Measurement Theory in Action: Case Studies and Exercises comprises twenty modules, each corresponding to entire chapters in typical measurement theory texts. The modules begin with introductory concepts and a review of statistics; progress through conceptions of content, criterion-related, and construct validation, validity generalization and test-bias; and conclude with more advanced topics such as multiple regression and item response theory (IRT). Each module is composed of an overview, case studies, exercises, Internet references, and suggested further readings. An extensive glossary of key terms is also provided for quick reference. An Instructor's Resources CD containing exercises and assignments is also available.
In the two previous modules examining validation, we discussed the role of expert judgment and the use of a correlation between test scores and a relevant criterion as methods for providing evidence of the meaningfulness of test scores. Cronbach and Meehl (1955) discussed a third validation strategy that they felt was best suited to the many instances when test scores assess an attribute or a quality (i.e., a construct) that is not readily operationalized. Their arguments were so powerful that today psychometricians view all validation efforts as evidence regarding construct validation.
“Psychology works with crude, half-explicit formulations” (Cronbach & Meehl, 1955, p. 294). Although this is an astonishing declaration, Cronbach and Meehl used this assertion not to urge the abandonment of psychological inquiry, ...