Developing and Testing the MOS 20-Item Short-Form Health Survey

A General Population Application

Published in: Measuring Functioning and Well-Being: The Medical Outcomes Study Approach / edited by Anita L. Stewart and John E. Ware, Jr., (Durham, N.C.: Duke University Press, 1992), Chapter 16, p. 277-290

Posted on RAND.org on January 01, 1992

by John E. Ware, Cathy D. Sherbourne, Allyson Ross Davies

When questionnaire length must be limited, investigators must choose between breadth (comprehensiveness and content validity) and depth (precision) of measurement. The short-form survey represents only the most important health concepts and included 17 items adapted from longer surveys used successfully in the rand Health Insurance Experiment (HIE): physical functioning (6), limitations in role performance due to poor health (2), general mental health (5), and current health perceptions (4). Social functioning was not included. Louis Harris and Associates included the 17-item short form in a 100-item telephone interview and administered it toward the end of the interview. It took 3-4 minutes to complete. We hypothesized that if the new short form is valid, (1) older persons should report poorer physical and role functioning and perceive health less favorably, but rate mental health more positively, (2) men should rate their current health more favorably and report better mental health than women, and (3) education and income should be positively correlated with all four scales. The pattern of results corresponds with the MOS hypothesis. The reliability estimates exceed conventional standards for group comparisons. There is a high correlation between the current health scale and the single-item rating of health. Administration times represent a reduction in respondent burden of approximately 90 percent relative to the full-length versions from which scales are derived. Limitations in social functioning due to poor health and pain are measures that were omitted that the MOS recommends be included.

This report is part of the RAND Corporation external publication series. Many RAND studies are published in peer-reviewed scholarly journals, as chapters in commercial books, or as documents published by other organizations.

The RAND Corporation is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.