The developer of a new occupational interest test conducts a study to determine which test items accurately distinguish between people who are employed in different occupations. The test developer is likely using which of the following methods of scale construction?
- empirical criterion keying
- homogeneous keying
- factor analysis
- multiple regression analysis
The Correct Answer is A.
The test developer described in this question is trying to determine which items distinguish between specific criterion groups (i.e., people in different occupations). Empirical criterion keying involves choosing items for different scales on the basis of how well the items distinguish between different criterion groups. As an example, this method was used in the development of the Occupational scales of the Strong Interest Inventory.
Answer B: Homogeneous keying is used when items are selected on the basis of the degree to which they measure the same domain.
Answer C: When factor analysis is used for scale construction, items that correlate highly with the same factor are included in the same scale.
Answer D: Multiple regression analysis is a multivariate technique that is useful when performance on two or more predictors will be used estimate an individual's score on a criterion. It is not a method of scale construction.