What can modern psychometric techniques add to health psychology research methods?

Abstract

Traditional psychometric methods have provided a useful and conventional framework of developing and evaluating self-report instruments. Nevertheless, the Classical Test Theory (CTT) underpinning traditional psychometrics is a theoretical non-testable theory comprising assumptions that are usually easily met by scale data. Therefore, utilising the CTT could potentially lead to weak conclusions regarding the psychometric properties of instruments used in patient research and subsequently contribute to type 1 and type 2 errors. Modern psychometric techniques such as the Rasch Measurement Theory (RMT) addresses all limitations of traditional psychometrics. Firstly, the RMT paradigm offers a testable model that can be utilised to verify the measurement properties of scales rigorously. Secondly, the RMT enables the development of linear interval-level measurement on the basis of ordinal-level raw data. Thirdly, within the RMT, item and person location estimates can be provided and this can lead to adaptive testing through the use of item subsets to reach measurement and fourthly, RMT enables individual-level measurement. Psychometric evaluation examples of questionnaire-based patient data are reviewed to designate the advantages of using both traditional and modern psychometric techniques.