2 ms·
I took a similar approach when I taught programming to non-CS majors last summer. The difficulty is that it required significant one-on-one time with each stude
by scott_s 16y ago
I took a similar approach when I taught programming to non-CS majors last summer. The difficulty is that it required significant one-on-one time with each student that needed help, and it required that they came to me. Luckily, my students weren't shy and I made myself readably available to them almost every day. But college students are generally better motivated to seek out help than high school, middle school and elementary students.
What took so long with each student was, based on their original question, systematically figuring out what their real misunderstanding was. This could be very time consuming, and took enormous patience on my part. It required much back and forth with the student so I could build a mental model of their mental model. Then I had to figure out how to build a bridge from their mental model to the correct one. That also was time consuming, because I sometimes had to build several bridges before finding one that clicked with the student.
Usually helping the Nth student on a project was quicker than helping the first because by that time I recognized what were common problems and built a bag of tricks to explain them. For example, it took me a surprisingly long time to realize the students had no practice running through an algorithm on paper, then translating that to code. (This was partially because the algorithm was so simple I hadn't even realized it was an algorithm.) So one technique I used was to, on paper, set up what was needed for the algorithm to work (list of numbers, table of results, etc.), then make them tell me what to do to get the correct result. I was acting as, basically, an intelligent computer that could be programmed in English. After doing this a few times, the students could finally "see" the algorithm, but it took a lot of time and effort with each student.