I am a bit hesitant to include this on so public a venue as I feel shakey sometimes about my own statistics background. I do so because after two affirming experiences this summer in AP Stat workshops I think I am at least on the right track. I would appreciate (and welcome, I think) suggestions or feedback. Heretofore my descriptions have been shared with another member on the list who, like me, is teaching the course for only a semester. I meet my students twice each week for 70 minutes each time. Our student body is from a gifted population, but most of the students in my Data Analysis class have not been terribly successful at our core math program and/or have gotten somewhat turned off to the rigor it requires.
Here goes (the following is a copy of the note I recently sent to Lisa). I apologize in advance for its length and wandering style.
----------------- Lisa, I am working through and trying to write down the syllabus I have developed over the past few years. Unfortunately it is not in useable shape at the moment. I will give you the basic outline.
I begin with a critiques of some data-based articles and reports and get student oral feedback on their strengths, weaknesses, sources, etc.
Students read and do problems from David Freedman's Statistics book chapters 1 and 2 - this is a great source for conceptual questions but is hard for the kids. I use a technique whereby they can critiques one another's answers for 10 minutes before turning theirs in. They can incorporate any suggestions or changes they want. These two chapters do a great job of hitting the key points of good experiments and observational studies. The questions force the kids to analyze the reading.
Students must take a current topic in school (block scheduling, new inquiry day we just instituted, changing graduation requirements - any topic in their world - and, with a partner, describe an experiment or observational study they would set up to test its effectiveness. They must show me through their design that they have taken into account all the good practice things dealt with in their reading either by incorporating them or by explaining why they did/could not do so.
I have shown the COMAP video #4, segments 18 and 19 on census and survey data gathering. I could have done the Freedman stuff through the video as well but wanted to vary the presentation style.
A colleague in our research department came yesterday to talk with the students about his experience with our longitudinal study. He has done this for me before so has a set of "bad" questions he asks the students to critique and edit. He talks about good survey design.
Students are now writing their own questions. My first major project is for pairs of students to determine who, in the class, is the Most Average. As they work on this, they will have written survey questions, taken (appropriate) body measurements, and in general found out a bunch of stuff about their peers. Turns out the Sept. 6-8 USA Today (weekend edition) has a survey of teens that I hope to use. I build in some measurements (e.g. "foot area") that are likely to have a variety of interpretations in the hopes that some pairs of students will decide the variability is too great to use it (lack of validity). I have some pairs of measurements (e.g. height and armspan) that would be effectively used as a ratio (the most "square" person might be more average). I make sure that categorical, count and measurement data are included and I make sure we gather alot of it so that choices are necessary.
Along with revising the questionnaire and gathering the data students are learning about exploratory data analysis (measures of central tendency, measures of dispersion, various kinds of plots, five number summary, standardized scores etc.). I will use some data sets from Workshop Statistics for this I expect. I "teach" this still stuff through a technique called "jigsaw" where students form groups of 3 (or 4) and send a representative to each of three "expert" groups to learn either central tendency, dispersion, or graphs (I might also include standardized scores). They then return to teach their "home" groups these ideas and turn in a "home group quiz." I expect students to use all of these techniques in coming up with their Ms/Mr Most Average. This project is fun to do, some of the items are a bit wacky, but the project as a whole usually gets good results.
In past years I have had students do oral presentations with a rubric (generally class developed that stresses presentation skillls and content understanding). I have 22 students this year and am going to try a "poster session" approach instead. Students may answer questions on their posters, but my idea is that we will save some time. I think I will have the students rank all 11 posters (privately) and make some comments as to why they ranked them that way - I'm not quite sure about this though. I may just have them rank the top 4 or 5 so no one gets hurt badly.
I then move into bivariate data, regression and median fit lines and some discussion of correlation - this has been a weak area for me in the past so I am still trying to figure out what form this part will take.
Probability through conditional probability is next. I have an interesting look at tree diagrams and an application to the effectiveness of a test for HIV that adds some motivation.
I use the Johnson and Johnson Statistics text and COMAP Statistics video #5 for the foundation and assignments on confidence intervals and the normal distribution. Don't know whether I will continue to use that book this fall. Then I close with as much of inferential statistics as I can do. Usually student t, chi square, f test and paired t. I may include the Mann-Witney non-parametric test this term because it is a favorite of a colleague of mine in science (he claims that is all he sees in the literature), and I think he has finally taught it to me well enough for me to use and teach.
I have good students, though many of them are not near the top of OUR population. I have made the test optional. They have the syllabus and we are checking off those things we are covering in the course. I have commited to them that I will help them with the "left overs" if they want to try the test. I expect some will choose to take it. I am sure there will be left overs but hope that if I stress the conceptual ideas they will not be too hard for the kids to get on their own or with my help.
Hope these ramblings are useful. Let's keep talking. One think I have found are some teaching "tricks" for telescoping the topics (like the jigsaw). I am still working on making my assessments rigorous enough to push the students into deep learning.
One thing I do is to require a journal of sorts. Students must make 2 entries each week. One must be reflective about something we did in class that week. The other must be about some data they have encountered outside of class (sports teams, newspaper, science class, anything). I grade it 2 points per entry - 1 point for having it, 1 point if it is "reflective." I have found that this activity makes students connect much more to what they are learning.
--------------- Hope this is helpful (to me for feedback as well as others).
Sue Eddins Illinois Mathematics and Science Academy 1500 West Sullivan Road Aurora, IL 60506-1000 (630) 907-5966 email: email@example.com