Sunday, March 18, 2012

A True Story

Back in the year 2000, there was a lot of fuss about "Reform Calculus." The Advanced Placement committee had announced changes in the AP test, and there were several calculus books being offered that were quite different from the Thomas book that most high schools had been using for years and years. After much deliberation, Evanston decided to make the change, and we adopted the Ostebee Zorn calculus book for BC Calculus. There were two of us teaching the class that year: Ron Selke and myself. We approached the year with excitement and fear.

As we worked through chapter one with the students, we both learned a lot about how to make calculus meaningful and understandable to our students. We had decided to collaboratively write tests, and so we did. The first test covered the ideas in chapter 1 rather well we thought, and we were eager to see how students performed.

To say the first test was a disaster would be an understatement. There was not one student who even tried to work all of the problems. Many students left three or four blank. Ron and I looked at the test, and it measured what we thought was important, but because of the difficulty it measured little or nothing and created considerable discontent among our students. And these are the best students we had. We adjusted the grading scale on the test, admitting that we had totally failed to create a fair test, amd promised that we would do better for the next chapter.

In considering how to fix the problem, we had several ideas. One was to break the chapters into two tests. We rejected that because it would mean giving up too much instructional time for formal assessment; we would spend the year writing and grading tests. Another idea was to only test the easy stuff. We rejected that approach as not being in the best interests of our students. We were committed to making the class a rich mathematical experience that matched the wonderful way Ostebee and Zorn were allowing the course to unfold. Then one morning Ron came to me with one of the best ideas I have encountered. And I resisted at first. I offered reasons why it was a bad idea. After all of that, I agreed to try it. I have never looked back.

Ron's idea was to make a collaborative problem part of the test. The original plan said the day before the test, we would give each student a collaborative problem. This collaborative problem would consist of several parts and would encompass the main ideas of the chapter. The problem would allow us to address some of the subtle concepts or more complicated aspects of the material covered in the chapter. Each student would be required to work with at least two other people, and each person would turn in one copy of the team's perfect, well-organized, well-written solution when the student came in to take the in-class part. It would count as about one seventh of the test grade.

After a couple of tries, we modified the conditions a bit. In particular: we gave students the collaborative problem several days before the test, and always so it would be in their hands over a weekend. We posted the collaborative problem on our websites to allow absent students to access it. We helped shy students find collaborators. We encouraged collaboration with students in the other sections of BC Calculus.

The collaborative problem turned in to one of the best educational experiences in my career. Most of the work was correct, making them easy to grade. The in-class part was now manageable, but we were assessing all of the material. More importantly, students were learning mathematics while taking a test. What an amazing experience! Ron and I listened to their conversations as they worked the problems in the math lab, we heard them talking at the beginning of class, and we flat-out asked them about their experiences with the collaborative problem. All of what we heard was exciting. There was an outcry when we did not offer a collaborative problem for a test we gave on a half chapter. We had found a way to help them consolidate their ideas before they took the in-class part, so the in-class tests were also done better.

We began to notice that the collaborative groups became entities in themselves. The students started to get together just to study. Some of them met during a common free period every day in the math lab or the cafeteria and went over homework questions. Parents praised us for the learning they saw taking place in their homes as students gathered. Collaborative groups compared reaults with other collaborative groups. Students made friends and learned that learning is not an isolated activity.

I also taught a class in Multivariable Calculus and another in Linear Algebra for those students who had finished BC Calculus and had not graduated. There was a clamor for a collaborative problem in that class, so I happily agreed to their demands.

The second year, we realized that we did not have to rewrite the collaborative problem except to fix questiosn that didn't quite go where we thought they would. By now these questions are establsihed. The students understand their value as learning tools, and so there is little evidence that they are looking at old tests.

One moral to the story: when you try something new, it rarely works the way you
intended it. But the thing to do is not to throw it out--then you wind up
doing the same things you were dissatisfied with before that led you to make
the change. Rather, you need to try and identify what's not working and
fix that piece, By iterating several times, you come up with a new
strategy that does accomplish your goals--and even gets you places you hadn't
realized you wanted to be!

By the way, our students performed better than ever on the AP test and have ever since. It is nice when one's observations are valideted by an outside source.