Learning depends on what you already know

by Stephanie Chasteen on April 15, 2009

Firstly, I just have to say how beautiful the view is outside the airplane window right now.  It’s been an extraordinarily bumpy ride on my trip from San Francisco back to Denver, and I’m a real nail-biter when it comes to turbulence.  But now we are above the clouds and the sky is white with a band of blue and purple at the horizon.  I’m thinking of light scattering and optics and it’s quite soothing.

But on to the meat of today’s post.    There are a lot of ways you can slice the essential “messages” of education research, and what we know about how people learn.  My colleague, Noah Finkelstein, slices it this way:

1 – Actively engaging students in learning is important
2 – What people already know affects what they learn
3 – Context shapes what students learn
4 – Teaching is an effective way to learn

I’m reading a paper right now that focuses on number 2 – learning depends on what people already know.  There is a whole body of thought (that’s quite well substantiated by research) that learning isn’t just the addition of new bits of information, but rather the process of changing people’s existing conceptions.  In that way, it’s important to know where you’re starting (what people’s naïve conceptions are) so that you can CHANGE them, rather than just trying to overwrite them.  People have their existing ideas for a reason – they make sense, and they have some predictive power.  So the new idea has to provide some advantage over the old idea in order for people to drop their old model.  The same content might make perfect sense to one student, who doesn’t have a previous conception that conflicts with it, and yet be actively resisted by another student. The job of a teacher can be quite difficult, then!  As if you, dear reader, didn’t know that already.

This is old work, from Posner in 1982.  But realize, there is some debate as to how well-formed these ideas are.  As this paper by Hewson argues,

because students have experienced and thought about the world, they do come to class with ideas, often hazy, ill-formed and inappropriate, but ideas nonetheless.

There’s some evidence that people don’t come to the table with well-formed “theories” of how the world works, but rather these sort of muddy “proto-ideas” which often shift depending on the context of the problem in question.

But anyway.  An example of a misconception is that “mass / weight = “heaviness”” or that some objects have mass and others (like air or hair) do not).  Scientifically, of course, all objects have mass, and mass is a measure of the amount of matter in an object.  So, how can we work with those naïve ideas to help students arrive at the scientific understanding of mass (or density or anything else)?

The paper that I’m reading right now (Hewson and Hewson, J. of Research in Science Teaching, vol 40, p. S86, 2003) proposes four main ways that teachers can teach in a way that uses students prior knowledge.

1.    INTEGRATE new conceptions with existing (presumably correct) conceptions
2.    DIFFERENTIATE existing conceptions into more fine-grained categories or bins.  This assumes that the student is applying a concept too broadly, and they can see that what was plausible in one situation is not plausible in a different, more complex situation.  (For example, “heaviness” is an undifferentiated concept which includes mass and weight).
3.    EXCHANGE one conception (presumably incorrect) for a new one.  The new concept has to be shown to be more useful and predictive than the old one for the student to become dissatisfied and toss their old model.
4.    CONCEPTUALLY BRIDGE the new idea with meaningful common experiences, to give the student a context for seeing why the new idea is plausible.

In this study, they wanted to see how these different techniques affected student understanding of density.  In the control condition, they presented the main ideas in density and floating (eg., density = mass/volume, all matter has mass), discussed them, and did experiments to demonstrate these ideas.  Pretty typical science class stuff.

In the other class, they took what they knew about students prior knowledge (eg., mass and volume of an object affects whether it sinks, density = “crowdedness”), and geared their discussion based off those ideas, arriving at the more scientifically valid ideas (density = mass/volume, density is the packing and mass of particles, all matter has mass and density).

They found that starting from the alternative conceptions was more effective in teaching students the ideas they wanted them to learn (as measured by a conceptual quiz), though it wasn’t as good as one would have hoped.  But still, better than the standard “tell them what you want them to know” approach.  But both groups still held on to their alternative/naïve conceptions that they started with, at least to some degree.

Hewson & Hewson, J. Research in Science Teaching (40)  PP. S86–S98(2003)

Hewson & Hewson, J. Research in Science Teaching (40) PP. S86–S98(2003)

So, the take-home message is, starting with students’ existing ideas may be more effective than the standard textbook approach of presenting the ideas that you want them to learn.

{ 1 comment… read it below or add one }

Derek April 17, 2009 at 12:41 pm

I’ve found it useful sometimes to think of “coverage” (the traditional model in which new ideas are “covered” in class) vs. “uncoverage” (in which student misconceptions are “uncovered” and addressed in class).

If I have a piece of software I’m writing that doesn’t quite do what I want it to, I don’t typically rewrite the entire software. Instead, I try to determine what’s not working and target my efforts to fix that.

In the same way, if a student hasn’t quite got a couple of concepts, it’s more efficient to try to figure out what those misconceptions are and address them. It’s less efficient to explain the entire section of the textbook to that student.

“Uncoverage” happens regularly in office hour settings, I think. When you have one or two students in front of you, it’s relatively easy to figure out what they don’t understand and address it in one of the ways you mention above. When you’ve got a class of 20 or 50 or 100 students, on the other hand, “uncoverage” becomes much more difficult. Scaling up to those settings requires additional tools, I think.

Leave a Comment

{ 1 trackback }

Previous post:

Next post: