Multivariate analysis of the predictors of verruca at 80(GEE) method.

Goodness-of-fit assessment: Number of clusters  = 57, number of observations  = 166, the estimated area under the Receiver Operating Characteristic (ROC) curve  = 0.875>0.7, adjusted generalized R2 = 0.512>0.3, and Hosmer-Lemeshow goodness-of-fit F test p = 0.350>0.05 (df = 9, 156).
Prediction: To calculate the estimated probability of being verruca (i.e., the predicted value,) given the observed covariate values, one can use the following formula. According to the above fitted multiple logistic regression model
the predicted value of observation i is
where log ZSD =  logarithmized standardized Z value, and θSD =  standardized θ value.