Item ordering of personal disturbance scale (DSSI/sAD) in a longitudinal study; using Mokken scale analysis

Abstract

In order to establish a common scale with a stable structure over time to use in repeated measures studies, this study uses information from a large community-based cohort of women. Mental health was assessed at six time points during 21 years, using the Delusions-Symptoms-States Inventory/States of Anxiety and Depression (DSSI/sAD) for 2698 women with mean age of 25.2 years at baseline. Mokken scale analysis is applied to the 14 items of DSSI/sAD at each time point, using a hierarchical clustering Algorithm and a Genetic Algorithm. Both algorithms resulted in one Mokken scale at each time point. After excluding items that did not meet the assumptions of Mokken scaling for polytomous data, a single common scale with 10 common items was extracted from scales, which was highly reliable at each time point. This measure with reasonable hierarchy of severity of items (symptoms) has a stable structure over time and can be used to track the changes of women’s mental health over time and analysed via repeated measures analysis.