Validity & Reliability

1. Construct Validity

1.1. Test measures based on a theory or rationale

2. Content Validity

2.1. Test measures instructional objectives

3. Criterion Related Validity

3.1. Concurrent

3.1.1. New "better" test = high correlation with existing test

3.2. Predictive

3.2.1. New test predicts some future expected outcome

4. Test-Retest

4.1. 1 Test = Given Twice

4.1.1. Issue 1: Memory

4.1.1.1. Too little time in between

4.1.2. Issue 2: Change in Trait/Skill being measured

4.1.2.1. Too much time in between

5. Alternate Forms

5.1. 2 equivalent tests

5.1.1. Requires development of two tests

6. Internal Consistency

6.1. 1 test split into 2

6.1.1. odd-even reliability

6.1.2. Spearman-Brown prophecy

6.1.3. Kuder-Richardson

6.1.3.1. KR20

6.1.3.2. KR21

6.1.3.3. Coefficient Alpha