Statistical Analysis on Tackling Diabetic Patients among Adults in Damaturu, Nigeria

Abstract

This paper assesses the study that was carried out based on the broad objective of using the Box M and Wilks Lambda tests for tackling diabetic patients among adults in Damaturu, Yobe state. In this research work, 874 diabetic patients data was collected from the Yobe State Specialist Hospital, Damaturu for ten consecutive years (2008-2017).The data analysis were done using Box's M test and Wilks' Lambda test. The diabetic patient's responses were categorized into two groups which include: healthy ("0”) and diabetic ("1”). The variables (predictors), that is; age, sex, weight, blood sugar and urine sugar are used based on discriminant analysis. By using Fisher's method of discriminant analysis, the classification table results were obtained while assumption of homogeneity of covariance matrices is tested with Box's M test. This test is very sensitive to meeting also the assumption of multivariate normality. For our data, the test is significant so we conclude the groups do differ in their covariance matrices, violating an assumption in discriminant analysis. Additionally, Wilks' Lambda of the function is 0.546 and has the value (Sig. 0.00), thus we can say that the model is a good fit for the data since p < .05. This multivariate test is a goodness of fit statistic, just like the F-test in the case of regression. In the light of the above it is recommended that doctors and clinics should adopt the use of the models built by this research to detect prevalence of diabetics among adults so that adequate measures for prevention and control of diabetics can be taken early enough to alert the danger of the full manifestation of the disease.