Some clashing points of view were aired at a panel discussion entitled "Assessment and Grading: What's the Relationship?" The panelists were Alfie Kohn, a former teacher and the author of five books on education and human behavior, including Punished by Rewards and ASCD's Beyond Discipline; Grant Wiggins, president of the Center on Learning, Assessment, and School Structure (CLASS) in Princeton, N.J.; Tom Guskey, professor of education at the University of Kentucky and editor of ASCD's 1996 Yearbook, Communicating Student Learning; and Bena Kallick, coauthor of Assessment in the Learning Organization and cofounder of EXEMPLARS, a company that supports teachers using performance assessment. Ron Brandt, assistant executive director at ASCD, served as moderator.
A question from a member of the audience provoked the spirited exchange that follows:
Audience Member: As a parent as well as an educator, I certainly want some normative information about how my daughters are doing in terms of developmental trends and moving along through their own developmental process. And I also would like to know how they're doing in terms of some commonly held goals that we have for our kids. Can you give normative and criterion-referenced information together, or are those really antithetical to one another?
Grant Wiggins: [In my chapter in ASCD's 1996 Yearbook, Communicating Student Learning,] I argued that you can and should do both. Because I as a parent feel the same way. I began my [chapter] by saying, here's a school that went to a complete narrative, idiosyncratic report on the kid, and at the parent conference the parent says, "Yes, but how's she doing?"
You talk to most parents and they want normative information. Why? For good reason. They need a context. The only person with a context is the teacher. The teacher knows what developmental growth should look like. The teacher knows whether somebody is slow or not. The parent only gets the individual piece of information about their child. And it's perfectly reasonable to say, "Is this acceptable development? I'm not an expert."
And that means there have to be developmental norms. Now, admittedly, that's sometimes used as an excuse for giving stupid letter grades, which are not tied to anything stable, rational, or empirically sound. But I'm puzzled when people say, "Here's a longitudinal rubric of literacy development. We shouldn't report out percentages of where people are." I don't know why you'd want to hide that information.
Alfie Kohn: I know why. Because of the harms of reporting that. When you report normative information on individual students, you are incapable—I would put it that strongly—of having a truly developmentally responsive and progressive approach to education. Because the point then is, "Am I doing better than others?"
And when you start saying, "Here's what every nth grader should know"—which is implicit in saying, "Here's what 80 percent of nth graders do know"—what you say is, "There must be a particular timetable that doesn't respect the differences in the kinds of talents kids have and the different speeds at which they acquire stuff."
Grant Wiggins: That doesn't follow. The examples in my [chapter] show how what you're saying doesn't follow. Here's a developmental continuum that has bands of predictable development, within a wide range, related to empirical data of the rate at which people learn in a band.
Alfie Kohn: What does that tell you though?
Grant Wiggins: What it tells me is if my kid is here and the band of people in the school is here and the mean is here, then it's predictable that my child may not meet exit standards. And if I see that data played out over time on the same continuum and the rate of growth does not keep pace with the band of the cohort, then I have some reason to be concerned as a parent and say, "What alternative interventions are you proposing?"
Alfie Kohn: I don't agree, Grant. The exit standards can be purely criterion-based, and we can give [parents] warning of problems up ahead. But saying my kid's doing better or worse than this number of other kids only provides a pernicious orientation that leads people to think competitively, not about learning. The parents, the kids, the teachers—
Grant Wiggins: I'm talking about the parent who needs to know where the child stands. If you give an idiosyncratic report, the isolated parent has no knowledge of whether the kid is on track to meet a standard. That's all I'm saying.
Tom Guskey: I think there's a compromise here. Parents do, as Grant suggested, need an anchor. If we provide, even at the elementary level, this sort of checklist of skills that kids have—and we might mark them at a variety of levels: "developing," "mastered," "extending," whatever—the parent looks at that list and at the end does say, "Great, how's my kid doing compared to everybody else in the class?" And the reason they ask that is because that is the system under which they grew up. That really is not helpful.
A better question, which does have a normative aspect to it, is to say, "How is my kid doing compared to expectations for this level?" Which could be developmentally based, which is what Grant outlines in his chapter. That allows parents to still have that anchor, but it also allows a class where basically every kid in the class could be doing really well. We don't have that when it's strictly based on the class, because then everyone's pitted against everyone else and it makes learning too competitive. So if we help parents rephrase that question and say, "Here's how your kid's doing compared with our expectations"—be they developmentally based or whatever—that's a much more informative anchor.
A question from another audience member about the relationship between performance assessment and peer assessment elicited this response:
Bena Kallick: We need to ask the question, "What can we do that allows students to act in powerful ways on their own behalf?" You can't really confer power, because that immediately implies a hierarchical relationship in which I have it and therefore I can give it to you. So the real question of empowerment is, "How can I help you to feel as if you have it and you can make use of it?"
You have to create what I would call an empowering environment. And in an empowering environment, you need to be able to be self-evaluative, and you also need to be able to evaluate the work of others in that classroom. But to have an empowering environment in which that can take place requires structure, a sense of community, and the ability to know what you're [evaluating] someone's work against.
For example, I see a lot of students who have peer conferences in writing, but in the absence of a structure as to what they're looking for in the writing, it's very difficult to have a good evaluative conversation. So it seems to me that it's very important to look at the climate of the classroom, the sense of community in the classroom, and the concept of what I refer to as "critical friends," which is the capacity for people to be critical and also friends, meaning they're advocates for the success of all the students in the classroom.
That's a shift in the way we think about the question of assessment as it relates to self-evaluation and peer evaluation.
Another audience member made a heartfelt plea for specific advice:
Audience Member: I'm a 3rd grade teacher in Plano, Texas. This is my fourth year teaching. Tomorrow I have to go back in my room and test the children on money and time. I've made this great test, and the parents are going to expect 100s and stickers. And I'm frustrated, because I agree with what I'm hearing, but I don't know how to give an assessment when parents need a number and a quantitative value to put on their child. And it hurts me that I set my students up to feel like other people will always judge them.
I want to know how to do it. I agree with everything you're saying, but I haven't heard yet a method to bring to them tomorrow. I need to know that today. [Applause]
Tom Guskey: Help your students understand that although this is an assessment, it is a part of the learning process, and it is not the end of learning. That this is designed to provide some information, including what things they know and can do real well, what things they need to do better, and more specifically how they can improve those things.
The assessment itself should be an information device. We're not going to use it to sort and categorize students, but we're using it as part of the learning process to help them get better. That means they have to have a second chance. What happens if a kid doesn't do well on this assessment? To me, that says just as much about our teaching as it does about the skills and talents of that individual. And so for that individual, I have to find other ways of approaching [his] learning to help [him] learn those things well.
Alfie Kohn: I have some concrete ideas. One: Never put a letter or number on anything a student does, and explain why you're [not] doing that. Narrative comments when you have time—otherwise, nothing. If you must give a grade at the end, explain to the students and engage them in conversation about how that works, and say: "If you have to know what number or letter this would get, because I have to give you one eventually, come up and talk to me."
Two: If you have to give a letter or number, [use] as few gradations as possible. If you're doing zero to a hundred, shift to A, B, C, D, F. If you're doing A, B, C, D, F, shift to check-plus, check, check-minus. If you're doing check-plus, check, check-minus, shift to "okay" versus "needs work." The more gradations, the more we are pushing this kind of extrinsic stuff in their faces. What we need to do is help them engage with the task itself.
Three: Ask the students how they can best show you, when we finish this unit, what they've understood, what they need help with, what questions they had answered, and what new questions [they have]. Offer them alternatives to tests so they don't feel obliged, based on their own experience, just to say more of the same. In fact, that should happen repeatedly, that we bring them into the conversation about when it is time for assessment, how that's going to happen.
Next, if you are doing grades or tests, make as little deal about it as possible. The last thing you want to do is talk about good grades as a cause for celebration or stickers. Rather, we'll talk about how it could have been better, how my teaching could have improved, how your learning can improve next week, and then we move on. Otherwise, the grade or the grade substitute becomes the point.