My internet search skills have failed me, and I was hoping someone here
might be able to help me find a paper (or maybe it was a blog post???) I
remember reading a while back. It was about the benefit of giving intro
lab students plausible but false models, to help them get out of the
verification mindset and think about falsification. In particular, it gave
an example of air drag -- you tell the students, completely straightfaced,
that a falling object with air drag will clearly accelerate more slowly
than an object without air drag, so they are supposed to figure out what
"a" is in 1/2at^2, since it will clearly be less than "g". Some students
will go to great lengths to verify that the height vs. time graph is a
parabola, when they should be disproving that an object falling with air
drag undergoes constant acceleration.
Anyway, I've done that lab many times, and it's great, but I have lost the
original article from which I got the idea. I remember the author gave
some more examples of "plausible but false" models the students could test.