Validity and Reliability

1. Content Validity Evidence

1.1. EASY!

1.1.1. The minimum for any test, is the content on there what you want to be there?

2. Criterion Validity Evidence

2.1. Predictive Validity Evidence

2.1.1. Use of one test to estimate how well the student will do in the future. Example: SAT scores used as a predictor for College aptitude.

2.2. Concurrent

2.2.1. Two tests at once. One established test is given at the same time as an experimental test and the results of both tests are evaluated together. Example: IQ Test given at the same time with hope that higher IQ scores better on actual test.

2.3. Numerical indication of validity

3. Construct Validity Evidence

3.1. Other information backs up test results. If a class does poorly on a test then spends time going over the information and does well, or if one subgroup does better based on past experiences.

4. Test-Retest/Stability

4.1. Testing followed by testing again with the same exercise without any additional training in between.

5. Alternate Forms/Equivalence

5.1. Using a different test form and expecting siliar scores. Seen in large scale standardized testing to reduce the chance of cheating.

6. Internal Consistency

6.1. When a test measures one concept it is assumed that if a test taker gets one question right, they will get similiar questions correct.

6.1.1. Split Halves Method: Single test split into two parts (Odd-Even method)

6.1.2. Kuder Richardson methods: Big equation. Ver.20 is more reliable but 21 is easier to perform.