Drinking too much alcohol can be damaging for our health. Doctors want to see what things are tied with alcohol drinking behaviors so that they can find ways to develop tips/strategy for a target set of people. They look at three factors, those being the age, the gender, and how long the individual has been in school. They get a random sample of 10k individuals and use a logit equation

here the dependent variable i 1 if the person drinks and 0 doesn't.

female equals 1 if the individual is female and 0 if they arent.

age is how old they are

education is the number of years of education

Lets say we find the mean values for age, education, and the number of females vs males.

mean age: 35.8

mean years in school: 9.5

female/males in sample: 0.46

What is the probability of drinking alcohol for the average participant in our sample?

Iteration 0: Iteration 1: Iteration 2: Iteration 3: log likelihood = -4791.6962 log likelihood -4642.3802 log likelihood -4638.9265 log likelihood = -4638.9187 Logistic regression = Number of obs LR chi2 (3) Prob > chi2 Pseudo R2 305.56 0.0000 0.0319 Log likelihood = -4638.9187 smokes Coef. Std. Err. yearsed age female cons -0.082 -0.0142 -0.172 1.129 0.006 0.001 0.030 0.102