Many ways to skin a cat: psychometric methods options illustrated

Abstract

Abstract Background The three articles in this issue from members of the Psychometric Special Interest Group (SIG) of the International Society for Quality of Life Research (ISOQOL) examine three different psychometric techniques researchers use to analyze item and scale properties of a patient-reported outcome (PRO) instrument. The articles illustrate their respective strengths and weaknesses. Main text Many published articles use one of the three methodologies analyzed by the authors and the reader should have a basic familiarity with the assumptions, approaches, and statistical techniques behind each analysis. These three papers shed light on some of the conundrums facing developers and users of PRO measures and data regarding what method and instruments to use. These papers have used a dataset on depressive symptoms to show that no attempt to measure such a complex feeling domain as depressed mood can cover the entire spectrum of the experience. Conclusions As a group, these three papers will help readers evaluate published articles on instruments using one or more approaches as well as providing general education on these statistical methods in application.

Keywords

• Psychometrics
• Classical test theory
• Item response theory
• Rasch measurement theory
• Patient-reported outcomes