Test Reliability And Validity

Dan Eble

PSYC 105 G

Dr. Waterstreet

Test Reliability and Validity

Correlations are important in establishing test reliability and related test validity. Test reliability has four methods of testing. The first is the test-retest method, which is when the same test is given to students at two different occasions and the scores are then correlated for production. The second is the alternate form test, which is when two forms of the same test is given to students and the scores are then correlated. The third is internal consistency method, which is given to determine if the test answers given by the students are uniform with each other. The fourth is the inter scorer reliability, which determines if the students have some sort of agreement in their answers, usually used in essay testing.

Validity is also an important issue in test correlation. Validity concerns what a test measures and how well it does so. The test is only valid if it has an applicable purpose in the course. Content is a major portion of proving validity. You must look at the material and see if the content covers interest in the subject matter and if it is applicable to the students. Criterion related validity has two types, concurrent and predictive validity, which are both based on correlation. Concurrent validity reflects a standard of the test results from students. Predictive validity is when a test is given to high school seniors to later be correlated with their college GPA.

After homework number two I have a better vision of correlation. I feel that after taking test number two that I will have a positive correlation. I did not do all to well on my first test, but I feel that I know what to expect from the next test.