Multivariate Regression Analysis of Babies Birth Characteristics in Bayelsa and Rivers States, Nigeria

This research work used multivariate regression analysis to analyze babies’ birth measurements which is appropriate since there are several numbers of dependent variables and independent variables, usually the case in a real-life situation. The dependent variables [Lean Body Mass (LBM) Y 1 and Body Mass Index(BMI) Y 2 ] for some fixed values of the independent variables Head Circumference (CM) X 1 , Age (Days) X 2 , Height (CM) X 3 and Weight (Kg) X 4 ]were collected. Furthermore, the work was able to fit a multivariate regression model (MRM) to the data sets collected, estimated the model parameters and obtained matrices of the fitted coefficients with its residual matrices. Evaluating the significance of the multivariate regression model parameters, by using Bartlett’s approximation test statistic to determine if any single variable Y i (Y i , i=1 and 2) is to be predicted from all the independent variables in the system. The research work was able to fit several multivariate regression models to the data obtained. From the models built, we notice that X 2 (Age in Days) for model Y 1 and also X 1 [Head Circumference (CM)] and X 2 (Age in Days) for model Y 2 are not significant for female and male in Bayelsa State, while X 2 (Age in Days) for model Y 2 is not significant for female and male in Rivers State. This result confirms that there are some independent variables, that is ij ≠ 0 and shows that some independent variables explain the dependent variables.