Comparing Validity and Reliability
Frequently reliability and validity are confused when individuals are considering tool characteristics. The following illustration may help us understand these concepts better.
|Reliable, not Valid: When measurements are consistent (clustered), but they don't hit the target they are reliable but not valid. This missing the target usually indicates that the concepts requires substantial rethinking.|
|Valid, not reliable: When measures are scattered widely around the target but they are not tightly clustered they valid, but not reliable. This wide spread indicates that the indicators are not focused on the core concepts.|
|Neither Valid, nor Reliable: When measures are scattered but not focused around a core concept they are neither valid nor reliable. This missing the target without clustering indicates that the entire tool needs to be rethought.|
|Both Valid and Reliable: The measurements are consistent and tightly focused around the core concept. This indicates that the tool is a solid measure of the concept.|
|Statistical measures of reliability
|Statistical measures of validity|