Health Status Measurement, Construct Validity

Just as it is important to ascertain that a measurement instrument produces reliable results across different situations, it is crucial to assess whether it measures what it is intended to measure—its validity. In the area of health status measurement, construct validation of measurement instruments underlies sound medical decision making. Validity is established through a process involving a series of experiments designed to test various relevant hypotheses about the structure and nature of the construct and its logical manifestations. The results of these experiments inform the level of confidence with which researchers make conclusions about the persons under study and the interpretation of instrument scores.

Early measurement of health focused heavily on disease and mortality rates for populations and on clinical variables representing disease activity for individuals. Over time, with the mounting challenges presented by chronic diseases and disorders, many health interventions have focused more on levels of physical, mental, and social functioning than on length of life. Under these circumstances, sound medical decision making about the value of healthcare interventions depends increasingly on the validity of instruments for measuring health status. The impact of validity on interpretation of clinical trials has been demonstrated empirically in psychiatry, with evidence that clinical trials using unvalidated measurement instruments were more likely to report treatment effectiveness than those employing validated measures. This entry describes the specific challenges involved in assessing the construct validity of health status measures, addresses the evolving conceptual framework for validity and associated taxonomy, explains the main approaches used, and provides additional resources for more in-depth discussion of theory and methods.