Originally Posted by
MoAlpha
That’s a good formulation. The problem for the general public and a distressingly large number of professionals is the inability to distinguish between reliable data, unreliable data, and sheer opinion. Sounds like your students have mastered that part.
I work in a field where we don't usually get to do experiments. We work with crappy observational data so have evolved a lot of methods and techniques to assess and work with crap data. That's why the statistical techniques used in non-experimental fields are so much more sophisticated than the stat used in experimental fields. In physics or chem or engineering, when they get crappy data they just re-run the experiment. We don't have that luxury. (As an aside, my approach to measuring bike drag uses data that everyone else thought were too crappy to use. I just used the methods we use for crap data and because the data from speedometers and power meters and barometers is so much cleaner than the stuff I'm used to working with, the results turned out pretty well).
Another thing I tell the students is that there are those whose confidence in their opinions varies not with how much they know, but with how much they think you know. Basically, I tell them not be mansplainers.