Find the Missing Case: Put the Claimed Arrow Under Pressure
“Why doesn’t this example prove the cause?” The fastest answer is a case the argument cannot explain.
First, state the argument’s causal claim:
Cause → Effect
Then search in two directions:
- Cause and not Effect: the alleged cause occurred, but the predicted result did not.
- Effect and not Cause: the result occurred even though the alleged cause was absent.
These are not automatically decisive. One exception may barely affect a claim such as “usually causes.” Ask whether the exception is large, comparable, and reliably recorded.
For a claim that a medication prevents migraines, these are the pressure points:
- Many patients took the medication and still suffered migraines.
- Many patients suffered migraines without taking the medication.
The first attacks the medication’s ability to prevent the effect; the second shows the effect can occur without the cause. If the argument says the medication eliminates migraines, even one well-established counterexample may be devastating. If it says the medication reduces migraines, look for a meaningful pattern, not merely an exception.
When evaluating an answer choice, run three checks:
- Does it identify the cause and effect correctly?
- Does it supply the missing combination-cause with no effect, or effect with no cause?
- Is the case large or representative enough to matter?
A choice saying “some people disagreed” is not enough. One identifying many comparable failures pressures the causal claim.
Audit the Numbers: A Rate Can Be Correctly Calculated and Still Mislead
When an answer attacks data, audit what was counted before judging the conclusion.
Check the unit first. Are the numbers people, accidents, households, purchases, or accidents per driver? Then check the denominator. A raw count can fall while the rate rises if the population shrinks. Next check whether the groups were measured in the same way: same definitions, time period, reporting rules, and inclusion criteria.
Also inspect collection. A study can exclude dropouts or people who never returned for measurement. It can rely on self-reports for one group and official records for another. It can report an average that conceals a different distribution. These defects matter because the apparent effect may come from measurement rather than the alleged cause.
Use this sequence:
- Write down the comparison the argument uses.
- Name its numerator and denominator.
- Ask whether the groups contain the same kinds of cases.
- Ask whether “success,” “failure,” or the effect was recorded consistently.
- Recalculate or reframe the comparison if necessary.
A bad statistic weakens most when correcting the defect removes the apparent improvement or reverses it. A choice saying the sample was “small” may tempt you, but size alone is not fatal if the sample is representative and the pattern is large. A choice showing relevant cases were counted differently attacks the evidence at its foundation.
One Audit in Action: Reject the Attractive Percentage
A delivery company required its drivers to complete a new safety course. In the year before the course, the company recorded 12 crashes among 100 drivers. In the year after the course, it recorded 10 crashes among 100 drivers who had completed the course. The company concluded that the course reduced crashes.
Practice
Which one of the following, if true, most seriously weakens the company’s conclusion?
High-Yield Points - ⚡ Biggest Takeaways
Search for cause-without-effect or effect-without-cause cases. Audit the numerator, denominator, definitions, and included cases before trusting a statistic. Prefer a data flaw that removes the apparent difference over a vague alternative explanation.
Unlock the full lesson and continue reading
Signup to continue reading this lesson and unlimited access questions, flashcards, AI notes, and more