so the beginning is just what the professor covered in the lecture . The actual

so the beginning is just what the professor covered in the lecture . The actual assignment is listed under assignment 4. The information is listed for you to execute but here it is again : use http://www.stata-press.com/data/mlmus4/bp , clear
Also this needs to be executed as well before starting the assignment :
encode treat, generate(treatment)
label variable center “center identifier”
label variable id “patient identifier”
label variable time “visit number 3,4,5,6 (postrandomization)”
label variable bp “diastolic blood pressure in mmHg”
label variable bp_base “diastolic blood pressure during second visit, prior to randomization”
label variable treatment “treatment group A, B, C”
sort center id time
the questions are :
Q1. Build a panel regression model with bp as the outcome; use bp_base, treatment, and center as independent variables. What is the population-average estimator effect of treatment on bp?
Q2. Use the same model as Q1, but use between estimator. What is the effect of bp_base on the DV?
Q3. Generate a new variable for bp and assign values of 0 if they are lower than mean of bp and the value of 1 if bp values are equal or greater than the mean of bp. Use panel logistic function and your new variable as DV and all other IV’s from Q1. What is the coefficient value of treatment effect on your new DV?