Understand the factors that contribute to underage drinking. Results can be used to assist schools and high school programs that are looking to decrease or prevent underage drinking among its students.
The following variables are statistically significant: gender, workday drinking, likelihood to go out with friends, weekly study time, quality of family relationship, student’s health status and type of guardian (mother, father or other)
With these inputs, a multiple regression model is able to account for 44% of the variation in weekend drinking, with 95% confidence.
A percentage increase in weekday drinking increases weekend drinking by 0.61 percent, holding all other variables constant.
Adjusting for the chosen covariate (free time), we can see that the number of drinks on the weekend increases at the Mousinho da Silveira School. This can be replicated usingn other variables as the covariate.