Analyzing differential item functioning of the Nottingham Health Profile by Mixed Rasch Model
No Thumbnail Available
Date
2018
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Objectives: This study aims to evaluate whether items in the Turkish version of the Nottingham Health Profile (NHP) function differently according to age, sex, and duration of pain using the Mixed Rasch Model (MRM).
Patients and methods: The NHP data of patients with low back pain from a previous study was analyzed. To analyze differential item functioning (DIF) within the items of the NHP, the MRM was used. Age, sex, and duration of pain were considered factors which could cause DIF. The most powerful factor to define latent classes derived from the MRM was estimated using the Rasch tree method.
Results: The two-class mixture version of the nominal response model was identified as the best fitting model for the physical mobility, sleep, and social isolation sections. For the physical mobility dimension, some items showed DIF by age and for the social isolation dimension some items showed DIF by sex. For the sleep dimension, latent classes were unable to be explained by age, sex, and duration of pain. Items in other dimensions of the NHP were DIF-free and no items showed DIF according to age, sex, and duration of pain.
Conclusion: The results of this study may shed light on explaining the different response behavior of patients on the items of the NHP. Age and sex were found to be variables affecting item responses of NHP. Contrary to expectations, duration of pain did not cause DIF for any item. From the perspective of DIF, it can be concluded that NHP scale is a robust scale in terms of validity.
Description
Keywords
Differential item functioning, Mixed Rasch Model, Nottingham Health Profile