A certain moment will play out thousands of times over the course of a primary teacher's career. If you're a secondary teacher and have children of your own, you know it, too. It's the moment when a 4-year-old child stands in front of you, paper held in front of him, demanding, "Look at my drawing!"
Sometimes you can tell what the drawing is intended to represent; sometimes not. Either way, it's a moment of creation and relationship, a moment that both sets the stage for the educational process and defines it. The elementary education majors in my literacy methods classes treasure this moment because it makes them feel needed—a child cares what they think!—and because the child and the drawing are so cute.
But what we do and say in this moment matters. And it's worth digging into because it reveals much about how feedback affects learning in all subjects and grade levels.
The Limits of Rewards and Punishment
To analyze this moment, we have to pause it. There's the child, face upturned, the demand—"Look at my drawing!"—which is really a question—"What do you think?"—hanging in the air. Before we respond to this child, we have some work to do. There is, after all, an art and a science to teaching, and we have more than instinct to rely on. Both research and our values will help us respond in a principled, informed way.
For instance, we know that grades tend to undermine creativity, risk taking, and performance (Butler & Nisan, 1986; Pulfrey, Buchs, & Butera, 2011). When we imagine how to respond to the essays this child will hand us in a few years, this conclusion about grades might be more difficult to accept. After all, assign, collect, and grade seem to go together like baseball, apple pie, and Chevrolet. But we're not balking yet: No one in her right mind would be tempted to grade this 4-year-old's eagerly shared drawing.
We know that formative assessment, unlike the one-dimensional judgment communicated by summative assessment and grades, supports learning by giving the student descriptive feedback about his abilities, performances, or products (Chappuis & Chappuis, 2007/2008). But the kind of descriptive feedback we give matters, too. Carol Dweck's well-known research found that feedback directed at innate ability—"You're so smart!"—often backfires because it reinforces a fixed concept of intelligence, which discourages risk taking (Dweck & Mueller, 1998).
Now we may be starting to balk a bit. Doesn't a caring teacher tell students how smart they are? But no real worries yet; not one of us is seriously tempted to tell the child that his drawing makes him "smart." Perhaps we can bypass the fixed-intelligence trap but satisfy our urge to praise by following Dweck and Mueller's recommendation and focusing on the child's efforts: "You did such a good job of working hard on that drawing!"
But it's too early to put our child back into motion with this cheerful bit of feedback. More research complicates our simple substitution. As Alfie Kohn (2012) reminds us, Dweck herself doesn't cling to the idea that teachers or parents should simply praise effort instead of praising smarts. In fact, several studies indicate that both kinds of praise often undermine performance and interest. In one simple but telling study, Leann Lipps Birch gave children a kefir, a yogurt drink. Children who were offered cash or praise consumed the most in the moment. But when the drink was offered to the children a week later without any conditions, those originally given no prize or praise drank the most, while those who had been induced with money or compliments drank the least (Kohn, 1999).
Now we're feeling confused, even a little angry. A century's worth of behaviorist education theory and practice has led us to believe that rewards and punishments are our only options. When confronted with behavior we like, we praise and reward. When confronted with behavior we don't like, we reprimand and punish. Certainly, none of us wants to reprimand the child standing in front of us for sharing his drawing; but if praise is likely to discourage him from drawing in the future, what's an informed and caring teacher to do?
The Power of Relationship
If we're willing to step out of our surprisingly deep behaviorist rut, the answer is clear. Children don't need prizes or praise, but engagement and relationship —with other people and with the work they are doing. If you love a child, you probably know this on a gut level. But the research confirms it.
In one study, children were given a book with a marker placed 250 words into it. Some of the children were offered a prize to read to the marker. Others were told that they'd have the chance to share their opinion of the book with an adult when they'd read to the marker. The promise of a reward created compliance: The rewarded children read to the 250 word mark—and no further. But the promise of engagement with an adult about the book inspired children's engagement with the book: The unrewarded children were twice as likely to engage in the reading, and they read twice as much as the rewarded children (Krashen, 2003).
The alternative to the behaviorist model of rewards and punishments, in other words, is a relational model of engagement—children engaging with adults and one another around ideas, activities, and knowledge. This model helps us understand another danger of praise—it can be a way of disengaging, a quick trick to end the interaction without appearing to be uncaring. Kara, one of my elementary education students, refers to praise as the gun in the teacher's holster. A child demands our attention, and we point and shoot—"Good job!" We turn away from the child, blowing smoke from the barrel. We've said something "positive." But we've just killed the conversation.
If the research doesn't convince you that the praise gun can be lethal, a moment's reflection should: Think about going home and telling a family member about the great fourth period class you just taught. You don't want him to say, "Good job!" and then keep watching Oprah. You want him to ask questions, to help you figure out the reasons the class went so well, to find out how you're going to follow up tomorrow. In short, you want others to engage with you, not to praise you.
Three Goals for Feedback
We're almost ready to return to our poor paralyzed child, drawing still grasped firmly in hand. We don't want to praise, and we want to engage with the child about his work. But what are we trying to accomplish?
We generally think about feedback only in terms of its effects on the student's product or performance—Did this feedback result in a better drawing? Psychology suggests a second consideration—Did this feedback support the student's healthy view of himself as an artist? I propose that a third consideration is the effect of our feedback on the student's conception of the activity, skills, or knowledge he is learning to master—Did this feedback result in an understanding of what it means to draw (substitute any subject-specific activity here) that the student can take with him and use from year to year?
A simple request—"Can you tell me about this?"—can set everything in motion. The 4-year-old I have in mind has a name—Jonah. I'm talking with Jonah as he draws, so I can see our interaction playing out as he works on his illustration.
Maja: Can you tell me about this?
Jonah: I forgot what it's called. It looks like … a flashlight!
Maja: It does look like a flashlight! There's that white part; that looks like the light coming out of the flashlight.
Jonah: But it's not a flashlight. It's the kind of thing that pokes. If you touch this side, it pokes you. Ouch! It pokes, see? It's a porcupine! Have you seen the movie Shadow?
Maja: No—can you tell me about it?
Jonah: [Speaks too softly to be heard.]
Maja: Does the movie title mean shadow like the shadow of your cup on this table?
Jonah: No, no! Shadow is a dog. There's a porcupine in it. But Shadow doesn't get porced by the porcupine. [Finger taps the upper right-hand edge of the drawing.] This is a porc, porc, porc! [Pause.] This doesn't look like a porcupine. It looks like a flashlight. [Adds lines to fill in the white part.]
Jonah: Now it looks like a porcupine. What does it look like to you?
Maja: It looks like a flashlight-porcupine!
Jonah: It is!
In this brief interaction, two of my responses give Jonah information about what I see in his product, but also end up giving him information about the nature of the work that he's engaged in:
- "It does look like a flashlight! There's that white part; that looks like the light coming out of the flashlight!"
- "It looks like a flashlight-porcupine!"
Without Jonah's guidance, I might have looked at his finished product and guessed that the porcupine was the sun, or an insect. But I wouldn't tell him that right away. Notice that each of my responses comes after Jonah asks some version of, "What does it look like to you?" or after he asserts what he intends. Once he orients me, I affirm that I can see the traces of his intention in his marks.
I cringe when adults take over children's drawings. There's a time for adults to show children how to do something. But not when the child is in the middle of figuring it out, moving back and forth between the image in his mind and the image on the page ("This doesn't look like a porcupine. It looks like a flashlight! "). The point isn't getting the porcupine right. The point is that Jonah gets good at exactly what he's doing in this interaction: forming an intention, working with the medium to try to give that intention shape, viewing the product and inviting an audience to report on what they see, deciding whether what others see matches what he intended, and figuring out how to solve the problems he's noted from this comparison. These are skills that creators in all media develop and take with them from project to project.
My feedback may not appear to directly help Jonah improve his product. But in the process of talking with me about how his drawing looks like a flashlight and not a porcupine, Jonah begins to narrate his interaction with the porcupine. Stabbing his finger into the edge of the shape he's drawn, he stumbles on the solution to his problem—as his pencil mimics his finger, he creates the porcupine's quills.
Not only is Jonah an active participant in his own learning here, but he is also fully engaged in the work that artists do. And his product improves: his porcupine begins to look less like a flashlight and more like a porcupine.
From Porcupines to Lily Pads
Let's shift from a 4-year-old's drawing to a teenager's essay. How might feedback to student writers accomplish our three goals: help the product improve, support a healthy view of the self as an active learner, and help the student develop an understanding of what it means to compose text? Straub (2006) presents excerpts from the feedback Peter Elbow offered on a college student's essay. The assignment was to inform the reader about a subject the writer was knowledgeable about, not to tell a story of personal experience:
Lake Ivenhoe is unique because the only thing between you and the fish are occasional patches of lillypads. The best solution to this problem is to work a top-water buzz bait in the early morning or late afternoon. I have hooked some big bass using this technique, but if the bass is big enough to give a good long fight it can be very difficult to get it through the lillypads. After fishing the lillypads that morning my next move was to work a plastic worm under the giant oak trees that hang out over much of Lake Ivenhoe. Bass like to hang out in these shady areas so they can better spot unsuspecting prey swimming by. This didn't produce the monster bass I was looking for, so my next move was to work a spinner-bait along the southeast bank of the lake. (p. 336)
In Elbow's feedback, he never praises or criticizes the text or the writer. Instead, he tries to figure out and then ally himself with the student's (confused) intentions:
I felt something interesting going on here. Seemed as though you had the assignment in mind (don't just tell a story of your experience but explain a subject)—for awhile—but then gradually forgot about it as you got sucked into telling about your particular day of fishing. (You'll see in my wiggly lines slight bafflement as this story begins to creep in.) (p. 338)
Imagining and describing the student's intention isn't just a way of being nice. It isn't just a way of framing the feedback so that it will be accepted, although it likely does that. By labeling the student's confusion as "something interesting," Elbow normalizes it as something all writers experience as they find, express, clarify, and shape their intentions through language.
After pointing out the problem of the distance between the teacher's assignment and the student's emerging intention, Elbow uses his readerly preferences to point toward a complex solution:
The trouble is, I like your stories/moments. My preference would be not to drop them ("Shame on you—telling stories for an expository essay") but to search around for some way to save them as part of a piece that does what the assignment calls for. Not sure how to do it. Break it up into bits to be scattered here and there? Or leave it a longer story but have material before and after to make it a means of explaining your subject? (p. 338)
The simple solution would have been to cut the story entirely. After all, that's what the assignment, literally interpreted, would suggest. Elbow's solution—integrate your goal (tell about my day of fishing) with the teacher's goal (students will learn to write exposition!)—emphasizes the centrality of a writer's intention in his or her work. And it may also lead to a better product than an essay that merely followed the original instructions. Contemporary expository writers like Bill Bryson, Jonah Lehrer, and Malcolm Gladwell have gained millions of readers through this particular solution, telling a story that sets a problem or question and then presenting expository information that explains or complicates the story.
Although Elbow's suggestion is intended to improve the essay, he frames it in a way that supports the student's view of himself as an active learner. The student can't simply follow Elbow's orders, because Elbow hasn't given any. The purpose of Elbow's uncertainty ("Not sure how to do it") isn't just to force the decision back on the student. A writer himself, Elbow knows that once you note a problem in the text, there are a dozen ways you could work around it. Elbow is telling the student, "As a writer, you can't know the right solution to this problem until you're in the middle of it."
Elbow's choice to frame possible solutions with his readerly preferences ("I like your stories/moments. My preference would be to not drop them") not only shows humility, but also gives the student an honest and useful view of the reader's role in the writer's work. Readers are human beings whose thoughts, feelings, and actions you are hoping your words will affect. Writers must constantly navigate this truth, asking, For whom am I writing? Whose feedback is most helpful to me now?
Keeping the Purposes of Feedback in Mind
Every student experiences feedback in different ways, and there's a real trick to knowing when no feedback is needed. Feedback of any kind can sometimes paralyze a student, interrupting the flow of his or her work. We need to understand and recognize when and why this happens. My sketch of three goals for giving feedback, then, doesn't capture everything. But it's a start.
We can't always accomplish all three purposes of feedback at once. A single comment may not simultaneously help the student make the performance or the product better, contribute to a healthy view of active learning, and tell the learner the complete story of what artists or writers or mathematicians or scientists or historians do. But we should be sure that we always give feedback with these goals in mind.
