AbstractBackground: Measurement invariance is an important psychometric property for scale validation and is necessary for group comparisons in clinical practice and research. Item-level invariance referred to as differential item functioning (DIF) is an essential but often overlooked aspect of measurement invariance. This study examines item- and scale-level measurement invariance of the minimum dataset (Patient Health Questionnaire-9 (PHQ-9), Generalised Anxiety Disorder-7 (GAD-7) and Work and Social Adjustment Scale (WSAS)) used in Improving Access to Psychological Therapies Services (IAPT) in relation to gender.
Method: The study involves secondary data analysis of IAPTUS data. Mokken scaling analysis and hybrid ordinal logistic regression were used to examine DIF. Structural equation modelling was also used to determine whether the related structure of the measures could be considered gender invariant.
Results: Items from the PHQ-9, GAD-7 and WSAS were flagged as demonstrating statistically significant DIF, however, the magnitude and impact of DIF was negligible. The results indicate that the interrelationship of the measures was consistent across gender.
Discussion: The results suggest whilst there might be some minor differences in the expression of symptoms for the IAPT minimum dataset (PHQ-9, GAD-7 and WSAS), the measures generally function the same across gender, allowing for comparable gender-based evaluations in an IAPT population.
|Date of Award||Sept 2021|
|Supervisor||Alan Robert Bowman (Supervisor)|