|
M
11/1/04
|
Day of rest.
|
|
|
T
11/2/04
|
HW due:
Read §5.1.
|
|
|
W 11/3/04
|
HW due:
Write #5.20, 5.22, 5.24, 5.25; start reading in §5.2, with no reading notes
in §5.2 required just yet.
Quiz today on these recurring
weekly features from the Nov. 2 issue of the Post: “The Dose” (page F2) and “Quick Study” (page F6).
|
|
|
Th
11/4/04
|
HW due:
Finish reading §5.2, with reading notes.
|
|
|
F
11/5/04
|
No school (faculty
meetings).
|
|
|
M
11/8/04
|
HW due:
#5.37, 5.38, 5.42, plus a written paragraph in which you describe an experiment
you would be interested in conducting. For the moment, do not worry too much
about the practicality of conducting the experiment. Choose something that is
really of personal interest to you.
The following paragraph is an example, though obviously a slightly outdated
one. You do not have to write quite this much. If you use blinding or
blocking (or matched pairs, which is a form of blocking), be sure to say so.
I would like to see whether the manner
with which a house is approached on Halloween has any effect on the amount of
candy given. My design would be to select a neighborhood in Bethesda and give
each house four treatments, with a spacing of at least 3 minutes between
treatments to reduce the chance that the homeowners perceive any linkage between
the testers. There will be four testers, namely three friends and I. The four
treatments are as follows: (1) gruff manner, no costume, not saying “trick or
treat” or “thank you” to homeowner; (2) polite manner, scary costume, saying
“trick or treat” and “thank you” to homeowner; (3) polite manner, non-scary
costume, saying “trick or treat” and “thank you” to homeowner; and (4) gruff
manner, scary costume, but growling “trick or treat” and “thank you” to
homeowner. The sample size will be 60 houses, and each house will receive all
four treatments. This is a matched quadruples design, and the order of the
treatments for each house will be randomized. To reduce the lurking variable
of tester differences, each tester will perform all four treatments; each
tester will give his treatment #1 to 15 randomly chosen houses from the 60,
treatment #2 to a different subset of 15, and so on. Data will be recorded
for each house, out of sight of the homeowner. Data will be compared at the
end to see whether there were any statistically significant differences
between the amount of candy obtained, both on a house-by-house basis and in
the aggregate.
|
|
|
T
11/9/04
|
Group 1: Martin B.,
Nick, David P.
Group 2: Martin M., David C., Sam
Group 3: Tim, Ricky, Pan
HW due: Groups should submit a
rough draft of an experimental design they would like to pursue for the next
group project. Your submission will be graded (at this preliminary stage) for
completion, not for quality. If for some reason you were unable to meet with your
group, either in person or on the telephone, then you may instead read at
least two of these recent articles and prepare to discuss the subject of
those articles in class.
|
|
|
W
11/10/04
|
Quiz on “The Dose” and “Quick Study” from The Washington Post of 11/9/2004. Groups 2 and 3 should turn in a more complete rough
draft of an experimental design today.
|
|
|
Th
11/11/04
|
Test on Chapters 3
and 4 (100 points). You have
already been quizzed on this material. Try these links if you would like more
review.
|
|
|
F
11/12/04
|
No homework is due today.
|
|
|
M
11/15/04
|
HW due:
Read §5.3 and the Chapter Review on pp. 318-319; write #5.56.
|
|
|
T
11/16/04
|
HW due: Write
#5.73. You will need to develop your own notation for keeping track of the
simulation. Make sure your steps are legible and easy to understand; you may
wish to have one person design and execute the simulation while another keeps
a clear printed record. Because this is such a difficult problem, you may
work in pairs if you wish (i.e., one submission will be accepted from two
people).
In class: Fresh Air with Terry Gross,
Aug. 18, 2004.
|
|
|
W
11/17/04
|
Quiz on Washington Post weekly reading in the Health
section, plus yesterday’s audio
interview, plus this
article from the latest issue of Psychology
Today. For the latter, be prepared to criticize or defend the methodology
of the article.
|
|
|
Th
11/18/04
|
HW due:
Project groups should prepare a draft timeline for the experiment.
In class: Discussion (and possible quiz) concerning the Psychology Today article,
followed by group meeting time to refine methodology for the experiment.
|
|
|
F
11/19/04
|
No additional HW due.
Please get a good night’s sleep! There will be an oral spot check to make
sure everyone has listened to the audio
interview from Wednesday, but since almost everyone was in class that
day, that should not require any extra effort from most of you.
|
|
|
M
11/22/04
|
HW due:
Read §6.1; write #6.1, 6.3, 6.6.
|
|
|
T
11/23/04
|
HW due:
Read Richard Morin’s lead article in the Washington
Post Outlook section (section B) from Sunday, Nov. 21. If you do not have
the newspaper in your house, then either borrow a copy from the Ellison
Library or follow this link, which is available for approximately two weeks:
http://www.washingtonpost.com/wp-dyn/articles/A64906-2004Nov20.html?referrer=emailarticle
(Note: If you follow the link, you
may be required to give your e-mail address and some demographic information
in order to receive the text of the article.)
Also, ponder and prepare a written attempt to solve each of the following
brain teasers. Half-baked answers or off-the-cuff guesses will not be
accepted. You must put some thought, and some written evidence of that
thought, into each one.
1. There are 100 passengers assigned to 100 seats on an airplane. The first
passenger who boards sits in the wrong seat. Each passenger afterward
attempts to sit in his or her correct seat, but if it is already occupied,
will sit instead in a randomly chosen seat. What is the probability that the
100th passenger to board is able to sit in the correct seat?
2. [Wording corrected 11/23/2004.] I have three pennies that are
indistinguishable to you. However, I have the ability to tell which one is
the winning penny. (Perhaps I have marked one with a spot that is visible
from the other side of a glass-bottomed table, and I don’t allow you to view
the pennies from that side. Or, perhaps the pennies were minted in three
different years, and I have previously recorded the year of the winning penny
on a piece of paper in a sealed envelope. Whatever. The particulars of the
execution are not important. The important thing is that I can recognize one
of the pennies as the winning penny, but you cannot.) We will play the
following game, and you need to decide what the best strategy is in order to
win the penny. First, I will let you choose one of the three pennies,
whichever you want. Then, I will remove one of the other two pennies from
consideration, taking care to remove a
penny that is not the winning penny. (Note that I will always do this,
regardless of whether the one you chose is the winning penny or not.)
Finally, I will let you stick with the penny you chose initially or, if you
prefer, switch to the other penny that remains. If you end up with the
winning penny, you may keep it, but otherwise, you need to pay me a penny.
What should you do?
(a) Switch to the other penny.
(b) Stick with the penny you originally chose.
(c) It does not matter.
3. In the final game, we will play with two standard dice. I will roll the
dice in such a way that they fall off the table, onto the floor where I
cannot see them. You will peek at the dice and will observe whether at least
one of the dice is a “3” (˙·.),
and if the answer is yes, we will play the game. (If not, we will not count
this as a game—I will simply roll again.) What odds should you offer me, if
you were fair, for me to bet on the outcome being a double “3”? Or, if you
don’t know how to compute odds, answer this closely related question: What is
the probability that both dice are “3” given that at least one die is known
to be a “3”?
|
|
|
W
11/24/04
|
No school.
|
|
|
Th
11/25/04
|
No school. Happy
Thanksgiving!
|
|
|
F
11/26/04
|
No school.
|
|
|
M
11/29/04
|
HW due:
First, perform a simulation, using
either real dice or your calculator’s random integer generator, to verify
that the answer to brain teaser #3 is actually 1/11 as claimed by Martin.
Record your raw data in a tally table similar to the one we constructed last
Tuesday. You will need several hundred trials in order to show strong
evidence that 1/11 is the true probability.
(Yes, I know that the dice table provides a much more efficient proof. That’s
not the point! Zero out of 9 students thought of using a dice table, and even
after seeing the dice table, almost nobody was able to apply it correctly.
However, I think all of you could have performed a simulation to find the
correct answer. Later in life, when you are faced with a difficult
probability question, I want you to consider using a simulation to find the
correct answer. Meanwhile, your coworkers will waste mounds of time trying to
come up with a “mathematically correct” solution—which, by the way, is almost
always wrong.)
Second, carefully reconsider your
answer to brain teaser #1, using either a “Gedanken” experiment or
a simulation to prove that your answer is correct. A problem similar to #1
was featured recently as a puzzler on NPR’s popular Car Talk radio show.
|
|
|
T 11/30/04
|
HW due:
Perform the HW assignment that was due yesterday, except this time do a
simulation for the second part. Be sure to write out the methodology for your
simulation, step by step. Record the raw data of the results, not an
aggregated/abbreviated record. Tim and a few others are exempt from having to
redo the first part. Nobody is exempt from the second part.
|
|