Team F

100/100 for I3 and I4.

Bassem Mikhael
Christopher Loncke
Abigail Daughtrey
Prerna Agarwal
Yong-Gyun Choi

General Comments:
----------------

This looks great.  Looking forward to I5 and beyond.

Specific Comments:
-----------------

* Nice job on I3

* Your sampling plan looks very good.  Note that the number 500 may
  change when you do a more formal sample size calculation (I5). 

  Another reason to think of stratifying is if you think caffeine
  consumption will be different for different groups on campus.  If
  that is so, stratifying by those groups can lead to more efficient
  estimation (narrower CI's).  However, if in the end you decide
  against stratification, that's OK.

* Good approaches to both nonresponse and non-representativenes.s.