Here is where we are with respect to the projects:
====================================================================
Each group should divide its efforts on the pills project into
(i) data and context - Use online searching, library, and possibly
telephone or personal interviews with area experts who you can locate
and who are willing to provide some information. If you contact someone
(e.g., with the government or with some suitable association or lobbying
group), make sure you first have done sufficient background reading that
you'll know the basics. An expert isn't going to give you a lengthy
tutorial and he or she will expect you to be familiar with the basic issues.
But if you ask nicely and if you show that you've made a sincere effort to
prepare, such people can be extremely helpful!
(ii) analysis - This can proceed in parallel with (i). I've already
mentioned the basic possibilities of confidence intervals and hypothesis
testing. However, you may find after doing some searching online, in our
text, or in the library, that there are other interesting methods.
(iii) display - Again, searching online, in libraries, and our text,
you can find a variety of ways to present your conclusions and to display
the data. The project will be initially a report to me (and yourselves),
but if it is sufficiently appealing, I'll let you also make a presentation
to the entire class!
Next week, I expect that each group will have some specific questions
related to these three aspects.
======================================================================
For the bus project, we are going to monitor two routes:
"D"s (between Dupont Circle and Calvert St.) at 37th and Whitehaven Pkwy
best location: Park on west side of 37th St. at Whitehaven Parkway
"30"s (Wisconsin Avenue across from the Holiday Inn just North of Safeway)
best location: Old cemetary on west side of Wisc. Ave. or maybe
the rooftop of the starbucks opposite the safeway (if you can see
Wisc. Ave. from there)
For each route, you need to note the
direction and time
of each bus passing your observation point, using U for uptown and D for
downtown, with U meaning the bus is going north (on 37th or Wisc.)
Then give me the data sheets you've collected which will be entered into
the computer to produce the two time series for bus times. Count _all_
buses you see except for "not in service" (i.e., an empty bus which is
not accepting passengers).
I expect each of the groups to divide into two-person units to collect
the bus data - one person can be the "spotter" while the other records
the information in one of the sheets I'll hand out in class tomorrow
(also will be available outside my office door, Reiss 258, on top of
the file cabinets). Two of the groups have an odd number of people so
it may sometimes be necessary (or convenient) to have three. Please
put in at least two hours collecting data, so you need only do an hour
as "spotter" and you can divide your work in shorter than two-hour
segments if that is more convenient.
Please only collect information on _weekdays_ as this will decrease
the variance of our study. Any convenient time is ok.
I hope everyone will be able to collect bus data by sometime next week,
but
Thurs. May 5 at noon
is the _absolute cutoff_ date since I'll need
time to input the data and to determine the corresponding statisitcs
for the two bus routes, where the statistics will measure the degree
to which the time series of buses is consistent with a Poisson distrib'n
for the route.
Back to the prob-stat page.