Journal of Official Statistics, Vol.20, No.2, 2004. pp. 379–391
How Good is Good? Comparing Numerical Ratings of Response Options for Two Versions of the Self-Assessed Health Status Question
Barbara Foley Wilson, Barbara M. Altman, Karen Whitaker, Mario Callegaro
Abstract:Eight sets of disability questions were tested to assess their relative value for measuring impairments, functioning, and behavior. Two versions of a self-assessed health status question (SAHS) appeared among the eight sets. While both had 5-point response option scales containing Good and Very good, the rank order differed within the scales. A small experiment was conducted to explore whether Good and Very good meant the same thing to respondents when they were presented within the two different response option scales. Participants wrote each response option set on two Visual Analog Scales at different times. Good received a lower numerical rating (5.4 on a scale from 1 to 10) when it was the third option after Excellent and Very good than when it was the second option (7.3) after Very good. Findings are presented in the context of past research on SAHS questions and rating scales. The results are relevant for making cross-survey comparisons.
Keywords:Cognitive testing, disability measures
Copyright © Statistics Sweden 1996-2018. Open AccessISSN 0282-423XCreated and Maintained by OKS Group