- Amanda Faza Fahirani1, Netti Herawati1*, Dorrah Azis1, dan Nusyirwan1
- DOI: 10.5281/zenodo.15079402
- Global Academic and Scientific Journal of Multidisciplinary Studies (GASJMS)
In Poisson Regression, assuming that the variant
value must be equal to the mean (equidispersion), then another alternative is
used for data that has a variant value greater than the mean (overdispersion).
Overdispersion in Poisson regression can be overcome by using the Zero Inflated
Poisson model. The purpose of this study is to model Zero Inflated Poisson
(ZIP) with New Modified Bias Estimator in cases of Tetanus Neonatorum in Bandar
Lampung, Indonesia. In addition, this study also compares the performance of
Zero Inflated Poisson with new modified bias estimator with Maximum Likelihood
Estimation (MLE) method in handling multicollinearity and overdispersion in
application data. The best model based on the smallest Mean Squared Error (MSE)
value. The results of the study indicate that the Zero Inflated Poisson model
with a new modified bias estimator is better in modeling Tetanus Neonatorum
case data in Bandar Lampung, Indonesia because it has a smaller MSE value (0.8819)
than the MLE (75.5678). The Zero Inflated Poisson model with a new modified
bias estimator shows that Tetanus Neonatorum in Bandar Lampung, Indonesia is
influenced by the percentage of pregnant women examined ()
and the percentage of neonatal examination (
).