I am an avid reader of your column on The Gloss and The Grindstone for quite some time now, but never before had I been in direct need of your advice. Here it is: you have mentioned several times in your columns that you are a tutor in a lot of subjects. I have just started tutoring students myself. It’s maybe not really the same situation: I am employed by the teacher of a class that I have taken three years ago, to supervise the exercises part of the course. I have 24 students, engineering freshmen, at once, for one hour twice a week.
I am very anxious to do the best possible performance, as I know that a good tutor can make a tremendous amount of difference in a freshman’s future academic record. Any advice?
Hmmn. In my tutoring, I watch the students do a series of mixed problems so I can diagnose what they don’t know. Then — and I don’t know whether this is really possible in your field — I take the thing they don’t know and write problems about it on the spot. I mean, the first 5 years I was tutoring, I didn’t have the ability to do this for every single question type, but now, it’s very easy for me to generate a clever problem with an exponential expression on top of another exponential expression, and everything magically turns out to be an integer even though it didn’t look like it was going to be. Or sometimes, I’ll just take the same problem the student just struggled with, change the numbers, and make the student do it again.
Often in tutoring, a student says, “I had trouble with this problem,” and you show the student how to do the problem, and then you’re both pretty satisfied, even though neither of you should be. The student is never going to see that particular problem again. The student needs to know what to take away from that problem, what elements she is likely to see again, and how they might be remixed in future problems. Generating unique, similar problems helps create “transfer,” the ability to apply skills gained from this task to different, future tasks.