Regression Fallacy
The regression fallacy credits an intervention for a change without accounting for regression to the mean. After an unusually extreme measurement, a later measurement may be closer to the usual level even if the intervention had no effect.
THE IDEA
Regression Fallacy
The regression fallacy attributes a change to an intervention without considering that an unusually extreme observation may be followed by one closer to the usual level. It is related to post hoc reasoning. The supplied passage omits its logical form and original example statements.
Example #1: Headache
The source discusses a headache that improves after an intervention. Symptoms can fluctuate or resolve naturally, but the account does not establish whether the treatment contributed. The quoted “mean” should refer to a typical measured level, not automatically to complete absence of pain.
Example #2: Pain after surgery
The source discusses pain subsiding after prayer. Healing and symptom fluctuations are alternative explanations, but this anecdote alone cannot determine the cause. Natural recovery over time is not necessarily statistical regression to the mean.
To assess a proposed “cause”, compare the observed change with what could have happened without it. Repeated measurements and an appropriate comparison group can help distinguish an effect from fluctuations or recovery.
When to use it
Evaluating the effectiveness of a new drug when patients naturally fluctuate in symptom severity.
Interpreting a student's test score increase after a tutoring program—considering regression from an unusually low score.
Assessing a company's sales spike after a marketing campaign when sales naturally fluctuate.
Judging a golfer's improved performance after switching clubs, accounting for regression to the mean after a poor round.
FROM THEORY TO EVERYDAY LIFE
In practice
Headache and Medicine
Someone takes medicine during an unusually severe headache and later feels better. The sequence alone cannot show how much improvement came from treatment, natural recovery, or fluctuation from an extreme symptom level.
Pain and Prayer
Pain after surgery subsides after family members pray. Attributing the change to prayer solely from the sequence overlooks healing and symptom variation. The account alone also cannot establish that regression to the mean explains the change.
A MOMENT TO REFLECT
Put it to work.
Think of a time when an intervention was followed by improvement. Could natural regression to the mean explain the outcome?
What are common situations in sports or business where the regression fallacy might lead to incorrect conclusions about performance?
How can we design experiments or observations to distinguish true causation from regression effects?
Often confused with
CONTINUE THE CONVERSATION
Source texts
- Logically Fallacious - The Ultimate Collection of Over 300 Logical Fallacies