Selection Effects

mental-model

Strong outcomes do not by themselves prove institutional effectiveness. When an institution admits exceptional people, its graduates may succeed largely because of who entered, not because of what happened inside.

Take the top 2,000 people from a country of 1.4 billion and send them to one college. Some will become extraordinary even if the college contributes almost nothing.

E1

The intake hides the value added

Observed success combines at least two forces: the qualities people bring in and the value the institution adds. Selective admission loads the first variable so heavily that impressive alumni can look like evidence for the second. The institution’s reputation then grows from outcomes it may only partly have caused. Selectivity itself becomes a signal, while the ceremony of choosing winners can produce an illusion that the process identifies future success with precision.

E1

Where it shows up

The elite-college result

A college drawing 2,000 exceptional people from 1.4 billion starts with an extreme sample. Judging its teaching from those students’ later success is therefore a filtered comparison—not a clean estimate of institutional impact. Looking only inside that admitted group can also invite distortions related to Berkson’s paradox.

E1

Selection is not the whole explanation

This model warns against attributing every strong outcome to the institution; it does not prove that the institution adds no value. With only successful alumni in view, you cannot separate selection from teaching, networks, resources, or the optionality the institution may create.

E1

Ask for the missing comparison

Before paying for prestige, compare outcomes for similarly strong people who were admitted versus those who narrowly missed admission or chose another path. That comparison is far more informative than a list of famous alumni.

Episodes that teach this