Validity and Reliability Flashcards
Validity
The extent to which the test measures what it is designed to measure.
Face Validity
The extent to which the teat appears to be measuring what is it is designed to measure e.g. an IQ test is designed to test intelligence, it appears to do to (face validity) and would be a valid test if it accurately measures intelligence.
Construct Validity
the extent to which results support a network of research hypotheses based on the assumed characteristics of the theoretical psychological construct.
Predictive Validity
How well does the test predict future behaviour or ability e.g. does the test for scholastic aptitude and later exam results.
Convergent Validity
How much do the results from your test concur with those of another?
Internal Validity
The reported outcomes are the consequences of the relationship between the Iv and the DV, with no alternative explanations as all extraneous variables have been accounted for. I.e. the change in the DV is because of the change in the IV - e.g. if you conclude that exercise reduces heart disease, you must be CERTAIN this is the cause and not something else.
External Validity
The extent to which the results can be generalized from the sample to the general population or from an experimental setting to the real world. Does the relationship between the IV and DV hold outside of the experimental setting?
Reliability
The extent to which a test provides a repeatable measure of what is being measured. E.g. does a personality test yield a similar result every time I sit it?
Inter-rater Reliability
the degree of agreement amongst raters. A score of how much consensus there is between judges.
Test-retest Reliability
Doig the test twice and seeing to what extent the score correlate - how similar are the score?m
Split-half Reliability
after the participants sat the test, arbitrary splits the test into two equal halves. How similar are the results in both halves?
Cronbachs Alpha
the expected correlation of two tests that measure the same construct. Cronbachs alpha is a function of the number of items in a test, the average co-variance between item-pairs, and variance of the total score.
