Validity and Reliability: Used for Test Evaluation

1. Criterion-Related Validity

2. Content Validity

2.1. Do the questions for the test correspond to what should be covered by the test?

3. Validity asks the question, does a test measure what it is supposed to measure?

4. Concurrent Criterion Related Validity

4.1. Yields a correlation coefficient called validity coefficient.

4.2. Does the new test match performance on an existing test?

5. Predictive Validity

5.1. Reveals how well a test predicts future behaviors of examinees

5.2. Can the test predict a students success or failure fot the next grade level?

6. Construct Validity

6.1. If the test results correspond to what you would expect.

7. Reliability asks if the test receives the same type of score consistently.

8. Three Basic Methods for Estimating the Reliability of Test Scores

8.1. 1. Test-restest: test given twice and the correlation between the two tests is determined.

8.1.1. Problems: memory or experience involved the second time around.

8.2. 2. Alternate Forms: unreliable when there is a large difference between a student's test score on two different tests that are supposed to measure the same thing.

8.2.1. Problem: very tough to develop two good tests and time consuming

8.3. 3. Internal Consistency: items on the test should be correlated. Two types, split-half method and Kuder-Richardson method

8.3.1. split-half or odd-even: split the test in half and determine the correlation between the two halves, or put the even questions together and the odd questions together and determine their correlation.

8.3.2. Kuder-Richardson: "extent to which items within one form of the test have as much in common with one another as do the items in that one form with corresponding items in an equivalent form." (Kubiszyn & Borich, p. 344)

8.3.3. Problem: should only be used if the test is covering only a single concept.