TITLE:
Cumulative Link Modeling of Ordinal Outcomes in the National Health Interview Survey Data: Application to Depressive Symptom Severity
AUTHORS:
Andre Williams, Louisana Louis
KEYWORDS:
Cumulative Link Model, Ordinal Outcome, Depressive Symptom Severity, National Health Interview Survey
JOURNAL NAME:
Journal of Data Analysis and Information Processing,
Vol.14 No.1,
December
23,
2025
ABSTRACT: This study investigates the application of cumulative link models with alternative distributions (hyperbolic secant, Laplace, and Cauchy) to model ordinal outcomes of depressive severity using 2022 National Health Interview Survey data. The primary objective was to assess whether these models provide a better fit to ordinal response data and more accurate predictions than their traditional counterparts with the logit link function. The results indicate that the logit model achieved the highest classification accuracy, correctly classifying 83.54% of the cases. The Cauchy model demonstrated the best model fit, i.e., the lowest AIC and BIC values. This study highlights the importance of considering both classification accuracy and model fit when selecting a statistical model.