Impact of Educational Attainment on Depression Measurement: Insights from Differential Item Functioning

Abstract

Level of education can impact access to healthcare and health outcomes. Increasing rates of depression are a major public health concern, and vulnerability to depression is compounded for individuals with a lower level of education. Depression screening is important in the management of health and more so, in Primary care and mental health. This paper aims to explore the effects of educational levels on the psychometric properties of depression measures namely BDI-II, CES-D, PHQ-9, and PROMIS Depression items. It is an objective of the proposed research to establish the sources of DIF and make recommendations on modification that can be made to eliminate bias. Specifically, extrapolating results from PROMIS 1, the NIH Toolbox, and the PROsetta Stone panel studies, it weighs the educations of participants and examines DIF values for each measure. Research findings indicate that the DIF values are higher among those with lower education levels, particularly those with high school education or lower. Using the BDI-II and the CES-D yields higher levels of education disparities than when using the PHQ-9. Such adjustments may consist in rewriting particular items to enhance their clarity, and in adapting the organisation of assessments to take into consideration disparities in schooling, in order to ensure fair and valid measurement of mental health in diverse samples.