suppose we want to estimate the effects of alcohol consumpti

Solution

i.

Yes we should check for a regression and correlation between alcohol consumption and should attend if the correlation is high. This will guide us to neglect attend as variable, but if it is not then attend should be included since, colGPA is depends on number of classes he attend.

ii

SAT and hsGPA must be included as explanatory variables since they states about overall intelligence of students. For conducting study we should not neglect other variables which could affect depend variables.