Opening Up the Discussion on Student Evaluation of Teaching

Yesterday I gave you links to sources pointing out some reasons why teaching evaluations--as currently done at many schools--do not always provide accurate or useful data.  Thanks for the comments!

Today, I'd like to open a problem solving conversation by relating  some information  I picked up on my way to getting a Ph.D. in higher education:  A) Measurement (assessment) is far harder than it looks; B)  It's impossible to measure something unless you first identify what it is and C) There's no point measuring something without an understanding of why you want the information. 

There is a vast literature available with ideas about evaluating faculty.  Have a look here, here  and at these thoughts from AAUP.   Our problems in legal academe with misusing student evaluations is that we aren't necessarily asking the right questions of the right people in the right way.   Evaluating curriculum is very different from evaluating teaching.  I think it's more than reasonable, it's important to involve all stakeholders (students, faculty, administrators, alumni, and the general population of future clients) in decisions about what to teach and how to teach it.  I will be writing more later about how medical schools have been doing this over the past 30 years.   But its not fair (or even helpful) to confuse assessing how well an individual professor is doing her job with what that job should be.

I think it would be very helpful if every school considered its own individual goals and objectives for classroom instruction.   And I'm going to put forward some suggestions below about how that might happen.

But first, in the words of noted  legal practice guru Prof. Laurie Zimet of Hastings, lets all remember "we aren't going to solve this problem today."

So what does it mean to decide first what we should be measuring?

Well, lets say a faculty decides that every class teaching a subject on the bar exam be organized according to how it will be tested.  In Torts, that would mean a professor who spent more time on intentional torts than negligence wouldn't meet this standard.   How could that be measured?  What about syllabus review?  

Or, maybe a faculty decides that each of our students deserves to have their outside of class questions answered within two business days.   Of course we ask the students themselves--but why just by anonymous survey?  What about interviewing a small group of students and ask them for examples of times they've asked for help and how the professor responded.

I'm not recommending punitive inspections.    What about having a peer visit the class a few times during a semester and then immediately meeting with the professor to discuss what they saw?  Maybe not for every class, but what about first year professors, professors up for promotion, or where there has been a problem reported?  

Is all of this more work than glancing at a scantron sheet--yes.  But deciding what needs to be going on in our classrooms and then finding out if that's happening  is very important.

Finally--how we evaluate professors is a key component in what kind of teaching we get back (same is true with students).  In other words, we are likely to get back what we measure.

So what do you think?  What should we be doing in the classroom--and who should be evaluating it?  How can we separate evaluation of teaching with evaluation of curriculum?  And where in your institution can you find the resources to review how well what you're measuring reflects what you really want to know?