Validity and Reliability

1. Test - Retest

1.1. Same test administered twice to same group of students

1.2. Scores remaining consistent indicate stability of test

2. Alternate tests

2.1. Different forms of the same test administered

2.2. Tests results correlated between two tests

2.3. Important to make sure that both tests are equivilant to each other in what they are testing

3. Internal Consistency

3.1. Split Half Method

3.2. Tests should be divided with equal difficulty on both halves

3.3. Two halves of the same test are compared for consistency often with Kuder Richardson Methods

4. Common Reliability Issues

4.1. Variations in the group size affects the reliability coefficient

4.2. Grading must be reliable, if tests are graded incorrectly or differently, reliability is comprimised

4.3. More testing items makes reliability more accurate

4.4. Making tests too difficult or easy decreases reliability

5. Concurrent Criterion

5.1. Compares a known valid test with one designed to measure the same outcomes

5.2. Does not require a time span.

6. Content Validity

6.1. Does the test measure what it says it will?

6.2. Is the test directed at the ability of the student?

6.3. Does the test measure instructional objectives?

7. Constuct Validity

7.1. Relationship to other available information corresponds well to an estastablish theory

8. Predictive Criterion

8.1. Does the test predict future accomplishment?

8.2. Does scoring well on this test for most students mean they will do well in a future endeavor?

8.3. Always requires a length of time