Title: Using the box-Jenkins methodology to predict the number of divorces
Authors: Maryam Mehdi and Nibras Talib Mohammed
Volume: 9
Issue: 8
Pages: 120-128
Publication Date: 2025/08/28
Abstract:
The box-Jenkins methodology was applied to analyze time series and predict the number of monthly divorces in Karbala governorate for the period (2025-2024) using Karbala court data (2020-2024). The analysis showed that the time series is unstable, and stability was achieved after applying the second-order divergence. By analyzing the functions of autocorrelation (ACF), partial autocorrelation (PACF) and information parameters (AIC, BIC, HQ), the ARIMA model(3,1,3) was identified as the best predictive model. The results showed the efficiency of the model through Leung-box tests and residue analysis, where the annual expected value of divorces in 2025 was about 410 cases with a confidence level of 95%. The study found an upward trend in divorce cases and recommended the establishment of family counseling centers and the adoption of social policies to reduce this phenomenon.