AP Psych- Types of Validity and Reliability
Internal consistency reliability
Asses the consistency of results across items within a test.
Internal consistency reliability example
Surveys often ask the same questions several different times.
Split-half reliability
Compare one part of the measure to another part. The subject should receive similar scores.
Split-half reliability example
Compare the number of odd questions you answered correctly to the number of even questions you answered correctly on any test.
Parallel forms reliability
Compare the results of different forms of the same test measuring the same thing.
External validity example
Effective teaching strategies utilized by Coach Hart may not translate outside of her classroom.
Construct validity
Extent to which a measure captures the concept it is designed to measure.
External validity
Extent to which a study can be generalized beyond the sample.
Construct validity example
In developing a new IQ test, you must operationally define intelligence so that the measure assesses your definition of intelligence.
Interrater reliability
Independent raters will evaluate something and their evaluations will be similar.
Predictive validity example
The GRE predicts graduate level GPA
Internal validity example
When the FDA runs drug trials, they attempt to eliminate all other variables to determine if the drug itself caused the intended effect.
Interrater reliability example
When you evaluated free writes, 3 scorers should have scored the same free write similarly.
Test-retest reliability example
You take a test on Monday. Then, you take the same test on Tuesday. (Doesn't Dr. G do this?)
Concurrent validity example
You take the Big Five Inventory and the Myers Briggs Inventory to assess your levels of in/extroversion. They should reveal similar levels.
Face validity example
You would expect the questions on the AP psychology exam to be typed.
Face validity
Degree to which the measure looks like it's supposed to.
Concurrent validity
Degree to which the score on one measure correlates to the score of another measure given at the same time.
Internal validity
Degree to which we are confident that the changes in the dependent variable were produced solely by the independent variable.
Parallel forms reliability example
I give one version of the quiz to period 1 and a second version of the quiz to period 2. There should be a similar range of scores.
Predictive validity
Degree to which a measure is correlated to a later outcome
Test-retest reliability
Give one test. Then, give the same test again. The subject should receive similar scores.