Standards, grades and tests are wildly outdated, argues 'End of Average'

Giant finger pushing test-takers into bubbles
LA Johnson

Todd Rose dropped out of high school with D- grades. At 21, he was trying to support a wife and two sons on welfare and minimum wage jobs.

Today he teaches educational neuroscience at the Harvard Graduate School of Education. He's also the co-founder of The Center for Individual Opportunity, a new organization devoted to "the science of the individual and its implications for education, the workforce, and society."

In other words, Todd Rose is not your average guy. But neither are you.

In fact, he argues, absolutely no one is precisely average. And that's a big problem, he tells NPR Ed: "We've come to embrace a way of thinking about ourselves as people that was intentionally designed to ignore all individuality and force everything in reference to an average person."

Admissions offices, HR departments, banks and doctors make life-changing decisions based on averages. Rose says that "works really well to understand the system or the group, but it fails miserably when you need to understand the individual, which is what we need to do."

Rose talked with us about his new book: The End Of Average: How We Succeed in a World That Values Sameness. The opening example you use in the book is that in the 1940s, when the Air Force designed cockpits based on the average measurements of the pilots, there were an unacceptable number of crashes. But when they went back and measured thousands of pilots, across 10 body dimensions, they found that zero of them even came close to the "average" on all 10. So they concluded that they had to redesign the seats and so forth to be adjustable to each person.

Body size is a very concrete example of what I call jaggedness. There is no average pilot. No medium-sized people. When you think of someone's size you think of large, medium, small. Our mass-produced approach to clothing reinforces that. But if that were true you wouldn't need dressing rooms.

So dimensions like height and weight and arm length and waist circumference ...

Yes, they're not nearly as correlated as you would think. Height is one-dimensional, but size isn't. People are jagged in size, in intelligence, everything we measure shows the same thing.

I'm going to quote a line from the book, said to psychologist Peter Molenaar, who is arguing for a greater focus on individual difference: "What you are proposing is anarchy!" How do you make decisions about people if you can't use statistics and cutoff scores and compare them to averages?

People feel like if you focus on individuality, everyone's a snowflake, and you can't build a science on snowflakes. But the opposite has been true.

It's not that you can't use statistics, it's just that you don't use group statistics. If I want to know something about my daily spending habits, one straightforward way would be to collect records of what I spend every day. To take an average for myself would be perfectly fine. So you can generalize across time, but not across people?

We've got to let go of putting a group into a study and taking an average and thinking that's going to be close enough to universal insight.

Now we have something better. We have a natural science of individuality that gives us a surer foundation. We've gotten breakthrough insights in a whole range of research, from cancer to child development.

How does what you term "Averagerianism" impact our school system both historically and today?

It's so ubiquitous that it's hard to see.

We design textbooks to be age-appropriate, but that means, what does the average kid of this age know and can do? Textbooks that are designed for the average will be a pretty bad fit for most kids.

Then you think of things like the lockstep, grade-based organization of kids, and you end up sitting in a class for a fixed amount of time and get a one-dimensional rating in the form of a grade, and a one-dimensional standardized assessment. It's everything about the way we test and move kids forward.

With standardized tests, I often hear teachers talking about students being two months behind or ahead, as if there's a very fixed timeline for progress that all human beings should fit.

It feels comforting. But if you take the basic idea of jaggedness, if all kids are multidimensional in their talent, their aptitude, you can't reduce them to a single score. It gives us a false sense of precision and gives up on pretending to know anything about these kids.

So alongside jaggedness, two other principles of individual variation you look at are "context-dependence" and "pathways." Talk about those.

It's meaningless to talk about behavior and performance without context. Let's take assessment. Carnegie Mellon [University] had this work showing that changing the way a question is asked can fundamentally alter how a kid performs. So if the [math] problem is about football players instead of ballerinas, you can't standardize on the item. That systematically affects the kids' ability to demonstrate what they know.

But at a macro level, I think [context] introduces an attention to things like the impact of stress and trauma.

And what about pathways? This sounds a lot like the talk around personalizing learning using technology and allowing each student to learn at his or her own pace.

I think people who care about personalized learning talk about it as: If we just collect more data, we're going to have this personalization. And that's not clear to me at all.

I think when you look at the idea of pace, we are so convinced that slow means dumb and fast means smart. We feel justified in pegging the time to how fast the average person takes to finish.

But this is where, with a better understanding of this and realizing, "Oh, pace really has nothing to do with ability, people are fast at some things and slow with others," you would build a very different system than the one we have.

Do you think the school system acknowledges the need to treat students as individuals? Two years ago I would have said no. But my colleague Paul Reville, who used to be secretary of education in Massachusetts, he's rethinking the architecture of school systems. In most states, people have put on the books goals about meeting every kid where they're at. Even the "Every Student Succeeds" [ESSA, the new federal law] approach is based on the assumption that we're meeting each kid where they're at, to give them what they need to be successful.

But we haven't thought through the system design that needs to be in place to do that.

We're trying to have a system to do what it was never designed to do.

What about in higher education?

In higher ed we have a brutally standardized system. It doesn't matter what your interests are, what job you want, everyone takes the same courses in roughly the same time and at the end of the course you get ranked.

This is personal for me. I have two kids in college. The idea that someone is going to click a stopwatch, compare you to other kids in your class, and the kids with the best grades can get the best jobs, that's not a good deal. I want my two boys to figure out what they love and what they're good at and be exposed to things and be able to turn that into a job.

You talk about innovations that are starting to catch on, like competency-based education and credentialing — basically, accommodating different pathways and different balances of strengths and weakness.

There's plenty of ways we're making smaller units of learning to combine in ways that are useful to you. To me, competency based education is nonnegotiable. I don't think you can have fixed-time, grade-based learning anymore. I don't see how you justify diplomas.

It doesn't mean students can take forever, but allowing some flexibility in pace and only caring whether they master the material or not is a sound foundation for a higher ed system.

There are so many examples of a lot of really interesting universities trying these things.

Yes, reading this book it struck me that in some quarters, it seems like we've already moved forward to a focus on individuality, innovation, creativity. You talk about how companies like Google are finding that GPA or school prestige or even ranking employees against each other is not useful, and instead they need to create, essentially, performance-based assessments for doing tasks in context.

There are bright spots where you can see the principles of individuality at work.

So for me it comes back to, well, wait a minute. So why is that not the mainstream?

What I think my contribution is, is to say: Our institutions are based on assumptions about human beings. Our education system is based on a 19th century idea of an average person and using 20th century statistics.

As long as people think you can understand people based on averages, or how they deviate from averages, it seems reasonable. It looks like accountability and fairness rather than absurdity.

And you're trying to show that there's an alternative? If we don't get rid of this way of thinking about ourselves and the people around us, it's hard to get the public demand to create sustainable change. That's the role that I and my organization want to play. We're making a really big bet. Copyright 2019 NPR. To see more, visit https://www.npr.org.

(2016-03-23 04:00:00 UTC):

A previous version of this post misidentified psychologist Peter Molenaar as Paul Molenaar.