Asian Journal of Probability and Statistics

This study focused on modeling the probability of infant survival at birth using the Logistic and Probit regression models. Status of Infant at birth was the response variable; while Systolic BP, Diastolic BP, Age of Mother, Sex of Baby, Weight of Mother, Mode of Delivery, Age/Week Gestation, Parity, and Weight of Baby were the explanatory variables. The Study Data used for illustration comprised of live deliveries at the Federal Government of Nigeria Model Primary Healthcare Center located at Isu-Njaba Town in Isu L.G.A. of Imo State, Nigeria. The test for relationship between any two of the variables used in this study showed that there were negative significant associations between Status of Infant at Birth and Mode of Delivery, and Birth Weight and Mode of Delivery; while there were positive significant associations between Systolic BP and Diastolic BP, Age of Mother and Systolic BP, Gestation and Systolic BP, Gestation and Age of Mother, Diastolic BP and Weight of Mother, Parity and Age of Mother, Parity and Weight of Mother, Parity and Diastolic BP, Birth Weight and Weight of Mother, and Birth Weight and Parity. Multiple logistic and multiple Probit regression models were fitted on the Study Data, and the results of the goodness of fit test using Likelihood Ratio Test at 5% level of significance showed that both models were of good fit. The result of the Wald test showed that Weight of Mother, Mode of Delivery and Birth Weight were significant to the response variable (Status of Infant at Birth – Dead or Alive) at 5% level of significance. The study concluded that Infant Survival at Birth was significantly influenced by Weight of Mother, Mode of Delivery and Birth Weight. Also, both the Logistic and Probit regression models were relatively the same; with AIC values of 164.0844 and 164.2984, respectively.