Representation validity

From Wikipedia, the free encyclopedia

This article or section needs to be wikified to meet Wikipedia's quality standards. Please help improve this article with relevant internal links. (November 2007)

Representation validity is concerned about how well the constructs or abstractions translate into observable measures. There are two primary questions to be answered.

• Do the subconstructs properly define the construct (if you break up the main abstractions into

smaller abstractions or definitions)?

• Do the observations properly interpret, measure, or test the constructs?

One way to argue positively, albeit a very weak argument, is to claim face validity for the construct/observable relationship. Basically this is making the following claim: on the face of it, it seems like a good translation. The weakness of this argument can be strengthened by a consensus of experts. Another way to argue positively is to claim content validity for the construct/observable relationship. To do this one must check the operationalization against the relevant content domain for the construct: to extent to which the tests (ie, the observable measures) measure the content of the subject being tested — ie, that all the important content areas are covered adequately.