Construct Validity


Health Economics

  • A construct that correlates well with other conventionally trusted measures of the underlying concept and that discriminates as one would expect (i.e. in ways predicted by theory) between cases having different characteristics is said to have construct validity. For example, a measure of ' health ' that had construct validity would correlate inversely with specific indicators of illness and incapacity. The validity of a construct might well vary according to the context in which it is being used.