Keynote Lecture by Dr. Ida Marais
Rasch Measurement Theory and the measurement of change.
The keynote lecture investigates two factors that affect the measurement of change in health outcomes assessment using Rasch Measurement Theory (RMT) and then shows how the measurement of change in intelligence and attainment tests can be enhanced using Rasch meta-metres of growth.
Part I.
Response dependence and person-item alignment: The two factors considered in the assessment of health outcomes are response dependence and person-item alignment. In the case that the same questionnaire is used on multiple occasions, the potential bias in estimates because persons tend to respond in the same way to a greater degree than their parameter values imply arises (Marais, 2009). Andrich (2017) presented a mechanism for assessing, controlling, and eliminating this kind of bias in repeated measurements using RMT for dichotomous responses. In this paper a generalisation of the method is presented where items have more than two response categories. Person-item alignment at baseline impacts the effect of response dependence and the measurement of change. A set of simulation studies demonstrating these impacts is presented. Data from the Longitudinal Study of Australian children (LSAC), in particular longitudinal trajectories of mental health, are examined in light of the findings.
Part II.
Rasch meta-metres of growth: One of the methodological challenges in measuring and comparing rates of growth is that growth is usually not linear. In Rasch Meta-Metres of Growth for Some Intelligence and Attainment Tests (Andrich, Marais,& Sappl, 2023) we take advantage of an insight made by Georg Rasch, where a logarithmic transformation of the time variable, which he called a meta-metre, results in linear growth in the meta-metre. In particular, individuals get a single rate of growth which can then be used to compare rates of growth among groups using standard methods. In this paper examples are presented where the meta-metre helpfully describes growth of different phenomena, such as the development of intelligence, reading, and mathematics.
Andrich D. (2017). Controlling response dependence in the measurement of change using the Rasch model. Statistical Methods in Medical Research, 1-17.
Andrich, D., Marais, I & Sappl, S. (2023) Rasch Meta-Metres of Growth for Some Intelligence and Attainment Tests. Springer Nature Singapore.
Marais, I. (2009). Response Dependence and the measurement of change. Journal of Applied Measurement, 10, 17-29.
Ida Marais BSc, (North-West), MSc, PhD (Auckland) is a Principal Research Fellow at the Psychometric Laboratory, Medical School, The University of Western Australia. She was awarded a University of Auckland Doctoral Scholarship to complete a PhD in Psychology and has since worked as a researcher and lecturer in the fields of Psychology, Health and Education in New Zealand and Australia. She also works as a psychotherapist in private practice. Her research work with David Andrich focuses on social measurement using the Rasch model for measurement. Her work has been published in top international journals as well as in two books, ‘A Course in Rasch Measurement Theory’ and ‘Rasch Meta-Metres of Growth for some Intelligence and Attainment Tests’. She is currently working on the measurement of change, in particular change in Patient-reported Outcome Measures in clinical trials.
Find her publications:
Please see her two two recent co-authored books: