Reasonable psychometric standards for self-report outcome measures in audiological rehabilitation

This brief review article addresses the quality of self-report rating scale outcome measures in relation to audiological rehabilitation with hearing aids. It is intended to assist those who may wish to evaluate, select, or adapt existing self-report measurement tools, or to develop new ones. The focus is not on specific scales but on the key issues in scale development and evaluation. A modern perspective is presented. Areas addressed include measurement goal definition, specification of the target population, the importance of conceptual frameworks, evaluation of reliability, validity and responsiveness, and production of norms. Reliability includes internal consistency and test-retest reliability concepts, related statistical measures such as the standard error of measurement, confidence intervals and critical differences, and some specific numerical criteria. Validity includes construct, content, face and criterion validities. Responsiveness includes some principles in the measurement of change, causes of poor responsiveness, reliability of change measures, and effect size. Concluding remarks touch on the practicality of self-report scales. It is emphasized that measurement goals and target population characteristics must be defined precisely, that existing measures should be evaluated carefully before undertaking new development, that the properties of any measure may depend strongly on its purpose and context of use, and that quantitative statistical criteria should guide the evaluation of measures as well as the design of experiments or clinical trials.