The target of the attack was the institution. The institution carried the individual's identity. The identity defended itself.
Berkson's paradox is a statistical phenomenon where a spurious negative association appears between two variables when conditioning on a common effect or selection criterion.
A Scene Worth Recognising
Imagine a sales team where representatives earn bonuses either by making a high volume of calls or by demonstrating deep product knowledge. Management looks only at the reps who hit their quarterly bonus threshold and notices that, within this group, those who make many calls tend to have lower productāknowledge scores, and viceāversa. The observed negative link appears because the bonus threshold acts as a common effect: a rep can reach it through either strength, so selecting only those who succeeded hides the underlying independence of the two skills.
What it means and how it works
The mechanism is collider bias: conditioning on a variable that is influenced by two independent variables (the collider) induces a statistical dependence between them, even if they are independent in the underlying population.
When two independent causes both increase the likelihood of a certain outcome (a collider), selecting only cases where the outcome occurs can create an apparent negative correlation between the causes. This arises because, within the selected subset, the presence of one cause reduces the relative need for the other to explain the outcome, leading to a misleading inverse relationship.
Why it matters
Understanding Berkson's paradox prevents misinterpretation of associations in stratified or selected data, which is common in clinical studies, epidemiology, and any analysis that conditions on a shared outcome or selection rule.
The verified research on this pattern supports the following:
- Berkson's paradox arises when conditioning on a collider variable induces a negative correlation between two independent causes.
Common misunderstandings
Misunderstanding 1: People often mistake the induced negative correlation for a causal inhibitory effect, or they overlook that the paradox is purely a artifact of selection rather than a genuine relationship.
See Berkson's paradox in everyday decisions
Pick a life context to see how this bias can show up outside the textbook.
When Help-Seeking Skews Screen-Time Findings
A researcher studying a wellness app notices that among users who reported high anxiety (score 8 or higher on a 0-10 scale), more screen time appears linked to lower anxiety, but this pattern disappears when all users are considered.
Scenario
Alex, a public-health graduate student, pulls data from a campus mindfulness app. He narrows his focus to the 200 students who completed an anxiety screening and scored 8 or above. Within this group, he observes that students who spend more than three hours a day on social media tend to report lower anxiety scores than those who spend less time. He wonders whether moderate social-media use might ease anxiety. Later, Alex examines the full dataset of all 2,000 app users. The negative link vanishes; screen time and anxiety scores show no relationship. The spurious negative association arose because he conditioned on high anxiety-a collider that both screen time and stress-reduction behaviors influence.
Where The Bias Enters
Conditioning on high anxiety (the collider) creates a statistical dependence between screen time and anxiety, making them appear negatively related even though they are independent in the full population.
Decision Check
Before drawing conclusions, verify whether the analysis is limited to a subgroup defined by a common outcome; if so, examine the full population or apply selection-adjustment methods.
This pilot example is illustrative and review-gated. It is designed to explain the pattern, not to claim a documented public case.
Sources
- Niroula, Rishab. REV 2.0 Topic Catalog. Hello to Halo.
- Kahneman, Daniel. Thinking, Fast and Slow. Farrar, Straus and Giroux, 2011.
- Gilovich, Thomas. How We Know What Isn't So. Free Press, 1991.
- Shermer, Michael. The Believing Brain. Times Books, 2011.
The next time this pattern surfaces, the move is not to fight it ā it is to notice it. Naming Berkson's paradox creates a moment of pause before the decision. That moment is usually enough.