Dr. John Hattie: Assessment should measure teachers' impact
The educational statistician says current approaches to personalization, assessment miss the point
Measurement statistician Dr. John Hattie has spent his entire career in academia, beginning at the elementary and high school levels in his home country of New Zealand before traveling to Toronto for his Ph.D. and settling down for a while in North Carolina. Since 2011, he's served as a professor of education and director of the Melbourne Education Research Institute at the University of Melbourne, Australia — and his research has shaped an educational mindset focused much more on progress and teacher expertise than on achievement as it's currently defined in the U.S.
"In this country, you have more tests available for teachers about kids' achievement, but how many assessments do you have to help teachers understand how kids learn? Hardly any, if any," Hattie told Education Dive. "But that’s their job. I'd resource that. I’d resource schools so they knew better about their impact and looked at progress."
We recently caught up with Hattie at the 2016 Visible Learning Conference in National Harbor, MD, to talk education policy, personalized learning, where assessment should focus and more.
EDUCATION DIVE: There's been a lot of focus on the American education system’s move toward “School 2.0,” away from the industrial model. Obviously, there’s a lot of contention over how to evaluate students. What are some things that you think, as the American education system moves toward that, administrators and policymakers should take a look at and reevaluate?
JOHN HATTIE: It’s fascinating both as an insider and an outsider. My children were brought up in the school system here in the U.S., in North Carolina. So I’ve seen it from the parent side, from that particular view. You’re not that different from many countries in the sense that there’s a desperate move to want to move away, but for the last 200 years, a remarkable group of superb educators have worked out how to make the industrial model work. So it’s very hard to change from something that’s working.
One of the things that fascinates me, for example, is open learning spaces, and the effect size of open learning spaces on achievement is zero. But when you go into one where the teacher’s dramatically changed what they do from the factory model, it’s more than impressive what the effects on the students are. I worry about our current factory model because for about 60-70% of the students, it’s working very well, but for 30-40% it’s not. And it’s not always 30-40% of the struggling kids — some of our brightest kids are failing in the industrial model.
How do we merge what we’ve learned so well from that model to bring what’s needed now? Take for example math and science. The job opportunities are dropping off dramatically for math and science graduates without social skills. It’s a chasm, dropping off. For math and science students with social skills, it’s an increase. What the workforce is looking for is interpreters and communicators, but unfortunately we still teach math and science as if they’re lone subjects to individuals working alone, doing their own assessments by themselves, their own work by themselves. And we’re not teaching that communicative, collaborative skill in those areas.
I see many of the things that have been happening in this country as not being conducive to the profession of teachers. The other thing that’s worth remembering is this move by government to intervene on the outcomes of schooling is a very recent phenomenon. In fact, it wasn’t until Bill Clinton raised the goals of schooling and the second George Bush put into place No Child Left Behind that government has been that intrusive in schools. So I won’t say it’s a recent phenomenon, what we’ve been seeing by the takeover by the policymakers.
As I was mentioning this morning, I despair at listening to the parents and what they want. They want a better factory model than they had, which won’t serve their kids well. Somehow, we’ve got a major job — both through the media and through the work that we do in schools — of informing the parents that learning [today] is a different phenomenon than what they had. It is more about collaboration. It is a noisier sense of learning. It’s a different kind of learning.
Along that line, there’s always kind of this push from the top for educators to innovate, but it seems like a lot of the policies that come down from the top kind of discourage or dissuade educators, as with the stringent accountability standards that came with No Child Left Behind.
HATTIE: Well, I think Larry Cuban has it right. It’s kind of like an ocean. At the top of the ocean, you can get waves and it’s a swirling set of activity going on, but right at the bottom of the ocean, when the classroom door gets closed, schools haven’t changed for the last 150 years. The trouble with the swirl of all the policies is that it gives the impression that you have a broken system. Certainly in America with the Gallup polls, 80% of the parents think the quality of education in America is going backwards, [but] 80% of the parents think the quality of the education their kid gets is pretty damn good. There’s a massive mismatch there.
The problem I have is that, as the debate gets more public, we get into solutions that don’t work. We get into structural solutions: Holding kids backwards, more assessment, different curricula — all the things that have small effects. We don’t prioritize expertise. Every parent thinks, “That’s what the government pays for, anyway.” But actually paying for expertise is very, very expensive. We are unfortunately moving away from the notion that we need to develop that expertise in our teachers and school leaders. Unfortunately, our school leaders often don’t privilege their own expertise. We’ve got a very strange profession where we deny our own expertise, and that’s what bothers me the most.
The policies, if you look at them in this country, they’re very much driven by the lowest-common denominator. They’re very much driven by the structural. The whole debate you’ve got about charter schools is a massive distraction. We give parents the right to choose schools, but the variance between schools is not nearly as great as the variance within schools. We don’t let parents choose teachers, so what a massive con it has been to give them the right to choose schools when it hardly matters compared to what really matters. And those are the kind of public relations exercises we take that kind of disparage the expertise we need to develop.
There’s also a lot of talk about class size and the need to provide more personalized education, because kids learn differently, not every kid is good with homework, etc.
HATTIE: It’s true! Of course, every kid is unique in that sense and all education needs to be personalized, but “personalized education” doesn’t mean “individualized.” A lot of kids learn dramatically well from other students, so the mistake often with personalized education is assuming that it’s just about individual by individual. But when you come with the language of “personalized,” you come with a whole lot of baggage about smaller classes.
It’s really an interesting topic to look at smaller classes. The first thing I want to say is that the research I’ve done has looked at what has happened when we’ve reduced class sizes, not what should happen. It should be more personal, there should be more feedback, there should be more interaction — but what has happened is the opposite. In fact, we know from looking at the comparisons between classes of 25-30 to 15-20, in smaller classes, teachers talk more. There is less feedback. There is less group interaction. Why is that so? That’s what the intriguing thing is. I absolutely agree with everyone who says it should make a difference, but it hasn’t.
I think we know why it hasn’t made a difference: If you take a teacher from a class of 25-30 and put them in a class of 15-20 and they teach the same way, who’s surprised that it doesn’t make a difference? And your country, like mine, has spent trillions of dollars reducing class size with no effect. For the last 200 years, our teachers have found out pretty impressive ways for kids to learn in those larger classes. Would I spend a penny on reducing class sizes? No ... If I’m given trillions to reduce class sizes and that’s the only choice, I’ll take it. But if I’m given trillions, I’d rather spend the money on developing teacher expertise — having them spend less hours face-to-face and more hours collaboratively working and looking at their impact.
Unfortunately, politicians often don’t give us that choice: It’s smaller classes or nothing ... If you look at the trillions we’ve spent on reducing class size in this country over the last 30 years with virtually no impact, you’ve gotta wonder why we wasted our money.
On that same note about kids learning differently, there’s also the question of assessments. The argument for some is along the lines of “You can’t judge a fish by its ability to climb a tree.”
HATTIE: I think that’s a very unfair criticism. I’m a measurement person. I don’t dislike assessment at all. My concern is the purposes of it. What I’ve seen in your country is that you’re using assessment, constantly saying better assessment of kids is going to help make a difference to them. There’s no great evidence for that. In fact, I would challenge any teacher, any parent, before a kid does a test, to ask the kid what mark they’re gonna get. They’re pretty accurate, so why do we bother?
If you developed assessments to better help teachers do their job, firstly, you’d get different kinds of tests. But secondly, you’d get much more powerful impacts. We developed the New Zealand assessment for elementary and high school, which is entirely based on helping teachers understand their impact. I’m very proud of the fact that it’s a voluntary system — schools don’t have to use it, but 80% of teachers in New Zealand are using it 16 years later. They don’t want to get rid of it. They like measuring their impact. It’s a different kind of assessment. And yes, kids do get information from it. But the main purpose of it is to help teachers.
So that’s the question I’d ask: Any time a teacher does a test, what did they learn about their impact? Who did they have their impact with? This is your personalized notion, how they have to change for groups of kids and individual kids. If that’s the purpose of assessment, it’s very powerful.
Your current system, where you privilege high achievement, favors schools where kids start very high, and it gives a very false impression. So yes, assessment can make a huge difference if it’s about helping teachers know their impact.
If you were given control of the education system, what three things would you change immediately?
HATTIE: I’d change the narrative about what a good school is from a school that has high test scores to a school that can demonstrate the greatest progress. I’d certainly then ask the community of educators to address that problem by demonstrating to their communities the nature of their progress. There is no one right answer.
Would I use test scores on that? Absolutely. Would I only use test scores? Never. There’s a mixture in there. What is the preponderance of evidence that a school is adding value to every one of your students in the school?
The second thing I’d do is I would change the system to value expertise. What we do in Australia, we’ve got four levels. All teachers are classified as graduate, proficient, highly accomplished or lead. It’s done independently of the school. If you do it in the school, if you do it within the system — in my state of Victoria, which still does it that way, 99% of teachers get an annual bonus, an annual increment. That is just not credible. When you do it as New South Wales and the other states are doing it, where there’s national moderation, what we’re doing in those states is we’re creating positions for which only a highly accomplished lead teacher can apply.
That’s where the salary increment comes. We’re trying to make it so you can be a lead teacher and get paid more than the principal, but you take on extra responsibilities of collaboration of the new teachers in school and getting teachers to climb that ladder of expertise. Every school deserves at least one lead or highly accomplished teacher in that school. And I want to do it in a way so that parents don’t say, “If my kid doesn’t have a lead teacher, I’m not gonna go to that school.” That would never work. The lead teacher isn’t necessarily the teacher in the school who gets all the kids. It’s the teacher in the school who works with other teachers to help them understand and maximize their impact.
Do I need a third thing? I don’t need a third thing. I’ll stick with two. I’m happy with those two.
Follow Roger Riddell on Twitter