ABSTRACT
This chapter reviews reliability as a critical psychometric aspect of any score derived in clinical assessment. Knowing which forms of reliability, such as internal consistency, test-retest reliability, and inter-rater reliability/agreement are relevant to a specific instrument requires an appreciation of the nature of the instrument, the nature of the measured phenomenon, and the nature of reliability. Measurement error is associated with all clinical assessment strategies and the clinician should carefully consider the potential sources of measurement error when interpreting the scores obtained from an instrument. Statistical procedures (such as using the standard error of measurement) are available to assist in the interpretation of scores. Reliability evidence, and an informed appreciation of the meaning of this evidence, is essential if the clinician is to make scientifically-informed choices when selecting assessment instruments and interpreting the measures obtained from them. Clinicians should not presume that a high level of score reliability means that the score is a valid indicator of the construct the instrument was designed to measure. The chapter ends with recommendations for interpreting the results of reliability data of measures used in clinical assessment.
