Negative Binomial Regression in Underdispersion(Case Study: Neonatal Mortality in Jambi City)

Neonatalis is birth before 28 days of a baby. Factors that are considered to affect neonatal mortality include the number visits in 1st and 3 rd trimester, pregnant women receiving Tetanus Diptheria Immunization, estimated infants with complications, Hepatitis B Immunization for less than 24 hours, BCG 1 visits. Neonatal still very rare so right analysis used, namely Negative Binomial Regression. This research aim investigate negative binomial regression underdipersion on at Jambi City. These two methods specifically used Poisson distributed data because they rare. The stages will be carried out distribution test equidispersion assumption, parameter estimation, model feasibility test, selection best model. results obtained without variable 3rd-trimester or who received immunization AIC was 36.3.