# Learning by Screwing Up

One of the signs on my wall states, “You can’t learn effectively without the freedom to make mistakes.”  It’s a new school year, and my new batch of students is working on learning this lesson.

The vehicle was a lab to measure the thickness of aluminum foil.  This was an inquiry lab, but because students arrive in my class with little or no experience with inquiry, I start off the year doing whole-class brainstorming of ideas for the procedure.

Not surprisingly, with a little prompting, they figured out that they could take a rectangular piece of foil, measure length and width, measure volume by displacement of water, and divide to get thickness.  So I got out foil, rulers, scissors, and 100 mL graduated cylinders and turned them loose.

Most students started with a piece of foil that was small enough to fit easily into the graduated cylinder, and found that they couldn’t measure the change in volume.  Some started out with a large piece and found that when they crumpled up the foil, the volume increased much more than expected (most likely because of trapped air).  It took most groups a few tries to get data they were confident of.

After the experiment, I asked questions like,

“Did it work on the first try?”
“When the volume didn’t change, did you figure out what was probably wrong?”
“Do you remember what you did, why you did it, and how well it worked?  If so, did it help that you made a few mistakes and had to fix them?”

Being free to make mistakes and learn by correcting them has a tremendous effect on learning.  I’ve blogged about this in previous posts, including A Video Game Approach to Learning from 2011 and  If At First You Don’t Succeed… from 2008.  What’s different this time around is that I deliberately gave my students a task that encouraged them to make mistakes and figure out how to correct them, and afterwards I made the students aware of the extent to which making mistakes helped them learn.

Another goal of the lab was to introduce my students to quantitative error analysis.  I (deliberately) gave them 100 mL graduated cylinders (marked in 1 mL increments), which meant that their volume measurement was significantly more approximate than their length and width measurements.  Students were careless about estimating their volumes to ±0.1 mL (which can be done).  Because they were only certain of the volume to ±0.5 mL, their uncertainties came out to around ±40%.  This provided an opportunity to ask some useful follow-up questions, such as “Which of your measurements would you focus on improving?” and “How could you improve that measurement?”  The learning that followed because they encountered these problems would have been significantly hampered if I had given them a pre-written procedure that told them how to avoid them instead.

Share it now!
0
0