Evaluating divorce dynamics through ODE modeling and statistical hypothesis testing

Abstract

To understand divorce, a quantitative method is used which considers the social, economic, and psychological elements impacting marital changes. We introduce a model based on Ordinary Differential Equations (ODEs), combined with statistical hypothesis testing, to examine divorce trends over two decades using longitudinal, real-world data. Model parameters are estimated through nonlinear least-squares fitting, resulting in a high predictive accuracy (R2=0.9878), indicating the model’s dependability. Robustness is further confirmed through residual analysis, Durbin-Watson (DW), Jarque-Bera (JB) statistics, and normality testing. Consequently, the results provide important understandings of how divorce trends are changing, supplying a data-supported basis for policymakers and researchers to develop helpful intervention strategies to foster marital stability and lower divorce rates.