Table 12-16 shows a sample of 30 randomly selected patients with chronic kidney disease (CKD). The gender, systolic blood pressure, and presence/absence of diabetes are also indicated. Higher the rating of CKD stage, the more serious the problem.

(a) Develop logistic regression models to predict the chance of occurrence of a certain CKD stage.

(b) Interpret the coefficients of Gender, Blood Pressure, and Diabetes.

(c) Interpret the odds ratio for Blood Pressure. Is Blood Pressure a significant predictor if α = 0.05?

(d) Are the models significant at α = 0.05?

Transcribed Image Text:

TABLE 12-16 Data on Chronic Kidney Disease (CKD) CKD Blood CKD Blood Stage Gender Pressure Diabetes Stage Gender Pressure Diabetes 2 4 1 3234 3 1 4 MPMMPMPMFFMMPMM M 128 N 3 F 132 Y 130 N 1 F 124 N 135 Y 126 Y 22 M 130 N 2 M 132 Y 138 N 4 M 150 Y 126 N 3 M 130 N 135 Y 2 F 130 N 140 Y 1 F 124 N 128 N 1 M 126 N 126 N F 132 N 130 Y 4 F 135 N 130 Y 1 M 125 N 142 142 3 145 ANY 223 M 134 Y 2 F 132 Y 3 M 134 Y