Ch 5: Reliability of Measures
Importance of Reliability
Measures that are not reliable cannot be used to study variables or relationships between variables. Any data gained from an unreliable measure would have to be thrown out, so it is a waste of time and money to use unreliable measures.
Reliability can also be defined by using two components:
True Score: the "real" score on the variable (how the person actually is) Measurement Error due to a variety of things such as whether the statements are clear, the reading ability, the mood, fatigue, etc. of the person taking the test.
Reliability can be achieved by
careful test (such as a personality test) or survey construction and administration of the test or survey.
Reliability
is the consistency or stability of a measure of behavior.
Training of all researchers, assistants, or therapists is important so that
the measurements are being given in the same way by all.
Reliable measures give the same result
when measuring the same subject.