap?

Exercise 26.3 — Network Mapping (Small Group, 60 minutes)

Choose a piece of misinformation from the last six months that has been documented by a fact-checking organization or academic study. Using publicly available social media data, map the diffusion of the false claim.

Step 1: Origin Identification Using social media search, attempt to identify the earliest verifiable appearance of the claim. Document: date, account, follower count, and what network the account appears to belong to.

Step 2: Amplification Nodes Identify the five to ten accounts that drove the most amplification of the claim (highest follower counts, most shares/retweets). Classify each as: political elite, media figure, influencer, hyperpartisan media outlet, or ordinary user.

Step 3: Timeline Reconstruction Build a timeline showing the claim's spread over the first 24 hours. At what points did major amplification occur? Was there an elite cue that dramatically accelerated spread?

Step 4: Correction Diffusion Identify when the first correction appeared. Map the correction's spread using the same methodology. Calculate the ratio of false claim reach to correction reach.

Deliverable: A written brief (600–800 words) summarizing your findings, including a timeline graphic and a brief discussion of what intervention might have reduced the claim's reach.

Exercise 26.4 — Inoculation Message Design (Individual, 45 minutes)

You are working for a civic organization preparing voters for an upcoming election in which misinformation about mail-in ballot security is expected to be prevalent (based on historical patterns in your state).

Design an inoculation intervention using the principles of inoculation theory:

Step 1: Identify three specific false claims about mail-in ballot security that have previously circulated (document your sources).

Step 2: For each false claim, write an inoculation message that: - Warns the recipient that they will encounter a misleading argument - Presents a weakened version of the false argument (the "rhetorical virus") - Provides a clear refutation with specific evidence - Does not inadvertently amplify the false claim through excessive repetition

Step 3: Write a brief (400–500 words) justifying your message design choices with reference to the inoculation theory literature.

Step 4: Propose a measurement design: how would you test whether your inoculation messages actually reduced susceptibility to the false claims? What would your control condition look like? What outcome measure would you use?

Exercise 26.5 — The Meridian Mischaracterization (Case Analysis, Individual, 30 minutes)

Meridian Research Group's poll showed Garza 47%, Whitfield 46%, undecided 7%, with a ±3-point margin of error. Social media posts characterize this as Whitfield "surging" into the lead.

1. Is the claim that Whitfield is "leading" technically false? Is it misleading? Use Wardle's typology to classify the mischaracterization and explain how within-margin-of-error results should be communicated.

2. Write a 280-character (Twitter-length) correction that Vivian Park could use. Apply the best-practices criteria for correction design.

3. Vivian has three options: (a) publish a full correction blog post, (b) issue a press release to journalists covering the race, (c) respond directly to the accounts mischaracterizing the poll. Evaluate the likely reach and effectiveness of each approach for different audience segments.

4. Carlos Mendez's tracking dashboard found that 62% of poll mentions that week characterized Whitfield as leading. What is the policy implication of this for how pollsters should communicate results? Should Meridian change its reporting format? Discuss the trade-offs.

Exercise 26.6 — Platform Policy Analysis (Research and Writing, 60 minutes)

Select one major social media platform (Meta/Facebook, X/Twitter, YouTube, or TikTok) and analyze its current political content moderation policy on misinformation.

Your analysis should cover: 1. What the platform's stated policy is (cite the specific policy document, with date) 2. What enforcement mechanisms exist (removal, labeling, demotion, friction) 3. What research evidence exists on the policy's effectiveness 4. What exceptions or carve-outs exist (particularly for political advertising) 5. What changes in policy have occurred in the past two years and what drove them 6. What critics from the left and right have said about the policy

Write a 700–900-word policy brief evaluating the platform's approach and recommending three specific changes. Support each recommendation with evidence from the academic literature or credible investigative journalism.

Quantitative Exercise 26.7 — Calculating the Correction Gap (Excel/Spreadsheet, 30 minutes)

ODA tracked the following data for the Garza-Whitfield false claim:

1. Calculate the correction-to-exposure ratio at each time point (correction reach / false claim reach).
2. Plot the cumulative reach curves for the false claim and the correction on the same graph.
3. If ODA had been able to publish the correction at Hour 2 instead of Hour 4, and if the correction's diffusion rate had been identical to what was observed starting from Hour 4, estimate the correction-to-exposure ratio at Hour 24.
4. What does your analysis suggest about the value of speed in fact-checking? What are the trade-offs between speed and thoroughness?