Construct validity
Reliability, prediction and an appealing label answer different questions.
The question
When does a score justify an inference about a psychological attribute?
Cronbach and Meehl’s 1955 paper helped formulate construct validation as a continuing relation between measurement and theory. Its historical importance does not make every early example a modern standard.
What the claims mean
Operational definition
A procedure supplies an observable variable. It can be useful even when the psychological interpretation remains uncertain.
Construct interpretation
Calling that variable anxiety, intelligence or self-control requires a pattern of evidence connecting it to the proposed attribute and distinguishing alternatives.
Evidence to examine
Cronbach and Meehl distinguish criterion-oriented prediction from interpreting an unobserved construct. Their account asks how a network of theoretical relations supports a test interpretation, rather than treating a single correlation as final validation.
Shared ground and disagreement
What the accounts can share
A repeatable score can consistently measure the wrong thing. Prediction of an outcome is valuable but does not uniquely reveal the mechanism producing the prediction.
What remains at issue
Which competing construct interpretations fit a measure, whether the measure behaves similarly across groups, and how a theory should change when expected relations fail.
Limits and next tests
Constructs are not made real by giving a factor a name. Conversely, disagreement about a construct does not imply that every measurement is arbitrary or equally informative.
What would help distinguish the claims?
Compare multiple methods, competing explanations and predicted changes after intervention. Test whether the interpretation survives new populations and whether apparent convergence is merely shared wording or method.
Textbook connection
ContextualOpenStax Psychology 2e: 2.2 Approaches to Research
The section provides background vocabulary. It does not cover the full debate or establish the status of later research.
Sources and access
- Cronbach, L. J., & and Meehl, P. E. (1955). Construct validity in psychological tests. https://psychclassics.yorku.ca/Cronbach/construct.htmFull text inspected
York University transcription: validation distinctions, theoretical network and interpretation limits inspected.