I've been doing this for about 14 years now. I teach introductory biostatistics in a School of Public Health. Since this is a graduate school, my problems are somewhat different from others, but most of the issues are the same.
None of the students is in love with statistics; they have put it off until graduate school.
Most of the students have trained in nonquantitative professions or disciplines.
They are often clinicians, nurses, nutritionists, etc, as well a physicians. This means they are experienced in dealing with individuals but have no concept of the measurement of or inference about a population.
The one characteristic that they most have in common is that they are usually itelligent and competent in the profession or discipline they have trained in before. Somewhat surprisingly, they are often unsure and uncomfortable in taking on biostatistics.
Another disadvantage is that they have all had a math class a long time ago and where they learned very bad habits and impressions of what math is and they can get quite upset when a graduate statistics class isn't what they know it should be. By this, I mean that they expect set-piece problems, which have known solutions, often in the back of the book, and they expect to succeed by doing problems of that sort.
I found TBL to be an especially valuable technique because it overcomes several of the student preconceptions and anxieties with statistics. But don't misunderstand. The students often dislike the whole approach at first, and a minority are actively hostile throughout the class and this is reflected in course evaluations. Others in this group have made similar comments.
The students learn more this way. The better ones take on much bigger challenges than they would with traditional didactic methods. They appreciate the opportunity to learn more and become more professional.
My principle innovation is giving exams following the same pattern as the RAs, namely individual exam followed by a group exam, after which I distribute an answer key for them to grade their own group exam. No one leaves the room with any doubts about how they did.
I found it hard to get started doing this in statistics because there was no one else with experience, though many others have used TBL with other sciences. I think the best analogs to biostatistics are physics and economics; I think biostatistics, at least as I teach it, falls somewhere in between.
I strongly recommend TBL for the standard material of statistics. I have some additional group activities that involve analysis problems of a bigger type and writing activities, which I now regard as the most critical activities in my class.
Statistical work requires clear written and oral communication, in both directions, using standard English and standard statistical and mathematical notation. I got this, and some other ideas, from the Reform Calculus movement which has been going on since the 1990s.
Regards,
David Smith
[log in to unmask]
Biostatistics Division
University of Texas School of Public Health
________________________________
From: Team Learning Discussion List on behalf of Smruti Patel
Sent: Tue 6/17/2008 8:25 PM
To: [log in to unmask]
Subject: TBL for Statistics course
I'm using TBL for the first time (having just "discovered" it a couple of months ago) for my Statistics courses this summer. I just had my first class today and it was great!
I was wondering if anyone else is using TBL to teach Statistics and if I could pick your brain, get some advice, steal some ideas, etc ....
Thank you,
Smruti
:)
|
