Monthly Schedule

(AP Statistics, Period B)

M 4/4/05	HW due: Read How to Lie with Statistics by Darrell Huff. If you were not in class on Friday 3/18, you will need to check out a copy from a public library or buy a copy from amazon.com or a bookstore (approx. $10). Or, if you make arrangements with me, I can lend you a copy over break. Graded Discussion today will cover the reading assignment.
T 4/5/05	Double Quiz (10 pts. + 10 pts.) on How to Lie with Statistics. The first question (10 pts.) will be an essay question in which you critically analyze a recent half-page Washington Post advertisement, using terminology from our class and/or from the book. Use of book terminology is preferred if applicable. For example, if the ad uses irrelevant data, you could criticize the data as being “semiattached” and explain why you believe they are irrelevant. If the ad uses an embarrassing survey question that people would be unlikely to answer truthfully, you could call the ad a case of “a sample with built-in bias”; hopefully, you would also write that this is a topic from last fall (namely, response bias). Hint: I would recommend being familiar with the points listed under the heading, “How to talk back to a statistic.” The other questions will be a total of 10 pts., testing how well you have read the book. Here is an example of a type of question I have used in previous years: Suppose that Republican Party affiliation in County X has grown from 24,000 registered voters in 1980 to 48,000 registered voters in 2000. A pictograph depicts this change by means of two elephants, one of which is twice as tall and twice as wide as the other. By what factor—according to the author—will readers perceive that Republican Party affiliation has increased? (A) no change (i.e., factor of 1) (B) factor of 2 (C) factor of 4 (D) factor of 8 (E) none of the above Answer: D.
W 4/6/05	HW due: Work four more problems from the review book (or your regular textbook, if you prefer) on confidence intervals, hypothesis testing, Type I error, and power (or Type II error). Try to pick one problem of each type.
Th 4/7/05	HW due: Repeat questions (a) through (e) that were due on March 15. A minimum of two illustrations per question are required. Neatness counts. If you did a good job with this the first time, or if you took neat and careful notes from our class discussion, you may not have very much work to do.
F 4/8/05	HW due: Read pp. 327-331 in the Barron's book. Then, read questions #1-9 on pp. 343-346 and determine (YES or NO) whether each one is answerable based on the reading you just performed. (Several are, and several are not.) Write your answers on your HW paper. I would encourage you also to work the problems that can be answered, but since the answers are already given on p. 347, that is not required for credit on the HW scan. In class: Use data gathered yesterday to determine whether there is evidence of a systematic lack of randomness in Smokey selections. If you are absent today for any reason, you must find a classmate to get the CSDELUXE (chi-square deluxe) program, or you must key it in yourself. Anyone who does not have this program on hand and working on Monday will lose points.
M 4/11/05	HW due: Run a complete PHA(S)TPC hypothesis test, using the larger data set from Friday. Let the p’s denote the true probabilities of Smokey selection, and let the hypotheses be as follows: H0: pB = 1/11, pC = 3/11, pafter C = 5/11, and pother = 2/11 Ha (standard version): (pB ¹ 1/11) OR (pC ¹ 3/11) OR (pafter C ¹ 5/11) OR (pother ¹ 2/11) Ha (alt. version): Probabilities are not as stated in H0. Ha (alt. version 2): At least one of the probabilities stated in H0 is not correct. Our data (n = 100) were as follows: 13 “B” students, 33 “C” students, 28 “after C” students, 26 other.
T 4/12/05	Test on Confidence Intervals, Hypothesis Testing, Type I Error, Type II Error, and Power. The study guide is the same as the one in the March 8 calendar entry, except that this time you are also responsible for Type II error and power.
W 4/13/05	Re-Test (same exact test as yesterday), plus the following HW due: First, read pp. 331-338 of the Barron’s book. Second, carefully write out null and alternative hypotheses for questions #1-9 on pp. 343-346. Third, solve questions #1-9 on pp. 343-346 if you have not already done so. Finally, revisit Monday’s HW assignment and make sure that your conclusion shows strong evidence (c^2 = 12.948, n = 100, df = 3, P < .005) that Smokey’s true probabilities are not as stated in H0. Your work (consisting of 4 terms added up to get the c^2 statistic) must also be shown. Where is the largest contribution to c^2, and what does this tell us about Smokey’s possible bias? Be very specific; refer to specific student name letters. Note: Only periods A, B, and C will meet today. Periods D, E, and F are being pre-empted for Diversity Day. HW will be collected at the beginning of the period, as you enter. I will grade it while you take your re-test from yesterday. You should be able to finish if we start promptly. You may discuss the test freely.
Th 4/14/05	No additional HW due. In class: Continue with c^2 goodness-of-fit.
F 4/15/05	HW due: Continue reading in the Barron’s book (pp. 331-338), through the end of the section on c^2 tests for independence and for homogeneity of proportions. These two methods are lumped together in the CSDELUXE program as c^2 tests for 2-way tables. All previously assigned HW is still due and may be collected. At end of class: STAtistics Jeopardy Bingo!
M 4/18/05	Old HW due: Read pp. 331-338 in the Barron’s book if you have not done so. You were supposed to have already done this, but something tells me that some students still need to do it. The responses in Friday’s Jeopardy Bingo competition were a clue. When the answer was, “It’s the df for a c^2 2-way test,” we had to go practically around the room before finally finding one student to say, “What is (rows – 1) times (columns – 1)?” New HW due (required of everyone; must be completely finished): Work the following problems. Allow 2¼ minutes for each multiple-choice problem. p. 371 #17, 22 p. 372 #26, 28 p. 373 #32 p. 376 #1 (allow about 13 minutes for this one, and show all work) Important: In addition to showing your scratch work, write down in your notebook the type of problem (e.g., 1-sample t test, c^2 g.o.f., c^2 2-way) that each one represents. Total length of new HW < 30 minutes for regular time, about 36 minutes for extended time. If you can’t handle this, I would rate the probability of our eventually adopting the MWF option as 0.000. In class: 2-prop. z test quiz (4 pts.), using an example that also works well as a c^2 test.
T 4/19/05	HW due: Read pp. 338-342 in the Barron’s book. This is the last topic of the year—hooray! Also do a full PHA(S)TPC writeup of the c^2 test that we began shortly before time ran out yesterday. I gave you H0 and Ha; if you were absent, be sure to get them from somebody who was there. The data were in a 2-way table as follows:
		Students	Faculty
	Polite	25	25
	Not so polite	50	15
W 4/20/05	Quiz on c^2 G.O.F. and 2-Way. Bring all your HW from this quarter. I will plan to grade a random selection of problems while you are doing your quiz.
Th 4/21/05	Before school: It’s JBAM (Jeopardy Bingo At McDonald’s)! HW due: Bring a bag of plain M&M’s candies.
F 4/22/05	Quiz on Linear Regression Tests. You are responsible only for tests concerning b (true slope) and r (true correlation coefficient).
M 4/25/05	FBK Day (no school).
T 4/26/05	HW due: Make study notes for yourself concerning the Must-Pass Quiz. A lame-o markup of the existing answer sheet does not qualify for credit. Oral Quiz (2 questions for each student) on the Must-Pass Quiz. You may use your notes for now, provided they are either handwritten in your own writing or stored in the memory of your calculator in such a way that you can access them quickly. I will be the judge of what constitutes “quickly.” In class: Review of Topics I and II (exploratory data analysis, experimental design).
W 4/27/05	Another Oral Quiz (2 questions for each student) on the Must-Pass Quiz. Yesterday’s rules still apply. In class: Review of Topics III and IV (probability, inferential statistics).
Th 4/28/05	AP Multiple-Choice Practice (100 pts.). Everyone will take this, regardless of whether you are planning to take the real AP exam or not. Format will be 22 questions in 49.5 minutes. Extended-time students will have 15 questions in 50.6 minutes.
F 4/29/05	Multiple Choice, Take 2. Same format as yesterday, except shorter (17 questions in 38.25 minutes for regular time, 12 questions in 40.5 minutes for extended time). If you were happy with how you did yesterday, you need not show up today. If you take both tests, the higher score will be counted. If you missed yesterday’s test for an excused absence, today’s score will be counted.

---

Return to the Statistics Zone

Return to Mr. Hansen’s home page

Return to Mathematics Department home page

Return to St. Albans home page

Last updated: 02 May 2005