1. Split Half (Odd Even)
1.1. Students take each half of test. Scores should correlate.
1.1.1. Importance to assessment -most frequently used as it only needs one test.
2. New node
3. Test - Retest (Stability) Evidence
3.1. Give test twice - correlation in scores each time
3.1.1. Importance to assessment - depends on time interval - If too long then scores unreliable
4. Alternate Forms Equivalence Evidence
4.1. Two equivalence tests given - scores should be the same for both.
4.1.1. Importance to assessmetn - If scores are different, test is unreliable. Standardized tests need this reliability.
5. Internal Consistency Evidence
5.1. Items correlated with each other in consistent manner - includes equal difficulty levels.
5.1.1. Importance to assessment - good for content evaluation - if one item correct others should be correct.
5.1.2. Importance to assessment - only use if similar items measure a single concept
5.1.3. Importance to assessment - achievement scores inflated due to speeded sections
6. Kudez - Richardson
7. Concurrent Criterion
7.1. Correlates test scores with a criterion measure at the same time.
7.1.1. Importance to assessment - measures two tests (an old and a new tests) for similar criterion rate.
7.1.2. Importance to assessment - if students rank the same on both tests then new test is valid.
8. Content Validity
8.1. compares test items to objectives
8.1.1. Importance to assessment - checks if what is taught is what is tested.
8.1.2. Importance to assessment - shows how well content taught has been mastered.
9. Criterion Related Validity
9.1. Correlates test scores with external standards
10. Predictive Validity
10.1. Predicts future behaviors
10.1.1. Importance to assessment - for aptitude tests, predictive results
10.1.1.1. New node
11. Construct Validity
11.1. Checks whether test results correspond with scores on other variables
11.1.1. Importance to assessment - results should correspond to what the teacher expects.