# Econ

1039 Words5 Pages
Preface In our data analysis we do some univariate analysis before proceeding to models. In survival analysis it is highly recommended to look at the Kaplan-Meier curves for all the categorical predictors. This will provide insight into the shape of the survival function for each group and give an idea of whether or not the groups are proportional. We also consider the tests of equality across strata to explore whether or not to include the predictor in the final model. For the categorical variables such as marital, eservice, plusservice and totalservice we use the log-rank test of equality across strata which is a non-parametric test. For the continuous variables such as age, address, income, education, employment, and reside we use a…show more content…
Consequently we find the better model after dropping variables with the p-values more than 0.05. Model 5 is the best model which has all the significant variables and has lowest AIC value of 1478.82 and BIC value of 1513.17. If the marital status is altered from married to unmarried, while holding all other variables constant, the rate of churn decreases by (100% - 66.5%) = 33.5%. As year in current address is increased by one year, all other variables are held constant, the rate of churn decreases by (100% - 94.5%) = 5.5%. As year with current employer is increased by one year, all other variables are held constant, the rate of churn decreases by (100% - 93.2%) = 6.8%. As people switch from non eservice to e service, all other variables are held constant, the churn rate decrease by (100% - 56.5%) = 43.5%. As people switch from non plusservice to plusservice, all other variables are held constant, the churn rate decrease by (100% - 54.5%) = 45.5%. As education level increases, all other variables are held constant, the churn rate increases by 12.83%. Finally, I adjust my model for heteroscedasticity by using robust function. The final adjusted model is presented below: Table 1: Final Model (Corrected for Heteroscedasticity) |Variable |Haz. Ratio |Robust Standard Error |Z-value |Significant | |marital |0.665 |.0790557 |-3.43 |** |