Assessment Validity and Reliability

1. Content validity

1.1. Systematic correlation to instructional objectives

2. Construct validity

2.1. Using a theory or rationale to account for correlation

3. Criterion-related validity

3.1. Correlate test scores with an external standard or obtain a numerical estimate of validity

3.1.1. Concurrant- Measurement collected at the same time

3.1.2. Predictive- measurement collected over a period of time

4. Validity

4.1. Measures what it is supposed to measue

4.2. Adequate validity evidence

4.3. Validity dependent on the strength of the validity coefficient

4.4. Consider the purpose of the test

4.5. Group variability affects the strength of the validity coefficient

5. Types of validity evidence

6. Internal consistancy

6.1. Estimates only single or unitary trait

6.1.1. Split-half- Data is divided into halfs and correlated with one another

6.1.2. Odd-even- Data is divided by odd and even numbers and correlated

6.1.3. Kuder-Ricardson procedure- determines the extent the assessment measures a single unit

6.2. Gives inflated results for speeded tests

7. Alternate-form estimates

7.1. Adminster two equivalent assessmnets to the same students

8. Test-retest

8.1. Administering the same test to the same students within a short interval

9. Reliability

9.1. Stability of the assessment over repeated administrations

9.2. As group variability icreases reliability decreases

9.3. As scoring reliability goes down, reliability goes down

9.4. The longer the test the more reliable

9.5. Reliability decreases if assessment is too easy or too hard

10. Methods of Estimating reliability