1. Validity
1.1. Content Validity
1.1.1. Does the test cover what it should?
1.1.1.1. Does the test match instructional objectives?
1.2. Criterion - Related Validity
1.2.1. Concurrent Criterion-Related Validity
1.2.1.1. Does this new criteria test perform like the old test?
1.2.1.1.1. What does the validity coefficients tell us?
1.2.1.1.2. Is the group being measured different, or are they alike?
1.2.1.1.3. Is the criteria relevant?
1.2.2. Predictive Validity
1.2.2.1. How well does the test predict future behavior?
1.2.2.1.1. What does the validity coefficients tell us?
1.2.2.1.2. Is the group being measured different, or are they alike?
1.2.2.1.3. Is the criteria relevant?
1.3. Construct Validity
1.3.1. Does the test use well-established methods of testing?
1.3.1.1. What does the validity coefficients tell us?
1.3.1.2. Is the group being measured different, or are they alike?
1.3.1.3. Is the criteria relevant?
2. Does the test measure what it should?
3. Reliability
3.1. Test - Retest
3.1.1. Give the test twice to see if the scores are basically the same.
3.1.1.1. How far apart were the testing times?
3.1.1.2. Has new information been obtained by those being tested again?
3.1.1.3. Are there clues that have reminded those being tested of how to get the right answer?
3.2. Alternate Form
3.2.1. Give two tests that are basically the same to two groups . If the scores are basically same, the test is reliable.
3.2.1.1. Can I create two sound, valid tests?
3.3. Internal Consistency
3.3.1. Do the students do as well on one part of the test as they do on another part of the test?
3.3.1.1. Split-Half (best way): Do students do as well on even numbered problems as they do on odd numbered problems?
3.3.1.2. Kuder-Richardson: how alike are the questions? Does the test utilize a consistent format?