Of That

Brandt Redd on Education, Technology, Energy, and Trust

12 September 2011

Breaking the Tyranny of the Bell Curve

If you take a random set of students, teach them all the same way and then give them all the same standardized assessment the results will follow a normal distribution or "bell curve" with a few excelling, the majority performing near average and a few failing. This is the tyranny of the bell curve.

There are all kinds of problems with this: Standardized tests result in normal distributions of scores because they are designed to do so. Not necessarily because human ability really follows a normal distribution. Indeed, human intelligence is malleable.

But let's set that aside for a moment and just go crazy theoretical. Suppose you had a large population of identical students. Then you put them in classrooms where instruction was delivered in identical ways. Then you gave them an identical assessment. The results would approximate a normal (or bell) curve. Why? Because a normal curve is what results when you average out a bunch of random errors. Instruction is naturally error prone. Students don't always pay attention. Even when they do, they don't always understand. Teachers make mistakes. People get sick or have bad days.

My colleague, Josh Jarrett, is fond of saying that high school graduates' knowledge is kind of like Swiss cheese with random holes in their understanding.

When looking at children, my natural inclination is to celebrate their differences. When they are dressed the same, in sports uniforms for example, I gravitate to the differences the color of their hair and their eyes, how they smile, who they cluster around, what grabs their interest.

Despite this diversity, our society needs all children to reach a certain standard of competency in core subjects of literacy and mathematics. Likewise, they need to have a basic understanding of the social and civic institutions and norms that are essential to prosperous society.
So, the challenge is achieving consistent results (academic achievement) while prizing the inconsistency of the inputs (our children). The obvious answer is that we adapt the education to the needs of each student. As a friend put it, "Every student should have an IEP."

But IEPs or Personalized Learning, as we prefer to call it, is prohibitively expensive, right? I believe that the principles of mass customization so successfully applied in other industries can also be applied to education. I'll be writing more on this in coming weeks.

Posts in this series:
Breaking the Tyranny of the Bell Curve
Tackling Bloom's Two Sigma Problem
The Personalized Learning Model

1 comment:

  1. I was recently talking with my maths teacher friend here in the UK about the differences in education vs the US. I was telling her about some of the problems with "no child left behind" and she told me about their initiative with a similar goal and name, "Every Child Matters" (http://en.wikipedia.org/wiki/Every_Child_Matters) - but the approach is from the individual rather than the test results. Her challenge and duty as a teacher is to find the individual needs of every child and teach to those needs. It seemed like a much better approach to me, although a great challenge - and maybe similar to what you are saying about personalized learning.