bepress|Social and Behavioral Sciences|Psychology|Quantitative Psychology | bepress|Social and Behavioral Sciences|Psychology | bepress|Social and Behavioral Sciences
Separating the signal in a test from the irrelevant noise is a challenge for all measurement. Low test reliability limits test validity, attenuates important relationships, and can lead to regression artifacts. Multiple approaches to the assessment and improvement of reliability are discussed. The advantages and disadvantages of several different approaches to reliability are considered. Practical advice on how to assess reliability using open source software is provided.