Assessing measurement in health: Beyond reliability and validity

Abstract

Background: Psychometric concepts have undergone a transformation in health fields, as articulated in a consensus report by an international panel of health measurement experts: COSMIN, the COnsensus-based Standards for the selection of health Measurement INstruments.

Objectives: The aims of this paper are to describe emerging ideas relating to the development and testing of new measures in health fields, to present a revised measurement taxonomy that builds upon COSMIN, and to explore the extent to which the new measurement concepts have played a role in psychometric assessments in nursing.

Design: A descriptive analysis of a sample of psychometric papers published in three major nursing journals was undertaken.

Methods: A new measurement taxonomy is presented and explained. A sample of 105 studies, representing a consecutive sample of psychometric studies published in the International Journal of Nursing Studies, Nursing Research, and Research in Nursing & Health between 2010 and 2014 was reviewed to ascertain the extent to which psychometric assessments in nursing map onto the new taxonomy.

Results: Most nursing studies reviewed adhered to traditional concepts of psychometric assessment, which focus on reliability and validity. The studies in the sample rarely involved assessments of longitudinal measurement aspects, namely the reliability and validity of change scores (responsiveness).

Conclusions: Many constructs of interest to nurse researchers are amenable to change—and these constructs are frequently the target of nursing interventions designed to foster change. Future psychometric work by nurse researchers would benefit from assessments of the psychometric adequacy of change scores.