|
AP Statistics / Mr. Hansen |
Name: _________________________ |
Pairs Project on How to Lie With Statistics
And P-value Calculations
For the final project of the year, which is worth 80 points, you will be organized into pairs (matched pairs?). Each pair of students (click here for your partner assignment) will read the slim book How to Lie With Statistics by Darrell Huff, preferably over spring break, and will submit the following graded items by Thursday, April 5:
|
1. |
(25 points per person) Quiz on Thursday, April 5, to test comprehension of the Huff book. Anyone who is absent on Thursday, April 5, for any reason, will have the option of taking an oral quiz or a slightly more difficult version of the written quiz. |
|
2. |
(25 points per person) Four mounted, annotated clippings from magazines or newspapers (i.e., two clippings from each person) illustrating misapplication of statistics or graphs. A mounted, annotated clipping means that you must neatly attach the clipping to a standard 8½" ´ 11" sheet of paper, with your name in the upper right corner. Circle the specious use of statistics, using blue ink if possible, and write a sentence or two of explanation in the margin. If the incident resembles one of the examples cited in Huff’s book, use his terminology to earn full credit, but of course there are many other ways that statistics can be misused. You are welcome to use any that you happen to see, regardless of whether or not they are mentioned in Huff’s book. Clippings without annotations, or clippings that are not properly mounted, will be scored at 0 points. |
|
3. |
(30 points per person) Independent evaluations of P-values from other groups. Each person will be assigned a group number (1-8) from the third quarter project list. Track down a copy of the raw data, and do the full VHA(S)TPC process to compute a P-value and a conclusion about the project. Note that you may need to read part of the methodology in order to understand the raw data. You are encouraged to work with your partner on both your P-value and your partner’s P-value, but each submission must be predominantly the work of the person who signs the paper. Do not copy P-values or associated work from other pairs who may be assigned to analyze the same experimental data. You may use Mr. Hansen as a "statistical consultant" if you are having trouble setting up your H0 or Ha. |
Collaboration with your partner is not required. You may work entirely in a vacuum if you wish. However, I strongly encourage you to check on your partner’s progress and share ideas. I will score all items except for the quiz based on the mean of the score of the two partners. For example, if you submit two wonderful annotated clippings and your partner submits none, each of you will earn 12½ points for that portion of the project. If you do a good P-value calculation (30 points) and your partner does a weak one (20 points), each of you will earn 25 points for that portion of the project. For the quiz, your quiz score stands alone.
Pairs will be assigned randomly. However, if there are people you simply could not possibly work with, write their names on the half-sheet provided, and I will attempt to avoid pairing you. Yes, I know you would like to determine who your partner is, but it doesn’t work that way.
My name is ________________________ . I normally attend (circle one) D period / F period .
I am willing to work with anyone in the class except
_____________________________________
_____________________________________
_____________________________________
_____________________________________
Check one:
c
Except for the name(s) listed above, if any, I would be willing (for 1 point extra credit) to accept a partner from the other period class.c
I prefer to have a partner only from the period that I attend.