What you will be able to do
interpret quantities with denominator, scale, and uncertainty; detect misleading visual or statistical framing; verify news claims and distinguish misinformation from disinformation
Quantitative Data, News, and Misinformation
How can quantitative reasoning expose misleading news and support accurate public conclusions?
Activate what you know
A headline says incidents rose 50 percent. List the numbers and context needed before deciding whether the change is large or meaningful
- denominatortotal or comparison base used to calculate a rate or percentage
- ratequantity measured relative to another quantity, often time or population
- percentage pointarithmetic difference between two percentages
- correlationmeasured relationship in which variables vary together
- causationrelationship in which one factor helps produce change in another
- misinformationfalse or misleading information shared regardless of intent
- disinformationfalse or misleading information deliberately created or spread to deceive
- uncertaintyquantified or described limit on precision and confidence
1
Read behind the Number
A count answers how many; a rate relates the count to population, time, distance, or another base. Comparing cities by raw incident counts can mislead when populations differ. Convert to the same rate and period, then inspect whether reporting practices or definitions changed. Percentage change and percentage-point change differ. A rate rising from 20 percent to 25 percent rises five percentage points and 25 percent relative to the original rate. Headlines should state the form clearly. Samples introduce uncertainty. Ask how participants were selected, how many responded, whether groups were underrepresented, and what margin of error or range is reported. Missing data can bias a result, while a large sample cannot repair a systematically unrepresentative method
2
Verify Both the Number and the Narrative
News writers translate technical data into public claims. Trace a statistic to its original agency or study, reproduce simple calculations, and compare the headline’s wording with the measured result. Corrections and later revisions matter, especially for preliminary data. Record the publication date because a once-accurate claim may become outdated. Correlation does not by itself establish causation. If park access and health move together, income, age, neighborhood design, or reverse causation may contribute. Strong causal claims need time order, comparison, mechanism, and methods that address alternatives. Misinformation can spread through error or carelessness; disinformation includes deceptive intent, which requires evidence about creation or coordination. Do not guess intent from falsity alone. Correct responsibly by sharing the verified fact, source, date, and context without amplifying harmful imagery or personal data
Demand the Reference Class
Every higher, lower, doubled, rare, or majority claim needs a denominator, comparison, time window, and definition
Use Four Questions
Compared with what? Out of how many? Over what period? Measured how? Then trace the answer to the original source
Accurate Numbers Can Support a Misleading Story
Selection, denominator, axis, time window, or causal wording can distort meaning without changing the plotted values
Headline and Graph Audit
Recalculate a low-risk claim, redraw the graph with an honest scale, trace its source, and write a corrected headline with uncertainty
Quantitative news literacy checks counts, denominators, rates, periods, definitions, samples, uncertainty, displays, and causal language, then verifies original sources without assuming deceptive intent
Practice and answer guide
Work through all 12 questions. Open an answer only after you have written or explained your response.
- 1. A rate rises from 20 percent to 25 percent. Give the percentage-point and relative increases
Check answer
Five percentage points; 25 percent relative increase because 5 divided by 20 equals 0.25
- 2. City A has 200 incidents among 100,000 people; City B has 150 among 50,000. Which rate is higher?
Check answer
City A rate is 200 per 100,000; City B rate is 300 per 100,000, so City B is higher
- 3. Why are raw counts insufficient in that comparison?
Check answer
The populations differ; a common denominator is needed for a valid per-person comparison
- 4. Draw two bar graphs of values 90 and 100: one starting at zero and one at 85. Explain the visual effect
Check answer
Both plot the same data, but the 85 baseline visually exaggerates the difference. The explanation should distinguish accurate values from misleading emphasis
- 5. A poll samples only visitors to one campaign website. Identify the main validity problem
Check answer
Selection bias: visitors to one campaign site are unlikely to represent the full target population
- 6. Can a huge sample fix that problem? Explain
Check answer
No. More responses from the same systematically unrepresentative group increase precision around a biased sample, not representativeness
- 7. A headline says policy X caused a decline because both occurred in 2024. Name three missing causal tests
Check answer
Time order, comparison group, mechanism, alternative causes, prior trend, and a method addressing confounding; any three
- 8. Distinguish misinformation and disinformation
Check answer
Misinformation is false or misleading regardless of intent; disinformation is created or spread deliberately to deceive
- 9. Why should intent remain unresolved without evidence?
Check answer
Falsity can result from mistake, outdated information, satire, or careless copying; intent requires separate evidence
- 10. A report changed its category definition halfway through a trend. What should a responsible graph do?
Check answer
Mark the break, explain both definitions, avoid joining incomparable values as one seamless trend, or recalculate using a consistent definition
- 11. Rewrite: Crime doubles! The count rose from two to four in one month; the usual monthly range is one to six
Check answer
A better headline is: Monthly count rose from two to four but remained within the usual one-to-six range; more months are needed to assess a trend
- 12. Build a verification note for a fictional statistic, listing original source, calculation, denominator, period, uncertainty, and corrected claim
Check answer
Answers vary but must include all six components and keep the corrected claim within what the fictional method supports