Determining the factors affecting the happiness levels of divorced individuals by ordered logistics regression analysis


Creative Commons License

Uyar B., Büyüksu C.

Manas Journal of Engineering, cilt.10, sa.2, ss.193-202, 2022 (Hakemli Dergi)

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 10 Sayı: 2
  • Basım Tarihi: 2022
  • Doi Numarası: 10.51354/mjen.1021287
  • Dergi Adı: Manas Journal of Engineering
  • Derginin Tarandığı İndeksler: TR DİZİN (ULAKBİM)
  • Sayfa Sayıları: ss.193-202
  • Van Yüzüncü Yıl Üniversitesi Adresli: Evet

Özet

In cases where the dependent variable is categorical and ordinal, the Ordinal Logistic Regression Model is used for model estimation. In order to predict the Ordinal Logistic Regression Model, it must provide the parallel lines assumption. In the study, the happiness levels of divorced individuals were estimated with the ordinal logistic regression model. The data set used in the analysis was obtained from the Life Satisfaction Survey implemented by the Turkish Statistical Institute in 2020. Brant's Wald Test and Likelihood Ratio Test were applied for the parallel lines assumption and the null hypothesis could not be rejected. In this context, the model ordinal logistic regression model was estimated. The statistically significant gender variable shows that divorced women are happier than divorced men. It has been determined that success and job variables tend to decrease happiness levels in divorced individuals compared to other factors. In general, when the education level is examined, it is seen that the level of happiness of divorced individuals increases as the education level increases. It has been concluded that divorced individuals who are satisfied with their health, education, income and social life are happier than divorced individuals who are not satisfied with their health, education, income and social life. It has been determined that divorced individuals are registered with the Social Security Institution and their happiness levels are higher than those who are not registered.