|
Th 5/1/03
|
Multiple-Choice Review of the Entire Year. By your vote, this quest (originally scheduled for
50 points) will be replaced with an intensive in-class review, followed by a 45-minute
homework assignment that is to be taken seriously. I reserve the right to
reinstate the grading of the quest, and a final exam if necessary for those
whose average slips below 80%, if the class as a whole fails to do adequate
work on the homework assignment.
You will need to answer 20 questions in 45 minutes. This is exactly half the
length of the AP multiple-choice portion. Please note that this assignment is
required of everyone, not merely those who plan to take the AP exam.
|
|
|
F 5/2/03
|
HW due: 20-question (45-minute) assignment from
yesterday. Follow the instructions
on the handout, and use the bubble sheet provided for your answers. When you
have finished, check your answers against these: BCEAD CDEEC EEDBE DEEDE.
Show scratch work on a separate sheet, and write a sentence or two of insight
for each of the problems you missed.
During class, we will begin our free-response review for the AP exam. Again,
your homework assignment will consist of a reduced-length AP simulation.
Allow yourself 51 minutes total for the three problems. As before, this
assignment is required of all students, not merely those who are planning to
take the AP exam.
|
|
|
M 5/5/03
|
Class will meet in Room
R today.
HW due: 3-question (51-minute) assignment that was handed out Friday. Show
all work, neatly and in AP style, on separate sheets of paper. Compare your
answers against the solutions on Barron’s pp. 465, 466, and 443. [There was a
typo on the handout; 433 should have been 443.]
For those who missed class Friday, the assignment was as follows:
1. Problem #1 on p.458 of the Barron’s AP review book. Allow approximately 13
minutes.
2. Problem #3 on p.459 of Barron’s. Allow approximately 13 minutes.
3. Problem #6 on p.436 of Barron’s. Allow approximately 25 minutes.
I would be most grateful if several students would give me advance copies of
their assignment before the start of A period. That way, I would have time to
make overhead transparencies or handouts that we could discuss (with names
removed, of course) in order to practice our critical thinking and “holistic
grading” skills. There will be an extra-credit bonus available for those who
comply with this request.
Also today: Start reviewing for your Must-Pass
Quiz. AP exam takers will find this to be a useful review sheet, but
everyone will need to pass this before May 23.
|
|
|
T 5/6/03
|
HW due: Handout from Monday, 5/6. (This handout was placed in your mailbox if you were
not in class). If you have lost your handout, or if you did not have a chance
to visit your mailbox today, here is the assignment:
1. Read the College Board’s Notes
on Student Performance.
2. Mark up the formula
sheets (pages 3-5 in linked document; requires Acrobat Reader). This task
is made vastly easier if you follow the handout I placed in your box last
Friday.
3. Read the free-response questions from 2002,
2002
Form B, and 2001,
thinking carefully about how you would approach them.
4. Under time pressure (13 minutes each for #1-5, 25 minutes for #6), do several of the questions from 2000 and
check your solutions on the AP
Web page. (Click the Scoring Guidelines link for each of the questions
that you worked on.)
In class: We will continue our review by working on free-response
questions from years 2000 through 2002 and by discussing the Must-Pass Quiz.
|
|
|
May
Madness!
|
Yes, Liam’s wild no-holds-barred review session
style is back. Join him approx. 7-9 p.m. Tuesday, or all morning Wednesday in the AV room,
as he jams the maximum amount of AP Statistics information allowed by law
into each attendee’s brain.
|
|
|
W 5/7/03
|
AP Examination, Trapier Theater, 12:30 p.m. sharp.
Bring pencils, calculator, and spare batteries only. If you arrive late, you will
be denied admission to the test and must report for class as usual.
Note that class will be
held today as usual for those not taking the exam. For people in that
situation, this would be a good opportunity to get the Must-Pass Quiz out of the way. There is a partial answer key available to help you
prepare.
|
|
|
Th 5/8/03
|
In class: We subject class
skippers to an on-the-spot administration of the Must-Pass Quiz. One failure ensues. But then
Ted M. hits one out of the park, becoming the first STAtistician to pass the
MPQ for 2003. Will W. follows suit with a borderline score (judged on appeal
to be a pass).
|
|
|
F 5/9/03
|
HW due: Re-think your
experimental design proposal in light of what we have learned this spring.
Decide what type of C.I. and/or test you will compute with your data: 1-prop.
z, 2-sample t, c2, or whatever.
We will perform some sample size and power calculations before we go forth to
gather data.
|
|
|
M 5/12/03
|
HW due: Now that you know
(approximately) what type of test you will use for your experiment, write
your null and alternative hypotheses and estimate the size of the effect you
wish to be able to detect. In class, we will perform power calculations to
determine the sample size that will be needed in order to achieve good power
against your estimated alternative.
|
|
|
T 5/13/03
|
HW due: Written analysis of
power. Marques’ group is using a chi-square test, so their assignment is to
state a believable 2-way table that might result if their alternative
hypothesis is true, and show that their test will have good power against
that alternative. All other groups must (1) state null and alternative
hypotheses, (2) state a specific alternative against which good power is
desired, (3) state numerically what you consider to be “good” power, and (4)
using a graphical and calculator analysis similar to what we did in class
yesterday, calculate the sample size needed to achieve your desired power. If
your group leader is not present today, he should have deputized someone else
to make sure that this written assignment is turned in on time. (If not, all group
members will receive a zero on this portion of the group project.)
Note: It is perfectly acceptable to use approximations when estimating the
required sample size. For example, we “guesstimated” the value of s =
6 bpm for the in-class example involving the s.d. of the difference between
two closely spaced measurements of heart rate.
Because the AP exam is now behind us, I will permit “calculator notation” in
your written power analysis.
In class: This did not go exactly as planned. Will and I waded into a swamp
involving the power calculation for his group’s extended-time experiment. The
main error we made was that the critical value (z*) must be converted to “question units” by the procedure of multiplying by s.e., not dividing as
we erroneously did in the midst of that horrendous invNorm expression.
Tomorrow, focus on your methodology and simply assume that you will have a
sample size of 25. Will and I will come around and perform some rough-cut
power calculations to verify that 25 is reasonable.
|
|
|
W 5/14/03
|
HW due: Final version of
methodology. This will be worth a maximum of 50 points per group member.
Point allocations will be determined by recommendation of the group leader.
If you wish to use the class period for soliciting feedback (and power
calculations) from me and other class members, that is fine; however, you
must bring a neatly typed, almost-ready-for-prime-time version of your
methodology. No more rough drafts, please! If you wish to make one last round
of modifications, I will accept the methodology document without lateness
penalty at the start of class tomorrow.
|
|
|
Th 5/15/03
|
Shortened class: We will
meet at 12:35 sharp to take roll, then disperse to gather data
and/or hold group meetings. I will be available for people who need a
statistical consultant. If you have not already taken the Must-Pass Quiz, today is your opportunity.
Today turned out to be the most exciting day in the entire history of
STAtistics: Alex K. passed the Must-Pass Quiz
on the first try! Bravo, Alex!
|
|
|
F 5/16/03
|
Similar to Thursday (roll
call at 12:45, followed by shortened class with free schedule).
|
|
|
M 5/19/03
|
HW due: Group leaders will orally
present the results of their data gathering effort so far. If a group leader
is absent for any reason, including sickness, he must designate a deputy to
fulfill this role.
|
|
|
T 5/20/03
|
Field trip to the National Cryptologic Museum. Bus will
depart promptly at 8:00 a.m. and will return to STA at about 1:00 p.m.
Where to board: Bus will load in front of the Lower School. We will try to send runners to a few other places
around campus to round up people who may be in the wrong place.
What to wear: Free dress, but please remember that you are
representing STA. I reserve the right to deny boarding for sloppiness. To be
safe, please wear a shirt with a collar, and avoid sandals, cutoffs, and
soiled or torn clothing.
|
|
|
W 5/21/03
|
Roll call at 12:35, followed by shortened class with free schedule.
Time is running short if you have not already passed the Must-Pass Quiz. Today would be a good day to
take care of this.
|
|
|
Th 5/22/03
|
HW due: Final written and oral group reports. Scoring will be a maximum of 75 points per group
member. Please keep the following in mind:
- Recommended length, including statement of
research question, methodology, analysis, and conclusions, is
approximately 5 to 6 pages plus appendices. Please use subheadings for
each section. You are free to alter these subheadings or to use
additional subheadings if appropriate for your project.
- You may “recycle” your approved methodology
section without modification.
- Omit the power analysis that you turned in last
week, since the reader will assume that your choice of sample size was
either arbitrary or the result of a power calculation. These details do
not belong in the final report.
- List names of all group members on the title
page, and identify the leader (e.g., by underlining).
- Please spare me the literary verbosity, and
stick to the statistical “meat” of your project. If you have a great
deal of philosophy to impart, feel free to include it in your oral
presentation (see below).
- Avoid passive voice and puffery: Say, “We
measured height and weight for 23 subjects,” not “The linear heights
above ground for the 23 subjects, both experimental and control, along
with the gravity-force weight upon the planet Earth of the selfsame
aforementioned 23 subjects, both experimental and control, were duly
measured, recorded, and tabulated in writing and were also duly
measured, recorded, and tabulated in electronic personal-computer
digital format by the group consisting of the names found upon the title
page of the present document.”
- Charts, if used, may be in an appendix or,
preferably, in the body of the report. Use a medium-size format, usually
no more than 4 in. wide by 3 in. high. Charts are not necessary for full
credit, but in many cases they are the best way of summarizing data and
error bars. Computer-produced charts are fine, but pen-and-ink charts
are also perfectly acceptable.
- Raw data tables are necessary for full credit.
Place these in an appendix.
- If you used human subjects, attach release
forms in an appendix.
- If you used a data-gathering instrument, e.g.,
a survey form or quiz form, include a blank copy in an appendix.
- I generally prefer black-and-white. If you use
color, make sure that the color adds clarity and is not gratuitous.
- For each figure, table, and appendix, be sure
to number it, label it, and
mention it at some point in the text, in an optional table of
contents, or both.
- Spelling, neatness, and grammar all count.
Please proofread carefully.
- Long projects generally score lower than short
ones, unless the subject matter truly warrants in-depth treatment.
- Each group leader must submit one paragraph or
more, on a separate sheet, recommending an appropriate point split and
giving well-supported justifications. (It is not adequate to say, “We
all worked equally.”) The group leader will forfeit 10 points if this
sheet is missing.
- You may submit early draft versions of your
written project for comment. If I delay your schedule by taking longer
than expected to respond with feedback, you may extend the due date on a
day-for-day basis.
- A designated person from each group—not
necessarily the group leader—must make a brief oral presentation to the
class today, regardless of the state of completion of your project. If
possible, please provide your raw data as an Excel file or as TI-83
lists so that we can independently verify your P-value calculations and/or confidence intervals.
|
|
|
F 5/23/03
|
Last day of school / last chance to pass the Must-Pass Quiz.
If you have not yet passed the Must-Pass Quiz,
and if your average is below B (80%), you must take a final exam. This is
true even if you took the AP exam.
If your average is B or above, but if you have not yet passed the Must-Pass Quiz, then what are you waiting
for? Surely you would like those points before the end of the day.
For those who have passed the Must-Pass Quiz,
you have earned yourself a boost to your quarter average. Congratulations and
have a happy post-STAtistics life . . . but remember that I may ask you some
of those “*” questions when you return for your 20-year class reunion in
2023!
|
|