check-ups and a biopsy at the end of the study. Suppose that the investigators fit a multiple logistic model with prostate cancer status as the binary response (1 if prostate cancer; 0 if no prostate cancer). The model has two predictors: Finasteride (1 if subject takes Finasteride; 0 if they take placebo) and Over60 (1 if a subject is over 60; 0 otherwise). The output below shows the estimates from the fitted model. Coefficients : Estimate Std. Error z value Pr(> z ) (Intercept ) -1. 88067 0 . 05075 -37. 059 0.05. O decreases the odds of developing prostate cancer compared to placebo, since the p-value is > 0.05. O decreases the odds of developing prostate cancer compared to placebo, since the p-value is