| Your favorite "correlation != causation" example |
[Sep. 14th, 2013|12:59 pm]
Paul
|
I have a request to make.
Next week, my 3 Psych classes (1 AP class, 2 mixed regular-and-lower curricula levels) will cover "methods and statistics." This topic leads up to one of my favorite* assignments of the year, writing a "Study Critique" essay. Students have to find a "study" (generally NOT a legitimate peer-reviewed study, usually a magazine poll or a newspaper article) and tear it apart it on the grounds of sample bias, poor question wording, misleading statistics, left-out information, hidden confounds, etc.
One fun-to-find possible flaw in a study is "correlation does not prove causation" (CDNPC).** However, I'm noticing that many of the examples I use in class to demonstrate CDNPC are a bit dull or trivial. I'd like to "punch up" my examples and have them more relevant to current (or recent past) Big Debates and Issues.
So, do you have any favorite examples of CDNPC to share, particularly ones taken from real world arguments about real world issues?
* This is a favorite assignment mine in a pedagogical sense-- if students come out of my class with better skills at critiquing analytically, that's actually more important to me than them learning anything specific to psychology. However, the essay is actually NOT at all my favorite assignment to grade, since reading them all tends to be a bit tedious and often disappointing. It's worth it, though, for the students who do "get it."
** And yes, I've already incorporated the relevant XKCD cartoon into my materials. Also this Onion-style spoof written by a colleague of mine back in grade school.
(Oh, there are so many other things I should post to LJ! Life is moving fast. Maybe later today, but no time right now.) |
|
