Considerations in design and use of scales in rehabilitative audiology.

Self-assessment questionnaires in rehabilitative audiology typically possess the measurement characteristics of scales based on forced-choice and rank-order. Data from these ordinal scales neither support arithmetic analysis nor arithmetic interpretation. However, powerful nonparametric statistics are available for application to data from ordinal level questionnaires. Thus, methodologic principles influence the way self-assessment scales are designed, constructed, and analyzed. They also influence the way clinicians interpret responses from individual patients. Within these constraints, well constructed questionnaires can provide information on a wide range of topics in rehabilitative audiology. Suggestions are given for generating interval level questionnaires that are more precise than existing, ordinal level scales.