Though measurement tools may already exist, some researchers may decide to create new ones. In doing so, researchers must be concerned with criterion validity. Discuss criterion validity and the role of concurrent validity and predictive validity in achieving criterion validity. Use an example to show you understand the concepts.

Question

What will be an ideal response?

Answer 1

Ans: Criterion validity refers to those cases when we are not satisfied with the way the construct has been measured. The measurement tool may have some flaws or it does not apply to our specific population so we decide to create a new measurement for a construct that has already been measured. In that case, we need to be clear that our new measurement tool will be compared to the old measurement tool, so we know it has high criterion validity. Concurrent validity requires that the new test works as well as the old test. Therefore, comparing the new test with the established measurement tool and examining the matching results are compulsory. Predictive validity is achieved after establishing concurrent validity; the new test is then tested a few times with different population samples. If we can claim concurrent validity with one sample, we need a number of groups of participants to test both tools multiple times. If our results over time still show that the new mathematics test is better than the old test, it establishes high predictive validity.