Www.medsci.org

Based on the result of binary logistic regression and formula of logistic model: logit(P)= In [P/(1-P)] = β0 + β1X1 +… + βnXn (the value of β comes from logistic regression, X is the independent variable, n is the number of independent variables), we can get the value of probability for dependent variable: P = elogit(P) /1+elogit(P). ................
................