Assessment Tests must be both valid and reliable

Assessment: Tests must be both valid and reliable. Validity: the extent to which a test measures what it’s supposed to. For example, the SAT says it measures “reasoning”… does it? Reliability: the extent to which a test yields consistent results. If Psych 4 gets an A average, while the other classes get Cs… what’s going on? Most tests are “standardized; ” what does this mean?

AP: Different types of Validity • Predictive: how successfully does the test predict the behaviour it’s designed to predict? – Example: how well does the SAT correlate with college performance? • Content: how successfully does the test assess what it thinks it’s assessing?

There is a difference between aptitude tests and achievement tests. What is the difference? • Aptitude: a person’s capacity to learn. These tests predict future performance! • Achievement: these tests assess learned knowledge! QUESTION: What kind of test is the SAT? MCAS? Midterms? ASVAB?

Results: standardized test results are compared using frequency distribution graphs, such as scatterplots and distribution curves. WAIS vs IQ

Most people achieve in the yellow area. Roughly the same number of people are either slightly above or slightly below the average. A few people are truly unusual, either in their excellence or their lack of excellence. IQ Results, Ohio, 1976

When I evaluate your tests, I just tip the same distribution curve on its side.