Wednesday, July 2, 2014

Is it possible to give too much feedback?

Recently, in the course of a presentation I was giving, I made a statement that is evidently controversial:
Many conscientious instructors give too much feedback to students on their work.

(I'm thinking mainly about student essays here.) Several audience members were taken aback (and this post at Philosophers' Cocoon suggests that at least some philosophers share such sentiments). But in my own defense, here's my rationale.

  1. That students respond positively to, appreciate, etc., extensive feedback on their work doesn't entail that this helps them learn. I don't doubt that many students like getting extensive feedback. But why do they like it? My hypothesis is that for the great majority of students, extensive feedback is welcome because it justifies the grade they receive on their work. Mind you, I think students are entitled to enough information that explains their grade. But the vast majority of students see feedback as an auxiliary to our summative judgment, not as a formative judgment on the basis of which to improve their subsequent work and become better student writers. (And I hate to say it, but: Those comments you meticulously prepared? Too often the paper will be thrown in the trash or the e-mail deleted.) This is apparent from the fact that student work often shows the same deficiencies time after time. If we want students to treat our feedback as formative, this is something we have to encourage them to do — and show them how to do it. And we have plenty of ways to show students we care about their performance besides our feedback.
  2. We can give informative feedback that isn't very extensive. A grading shortcut can be just as informative as thorough feedback. A well-constructed rubric, particularly augmented with marginal grading codes, can tell students as much as narrative-type feedback. Rubrics and the like may be perceived as impersonal -- but we shouldn't think that's an entirely bad thing.
  3. The ultimate goal of teaching should be to (slowly) create autonomous learners who are increasingly non-reliant on our feedback. This is really the heart of my case against giving (too) extensive feedback. One saying I like about teaching and learning: "She who does the work does the learning." Giving extensive feedback is a great deal of work for an instructor  — and what it shows is that the instructor has mastered the task at hand  and can evaluate the quality of learning performances. But unless the instructor mandates it, the feedback generates no work on the part of the student and so does nothing on its own to help the student become better at the task in question. If our hope is that students will become more expert-like in their academic performance, they have to develop metacognition: the capacity to evaluate their own work midstream, in the very process of producing it. Only then can they loosen their dependence on our judgment on their work and begin to assimilate the standards of excellence relevant to their work so that their judgment becomes reliable.
Add to this the fact that many of us are not all that good at this sort of feedback and I think the case against extensive narrative-type feedback is a strong one. What do you think?


  1. As my comment on the aforementioned Philosophers' Cocoon post suggests, I agree with your position. In particular, I think that providing detailed comments is usually a counterproductive practice: it takes up a lot of time, runs the risk of overwhelming students with information (making it more difficult for them to identify the most pressing weaknesses in their work), and is often just outright disregarded. I've had a lot of experience with students making the exact same mistake on Paper #2 that I flagged on Paper #1. I'm also sympathetic with your second point: for various reasons, students seem better able to digest concrete grading rubrics than a paragraph of comments. Rubrics also make it clear how significant the mistakes in particular areas were and directly shows how these mistakes impacted the student's grade.

  2. Stephen Bloch-SchulmanJuly 20, 2014 at 10:38 PM

    I agree. I often find that if I suggest 25 or 30 things to improve, students look at the list, pick the easiest 2 or 3, and fix those. And the easiest 2 or 3 often do little about the overall structure or quality of the paper. Instead, I list the 1 or 2 things the student has done that has helped the paper the most (even if it is not good, something has almost always gone right) and the 1 or 2 things that the student should focus on to have the biggest impact for improving next time (even for the best papers, there are things to improve. I tell that that my comments are intended to move them one big step forward, not to address everything. E.g., if they wrote a C paper and my comments get them to a B, they (and I) have done great revising/editing/learning. Once they get to the B, then we can work on moving from a B to a B+ or an A-.


If you wish to use your name and don't have a blogger profile, please mark Name/URL in the list below. You can of course opt for Anonymous, but please keep in mind that multiple anonymous comments on a post are difficult to follow. Thanks!