Team H

100/100 on I3 and I4.  

Jason Sun
Erika Tang
Zhiyi Tang
David Zimmerman

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

Excellent job of dealing with *both* topics.  Looking forward to I5.

Specific Comments, Undergrad Prospects:
--------------------------------------

* Generally good answers to I3 and I4.

* I4 B It's not clear to me how you are thinking of carrying out
  sampling within each year.  In II5 you will need to estimate and
  justify a sample size for example.

* In I4 E I like that you are trying to think through a reasonable set
  of questions to be asking yourselves as you analyze the data.

* In I4 F & B, I don't see why you would choose random years.  I do like
  that you are looking at multiple years though.  If you want to
  consider just 5 years' of data take the even numbered years (or the
  odds!).   Otherwise just take all of them.

* Draft IRB form is fine except that part 8 you can't realistically do
  informed consent, so you should tick the boxes that indicate you do
  NOT intend to use consent forms.

* As soon as you have the data, I would like you to give me a complete
  report on the data, including

  - how large is the data set in each year -- how many students in
    each of the years you have?

  - listing of all the variables collected and their definitions (if
    there are differences for different years, list separately)

  - What was CMU's sampling design?  non-response rate? sampling
    weights? etc.  Or did CMU expect it to be a census?  (note that
    non-response is still an issue for a census)

  [this isn't part of most teams' project assignments but in this case
  I need that information to provide you good feedback from here on
  out]

Specific Comments, Pedestrian Patterns:
--------------------------------------

* I3 and I4 look good here too.

* If you continue to II5 with this project you will need to calculate
  a sample size (using methods in class, and pretending to bw working
  with a SRS w/o replacement, for the purposes of sample size
  calculation).  2000 may be a bit big.

* IRB draft looks good except for
  - you will have to recalcualte the sample zize
  - Very likely the waiver of informed consent (sect 8,. second from
    end of page) would be denied.   Waiver of *signed* informed
    consent is more likely.

* I like the elaborate set of questions you are developing.

* I am a little concerned about the coverage of the locations you have
  listed.  Perhaps there is a way to get additional locations
  (e.g. secondary entrance to Baker) into your design, without
  expanding the work too much.

* understood that this is a backup project in case the admin record
  study doesnt' happen.