Why Feedback Often Doesn’t Help

By Alfie Kohn

[This is an expanded version of the published article, which was given a different title.]

A lot of people make a living by offering advice about how teachers should give feedback to students — or how administrators should give feedback to teachers. Unfortunately, a body of compelling theory and research raises troubling questions about the value of much of that advice. It turns out that hearing how well we’ve done (typically from someone in a position of power) often doesn’t lead us to improve.

The word feedback, which originated in the world of mechanics, refers to a self-regulating system like a thermostat in which output affects input. Thus, when applied to human communication, the word would seem to apply only to information — as in “Here’s something you did that I noticed…” When feedback is contaminated with evaluation (“Here’s what I think about what you did…”), it tends to become not only less effective but often downright damaging — both to future performance and to recipients’ interest in whatever they were doing.

For decades, studies have shown that praising people when they succeed can be just as counterproductive as criticizing them when they fail. Nor does it help just to tweak the phrasing or to praise one thing rather than another (for example, effort rather than ability). In the 1980s, researcher Ruth Butler found that students often became more intrigued by a task when they received simple comments describing what they had done, whereas praise “did not even maintain initial interest at its baseline level.”¹ More recently, two Vanderbilt researchers reported that students, particularly those who were reasonably proficient, did more poorly at math if they had previously received praise for succeeding.²

What is true of the judgment inherent in praise is also true of the judgment inherent in grades. A series of metaanalyses published in 2020 by Duke University researchers showed that substantive feedback without any grade attached was preferable for promoting both motivation and achievement. “Compared to receiving grades…or grades along with comments…comments [alone] led to heightened continuing motivation” — a whopping 1.5 standard deviations higher, in fact. And students who received comments without grades also showed higher achievement than those who received grades without comments. In fact, getting a grade was more damaging to motivation than receiving no feedback at all, particularly for struggling students.³

Notice that the relevant question here isn’t whether students got high grades or low grades. The problem is with grades, period. That’s true for the same reason that praise isn’t much better than criticism: The most striking feature of a positive evaluation is not that it’s positive but that it’s an evaluation. (One psychologist remarked that kids would likely come to find it unpleasant to watch TV if they were regularly evaluated for how well they did it.)

But the central point here applies to adults as well as children, which is why teachers often bristle at having an administrator sit in judgment of them. Some consultants have argued for a model of peer coaching that excludes offering explicit feedback — precisely because evaluation, which is counterproductive, has a way of slipping into the feedback.⁴ (What’s remarkable, when you think about it, is that some of these same teachers, understandably resentful at being judged, then turn around and subject their students to a constant stream of evaluations and expect it to help them.)

So why do evaluations backfire? First, because they, like other rewards and punishments, are often experienced as controlling — and people don’t like to be controlled.⁵ Second, to receive a pat on the head (an A or a “Good job!”) for doing a task well serves to devalue that task; it’s now been reframed as just a means to an end, a prerequisite to receiving a reward. Finally, evaluation creates pressure to keep up the good work, which, in turn, leads to risk avoidance. If the point is to perform well, better to stick with what one is likely to succeed at — a posture that’s not exactly conducive to learning or growth.⁶

Feedback, then, is better than evaluation. Unfortunately, that doesn’t allow us to conclude that feedback is always constructive. In fact, a comprehensive review of the research, comprising more than 600 experimental comparisons, found that even pure feedback had a negative effect on performance “in over one third of the cases.” And even when the effect was positive, “its utility may be marginal” or it may create “only shallow learning.”⁷

That result will be as surprising as it is unwelcome to anyone who has read, and accepted at face value, sweeping assertions about the overwhelming positive impact of feedback. When you collapse all the available studies into a single summary statistic, yes, feedback seems to improve performance. But that’s misleading for several reasons.

First, such a reduction overlooks all the variations in how it’s given and, equally important, how it’s received. Depending on these details, feedback can produce results that are often no better, and sometimes even worse, than receiving no feedback at all.

Second, such claims often ignore feedback’s effect on something of arguably greater importance than achievement on a given task — namely, a learner’s interest in that task (or in the topic, or in learning itself). Feedback’s effect on that outcome is even more likely to be disappointing than its effect on achievement: As one group of researchers reported, even “communications that provide clear, positive feedback about an individual’s competence at an enjoyable activity may not always have their intended effects on self-perceptions of competence, task involvement, and continuing motivation.”⁸

Finally, the eminent assessment expert Lorrie Shepard noticed that most studies of feedback “are based on behaviorist assumptions. Typically the outcome measures are narrowly defined, feedback consists of reporting right and wrong answers, and the end-of-study test may differ only slightly from the prior measure and from instructional materials.”⁹ Thus, even when feedback does seem to “work,” it may do so only on tasks of dubious value, such as cramming forgettable facts into short-term memory. An overview of the research doesn’t allow us to conclude much about whether feedback promotes meaningful learning.

Given the variations in results, are there any guidelines that can predict whether, and to what extent, feedback might really be beneficial? I’m glad you asked.

* All else being equal, hearing that you succeeded at a task is more apt to strengthen interest than hearing that you didn’t.¹⁰ (The supposed benefits of failure are wildly overrated.)

* Sometimes it’s obvious whether your efforts paid off: Either the seed you planted sprouted or it didn’t; either readers are surprised by your ending or they aren’t. Such feedback is more likely to enhance interest than having someone else tell you how well you did, which pulls you out of the learning experience.¹¹ Students are then less engaged with what they’re doing and more concerned with how well an individual with more power thinks they’ve done it.¹²

* Feedback is likely to backfire when it’s given publicly, or in comparison with other people’s performances. Contrary to a widespread American myth, competition tends to undermine both intrinsic motivation and achievement — for winners as well as losers.

* Feedback works best when it’s just one step in a learning process rather than a final judgment, although even the formative kind isn’t always beneficial (particularly if it’s based on a test rather than a more authentic type of assessment).

* It matters not only how feedback is given, but why. If the rationale is experienced as manipulative (to meet someone else’s standards), it may be damaging. The ideal scenario is for information to be offered at the recipient’s request. In general, effective teachers and managers do a lot less telling and a lot more asking: “How can I help?” “What do you need to know?”

As with so much else in education, paying too much attention to perfecting a method can distract us from reflecting on our goals. Moreover, it makes sense to consider not only the quality of learning but the experience of the learner. That’s why I keep thinking of a confession once offered by educator Cris Tovani: “I was concentrating so hard on…trying to perfect the feedback…[that] I forgot to focus on the kids.”¹³

NOTES

1. Ruth Butler, “Task-Involving and Ego-Involving Properties of Evaluation,” Journal of Educational Psychology 79 (1987): 474-82.

2. Emily R. Fyfe and Bethany Rittle-Johnson, “Feedback Both Helps and Hinders Learning,” Journal of Educational Psychology 108 (2016): 82-97. In a follow-up study, the researchers tried “taking the praise out of the feedback and just stating the correct answer.” Here, the results were even more striking: Such evaluation-free feedback produced a positive effect even with children who had high prior knowledge (Emily Fyfe, personal communication, January 2016).

3. Alison C. Koenka et al., “A Meta-Analysis on the Impact of Grades and Comments on Academic Motivation and Achievement,” Educational Psychology 41 (2021): 922-47.

4. Beverly Showers and Bruce Joyce, “The Evolution of Peer Coaching,” Educational Leadership, March 1996, p. 15. “Remarkably,” they added, “omitting feedback in the coaching process has not depressed implementation or student growth.”

5. Some early research found that praise is most likely to undermine interest when it’s presented in a controlling fashion. For example, see Richard M. Ryan, “Control and Information in the Intrapersonal Sphere,” Journal of Personality and Social Psychology 43 (1982): 450-61.

6. For an extended discussion of these three explanations, along with supporting research for each, see my book Punished by Rewards (Houghton Mifflin, 1993/2018), esp. pp. 76-81, 98-101.

7. Avraham N. Kluger and Angelo DeNisi, “The Effects of Feedback Interventions on Performance,” Psychological Bulletin 119 (1996): 254-84.

8. Judith M. Harackiewicz, George Manderlink, and Carol Sansone, “Competence Processes and Achievement Motivation: Implications for Intrinsic Motivation,” in Achievement and Motivation, ed. by Ann K. Boggiano and Thane S. Pittman (Cambridge University Press, 1992), p. 134.

9. Lorrie A. Shepard, “The Role of Assessment in a Learning Culture,” Educational Researcher 29 (2000), p. 11.

10. For a partial list of studies confirming this rather unsurprising fact, see Roberta K. Deppe and Judith M. Harackiewicz, “Self-Handicapping and Intinsic Motivation,” Journal of Personality and Social Psychology 70 (1996), p. 869.

11. The late research psychologist Mihaly Csikszentmihalyi put it this way: What’s important is for people to pay “attention to how they are doing in relation to their goals within the activity, and they learn to read that feedback without anybody having to tell them. That is necessary to keep focused on the task; without it, you wouldn’t know what to do. However, external feedback tends to involve the self. When you make children aware of whether they’re doing better or worse than they’re supposed to, then [they] switch completely from attention to the task to attention to the self — and other people’s opinion of [themselves]” (personal communication, June 21, 1993). For more on the contrast between a task orientation and an ego orientation, see the work of John G. Nicholls.

12. For a review of research demonstrating the counterproductive effects of leading students to focus on the “how well” (their achievement or performance) as opposed to the “what” (the learning itself), see chapter 2 (“Getting Motivation Wrong”) of my book The Schools Our Children Deserve (Houghton Mifflin, 1999).

13. Cris Tovani, “Feedback Is a Two-Way Street,” Educational Leadership, September 2012, p. 48.

To be notified whenever a new article or blog is posted on this site, please enter your e-mail address at www.alfiekohn.org/sign-up .