Validity and ReliabilityHow Can You Tell?Validity and reliability refer to the accuracy and repeatability of tests and instruments. Both are key issues in research and assessment. There are ways to show evidence of them before you conduct your study or administer a test. Showing ValidityValidity is the extent to which a test or instrument measures what it is supposed to measure. How can you demonstrate it? There are several ways:Construct validity--does the test accurate measure the quality or trait of interest? A panel of experts can rate test items to determine the extent to which the theoretical concept (e.g., leadership, anxiety, fitness)is represented. Content validity is the extent to which items represent the content of information of interest. Again, a panel of experts rate items to determine how appropriate they are. Concurrent and predictive validity both use correlations to demonstrate the accuracy of an instrument. Concurrent validity is demonstrated when the results of one test are highly related to another that measures the same thing (e.g., correlating two different intelligence tests). Predictive validity uses a current measure to predict future performance (e.g., performance on the LSAT as a predictor of success in law school). Showing ReliabilityRepeatability is an essential element of validity. There are several ways to demonstrate that an instrument yields consistent results by using correlations.Test-retest reliability means that you can administer the test to the same group twice and come up with very similar results. Alternate forms are two very similar and equivalent versions of the same test (form A and form B). They are both administered to two different groups to demonstrate stability and equivalence. Split Half reliability uses one form of a test and breaks the items into two groups, such as odd-even, and scores them separately to determine consistency. Top of Validity and Reliability
|
