A teacher seems to be required for for learning nearly all things. This is especially true for "school subjects," but it applies in most areas of life. Even people who are self-taught usually spurt in their knowledge and/or skills once they hook up with a teacher, coach, or mentor. Teachers seem to be necessary.
Consider the following e-mail circulated by Thomas McGowan, Chair of TLTE at UNL (e-mailed September 7, 2010).
Nebraska's recent reading scores by district---with Free and Reduced Lunch rates factored in. Clearly, the correlation is different from 0 (several conservative columnists in Berliner's home state has indicated there is "no correlation" between the two); Berliner demonstrates that the correlation is actually .95.
I think the coefficient is actually -0.94. The reason this was circulated was to point out that other factors -- like socioeconomic status -- play a very large role in learning outcomes.
My reaction to this was quite different. Having long held the view that teachers were really important and using some pretty soft data (based on awards) to view myself as a top teacher, the thought that struck me was that teachers don't matter. If I were an arbitrager and I could account for more than 80% of variances, I'd be rich! If these data are representative, there is very little variance to account for teachers. [Right off the top I'll give you that "good" teachers go to "good" districts, and these things are not really independent of one another. ]
David Fowler has repeated a study many times in which he shows CAI instruction superior to what it was compared with (some form of) "conventional" instruction. I've never understood why the CAI programs didn't just take off. There is no sense in having an instructional technology seminar dealing with superior tools when they are not adopted even when proven superior.
It also is quite clear that, when college students evaluate teaching, their evaluations are more about love (and grades) than measured learning.
Once I started asking myself to think these aberrant thoughts, I worked another tack. How many of my teachers were outstanding? How many were memorably awful? I had three that were outstanding and none that were memorably awful. How many teachers have you had? Most people guess high. You probably had 7-10 in grades K-6, with maybe a coach and a music teacher thrown in. You probably had about 20 in grades 7-9, and maybe another 25 in grades 10-12. You probably had about 25 in college. Add in a few coaches and mentors. You get an number around 100.
When I ask others this question, the number of "great" teachers is usually 5 or less, and the number of memorable "dogs" 10 or less. Most teachers were competent, but did not stand out. Further, I'm sure that some learner's great teachers were other learner's dogs. That certainly would be true of Gilbert Stork who was the most influential teacher I had. That is, Stork more than any others of my teachers imparted a way of thinking in me.
The biggest problem I see is that we don't measure learning. The currently popular high stakes tests do measure learning. However, they not only measure a slice (sometimes a thin slice) of what a learner knows, but they also are needlessly high stakes. After all, if you can take a graduate record exam any time, why should you have to take a reading test at a particular hour on a particular day? Moreover, when you use computers to deliver instruction, you can make what are essentially continuous, on-going assessments of learning. Why are we doing any of this testing business instead of tracking learners as they learn? Furthermore, if I take McGowan's data to heart, then for sure I want to teach the rich kids if my rating depends upon their scores.
I know a great deal about learning. I'm the managing author and a major contributor the the book, The Unified Learning Model. (Duane Shell has to be regarded as the principal author of that book.) That book presents what might be the best if not the only unified way of looking at leaning. The more we look at learning, the more we appreciate how little we know about teaching.
A seminar is -- literally translated -- a seed bed. I'm of the opinion that professors should never lead seminars in which they know all of the answers. I use seminars to learn. The goal here is to try to decide something about teachers that should have been decided long ago. This is too ambitious to be realistic, but there are many subgoals -- the principal one being to become aware of the available data.
We will read the following (a list that will grow) [Downloads require name and password from instructor and are restricted to course participants]:
This page is static. Report problems to dbrooks1@unl.edu.