When I was in grad school, my teaching assistanceship was for a couple of introductionary graduate level statistical/quant analysis classes. I loved it, and yes, I am a dork. A current co-worker in another department was one of my students. Several other co-workers are taking an intro stats course as part of their continuing education and have been using me for help. It is a good deal, I get lunch, they get a better understanding.
I see the purpose of most introductory classes is to give students a basic sketch of the field, an overview of the relevant problem solving thought process, and an introduction to some basic techniques. To me, most intro classes seek to instill a BS detector and a thought process and if either one succeeds, the student should get a C, and if both are installed, the student deserves a B+/A-.
So what are the key take-aways in the thought process people should get from Stats 101?
* Probability is not certainty
* How to figure out what a question is asking (Is X greater than Y, Is X in between A and B, is X less than K, and does it matter?)
* How to normalize different data sets
* What is wierd/unusual/different (statistical significance)
* Uncertainty dominates stats
* What is acceptable uncertainty
* When is something obviously bullshit
Anything else I should be adding? Anything I should remove from this list?
No comments:
Post a Comment