Construct validity
Construct validity asks whether a measure actually captures the thing it claims to capture. A questionnaire that reliably produces the same number every time can still measure nothing in particular, so reliability is not enough. The problem is sharpest in psychology because most of what it studies, such as anxiety, intelligence, or grit, cannot be observed directly and exists in the theory only as whatever the measure captures, which makes the reasoning circular unless the construct is tied to independent predictions. Structuralism collapsed on exactly this point: trained introspectors in different laboratories reported different contents for the same stimulus, and there was no way to decide who was right.
Why it matters
It is the first question to ask about any psychological finding, because if the measure is not measuring the construct, then nothing downstream of it means anything.
Also written: validity, operationalisation, operationalization
Where it comes up
Read about Construct validity on Wikipedia
Sources
- Construct validity Wikipedia
- Psychometrics Wikipedia
- Philosophy of Statistics Stanford Encyclopedia of Philosophy
