New ways to assess your students’ learning

by Stephanie Chasteen on February 12, 2009

Session:  Eugenia Etkina - Pedagogical Content Knowledge

Session: Eugenia Etkina - Pedagogical Content Knowledge

Why is it that we make tests multiple choice, when the world isn’t multiple choice?  We’re not an industrial society anymore, where following directions is the key to success.  Instead, we’re problem-solvers, and knowledge is the important commodity.  But our education system hasn’t quite caught up with this shift.  We’re still testing our students like they’re good little worker bees, for whom “knowing the answer” is the important skill.

But if we want to train students to be good problem solvers, we have to assess them on it.  One of Joe Redish’ teaching commandments is that if you want students to learn something, that has to be part of your assessments.

So, how to we assess problem solving skills?  Eugenia Etkina has two examples which were pretty nice.

1. Jeopardy
This is a fun game, once your students get used to it. Take this for example:
900N – (50 kg)(9.8 m/s^2) = 50 kg * v^2 / 12m.
Sketch a situation this might describe
Write in words a problem for which equation is a solution.
(Can’t figure out the answer?  Write in the comments).
This requires students to put together pieces and make conections between the formulaic descriptions of physics and physical situations.

2.  Deep problem-solving assessment

Falling ball and pendulum

Falling ball and pendulum

Watch this video. Decide what data you need to estimate the height of the table two different ways.  How do you decide whether the two methods agree?  Hint:  You can step the video forward frame by frame using your arrow keys.  All the data you need is on that page.
(Need help getting started?  Write in the comments).
Unlike a traditional exam, this assessment

  • Doesn’t tell you what to do (the student has to come up with the method)
  • Doesn’t tell you to neglect air friction, etc. (this is a genuine situation)
  • Tests whether they really know what uncertainty in measurement is, and how it’s different from experimental error
  • There is no correct answer
  • Students can learn during the assessment

If these are the skills we want them to learn, then we have to test them on it! Physics isn’t just about knowing facts, concepts, or how to solve formulaic problems.


Anonymous Coward February 20, 2009 at 12:50 am

Re: #1
Figure out the answer? I’m still waiting in suspense for the parenthesis to close!

Re: #2
Can I assume that this video was shot on the surface of earth? Can I assume that the person shoving the ball is an adult human? I mean, geez, that table height could be anything.

sciencegeekgirl February 23, 2009 at 12:40 am

Re #1… Well,

900N – (50 kg)(9.8 m/s^2) = 50 kg * v^2 / 12m.
can be rewritten as
F – mg = mv^2/distance

I wrote “distance” because we don’t know if that’s a height, radius, etc.

So, starting with the left hand side, what does that look like, physically?

And the right hand side, what could that be an expression for?

Then put it together…

Re: #2
There are two equations of motion you can use to solve for the height of the table. One involves the equations for the motion of the pendulum. You can figure out how many seconds the pendulum takes to swing by looking on the side of the video in the link — it tells you the frames per second. You can step through the video frame by frame and count.
You can similarly figure out how many seconds it takes the ball to drop, and you know an equation that will tell you how time relates to distance.

The neat thing is that when the whole class does it, you’ll get wildly varying estimates for one of the measurements, and very similar ones for the other measurement. It’s a great way to talk about error and why one measurement is more error-prone than the other.

Comments on this entry are closed.

Previous post:

Next post: