Feedback Helps Performance But Not Learning

I’m writing this paper on what neuroscience and cognition can teach us as writers of nonfiction—who, it’s been said, write essays that “show a mind at work” without, from what I can tell, learning much about how the mind even works. A colleague of mine in the psych dept at USF turned me on to the work of Robert Bjork down at UCLA, who developed the notion of “desirable difficulty” as an aid in learning, and today I found his seminal paper on the topic: “Memory and Metamemory Considerations in the Training of Human Beings”.

What I like about psych papers—particularly ones that look at learning and cognition—is how they demistify and take the romance out of what I and my colleagues have a tendency to romance and mystify: our training human beings to be writers.

I read this:

Richard Schmidt and his collaborators … have found that … reducing the frequency of feedback makes life more difficult for the learner during training, but can enhance posttraining performance. They have demonstrated that providing summary feedback to subjects … or “fading” the frequency of feedback over trials, impedes acquisition of simple motor skills but enhances long-term retention of those skills.

Schmidt works on motor skill learning, but Bjork found correlation to verbal learning skills, too. In other words, continuous feedback (like the kind creative writing students get in workshops) feels good and makes the task of learning feel easier, but there’s plenty of evidence that it doesn’t actually help learning.

The basic point of the paper is that there are lots of ways that trainers can impede (short-term) performance during training that then facilitate (long-term) performance once training ends. Put more specifically:

Manipulations such as blocking practice by subtask, providing continuous feedback during training, and fixing the conditions of practice act like crutches that artificially support performance during training. When those crutches are absent in the posttraining environment, performance collapses.

These findings correlate completely with William Stafford’s (above) “No Praise, No Blame” pedagogy (which my former colleague Michael Martone turned me on to) that I feel comfortable using them as an irrefutable, scientific basis for the new non-workshopping workshop course I’m trying out next year.

Workshop feedback serves the workshopper more than the workshoppee, being as it is a honing of critical faculties and not creative ones. And from what I’ve seen, the demands of the academic calendar lead students to submit first drafts when what’d be more helpful not only to their work on that particular essay but to their growth as a writer in the long run would be to submit drafts they’ve spent way more time on (here I’m parroting Carol Bly).

We tend to repeat as teachers what we’ve learned as students. We also tend to perform in the classroom the practices that make us feel effective in the short term—after all, we’re evaluated after just 15 weeks, whereas what we’re hoping to produce is a classroom full of lifetime writers. How do we know that our practices were effective when our students are done with their training and we send them out in the world to be working writers?

I’m reminded of another thing I learned form Martone about his approach to UA’s mandated Learning Outcomes on our syllabi. His had only one such outcome: You will still be writing in twenty years. How does extensive feedback on first drafts help make that happen? Bjork and Stafford argue pretty effectively that it doesn’t.