Ted & Takk handbook - construct validity: "what an instrument truly measures." (does it measure what it claims to measure - aka construct validity)
I guess theoretical Lx AJTs do in a sense lack construct validity.
"The degree to which the constructs under investigation are captured/measured; the degree to which inferences may be made about specific theoretical constructs on the basis of the measured outcomes" (p 298)
Kaplan - 3 things that scientists measure:
1 direct observables -- the color of an apple, somebody's answer on a questionnaire
1 indirect observables -- I thought sb was a man but she checked 'female' on the questionnaire?
3 - "constructs are theoretical creations based on observations but which cannot be observed directly or indirectly" (p 119)
BUT what is a concept vs a constuct??/
lesley andres never satisfactorily explained this either.