Wednesday, September 12, 2018

On the science of teaching science


Ask a scientist what makes science so special, so effective, so powerful?
What makes science a science?
The top two answers are “data”, and “reason”.
A scientist believes that with the right amount of data and with the right approach to analyze the data science can explain everything.
Scientists generally believe that a scientific approach can be applied to any natural phenomenon, to any social practice.
Well, except one.
Science cannot be used to provide a scientific description of how scientists teach science.
When I say “cannot” I do not mean “impossible”, I mean “restricted”.
A scientific approach to analyzing how scientists teach science would require collection, accumulation, and analysis of a large amount of data; a scientific approach would require collecting as much data as possible because in science, no amount of data is ever enough – one or two new facts may force the paradigm change (it happened!).
But take any college or a university and check what data does administration collect on the teaching practices of the faculty? All you find is the set of standard questions in the end-of-a-course student evaluations. And that’s that.
Those evaluations serve two goals: (1) to display that the college or university cares about the quality of teaching (“See, we listen to our students!”); (2) to ensure that a faculty does not stimulate in students a large amount of hatred toward the college (“Not great at teaching, but OK”).
That is why all college and university administrations limit the set of the questions in evaluations to the bare minimum. Why bother if the answers do not make any difference, anyway?
Data coming from end-of-a-course evaluations are vague and limited. But they are data! Even they could have been used for some analysis. But they aren't. No comparison of any sort; a year to a year, faculty to faculty, a department to a department, a college to a college. Hence, there is no system in professional development of faculty. Of course, there is an official, or a faculty, or even a structural entity which purpose, supposedly, is helping faculty becoming better teachers. It regularly issues advises, links, or invites to talks. But no one assesses the impact of those actions in any way, and no one has a specific data-based strategy for professional development, so, in the end, it is just a facade ("See, we work on professional development of our faculty").
If no one analyses even that limited data which comes from limited and vague evaluations, no one spends any energy on designing more informative assessing techniques.
Could it be done?
Of course!
Why it is not happening?
Because no one wants it – not administration, not faculty, not students, not parents, not accreditation agencies, not the government, not politicians.
There are many possible reasons for WHY no one wants to conduct actual scientific study of how faculty teach (e.g. read “A Continent Lie” or “What Research University Faculty Tell Themselves About Their Teaching”).
Reason #2, of course, is that there are only few faculty who are actually good at teaching and making this fact official could shatter the system (reason #1 is that no one really cares about the quality of teaching at a college or a university level if that is a research college or a university; teaching practices at public or for-profit colleges or universities are heavily guarded by local politicians or the owners).
But there are faculty who are good at teaching (I am one of them).
Hence, there are faculty who are not afraid of sharing their evaluations (here are mine).
Plus, there are faculty who could share their evaluations anonymously (for the sake of science!).
If some college or a university would start collecting and analyzing the data coming from those evaluations, it already could lead to invaluable insights on the teaching practices at a college and a university level. That could, in turn, lead to the improvement of the teaching practices of more faculty; which would lead to more data, etc. A positive feedback loop could eventually break the current restrictions self-imposed by administrations on designing, collecting, analyzing, and sharing data on teaching practices of faculty.
Who knows, maybe even teaching practices of scientists teaching science could eventually become a science, too?