Validity and Reliability

1. Criterion-Related Validity

1.1. Correlation of scores to established external standards

1.2. Concurrent Criterion-Related Validity

1.2.1. Scores from new test and a test with verified validity correlated

1.3. Predictive Criterion-Related Validity

1.3.1. Scores from initial test and subsequent test compared to predict behavior

2. Construct Validity

2.1. Test results consistent with expectations

3. Content Validity

3.1. Test items match instructional objects

4. Validity

4.1. Assessment measures what it is supposed to measure

5. Importance of Validity & Reliability

5.1. Test must be valid and reliable to have usefulness

5.2. Increase in score accuracy when test is valid and reliable

5.3. More informed decisions regarding student learning

5.4. Accurate results allow for proper instructional modifications and adjustments to assessment

6. Internal Consistency

6.1. Correlation of test items corresponding to same concept within the test

6.2. Split Halves

6.2.1. Correlation determined after test split into two equal parts

6.3. Kuder-Richardson Method

6.3.1. Consistency among the measures of test items within the same test

7. Alternative or Equivalence Reliability

7.1. Correlated scores from two equal versions of a test administered

8. Reliability

8.1. Assessment produces similar scoring rankings on multiple occasions

9. Reference Kubiszyn, T. & Borich, G. (2010). Educational testing & measurement: Classroom application and practice (9th ed.). John Wiley & Sons, Inc., Hoboken, NG.

10. Test-retest Reliability

10.1. Comparison of scores from same test administered on two separate occasions