Friday, September 21, 2012

Flipped Classroom

A "flipped classroom" is one where students watch lectures at home and do their homework in class. The Kahn Academy has done wonders in helping instructors with flipped classes.

I decided to try it with my class that introduces computer programming. My hypothesis: I could get more students to post excellent grades with this method than the traditional method.

Over the summer I worked on this website.

I put the lectures on-line. A total of 55 videos, over 6 hours of run time. Most videos are short. I used a graphics tablet to mark up the screen during the video. I got a good microphone and edited the videos to remove annoying sounds, mistakes, or other distractions.

The full text of the book is on-line. There are code examples and JavaScript animations that animate the code to visualize what is happening.

I created a multiple-choice quiz engine to test the students on-line. That helps make sure they actually went through the material.

The website got a lot of positive feedback from the community-at-large, I was helping more students learn programming on-line than I was in the classroom.

It seemed like everything ready for a full test of a "flipped" class this semester. The first real results would be from Test 1.

I started getting worried when there were students doing worksheets in class that had clearly not watched the lectures or gone over the material first. Plus my "draw a picture" lab had fewer images that really showed off the creativity of what students were able to do.

The results of Test 1 confirmed my worries. It showed a noticeable down-tick in grades compared to prior years. Was it terrible? No, there were still a lot of "A" grades. However there were more students on the C-F range than what I normally get. It seemed like a "flipped" class was allowing more students to slip through the cracks than before.

My conclusion, the data did not support my hypothesis.

Here are more hypotheses I have:

  • Students who don't want to do work outside of class still don't. They gain more from having the traditional lecture in-class and not doing the homework. In a flipped class they skip watching the lecture and just copy worksheets resulting in even less comprehension.
  • Fewer students will become inspired by the material and want to explore a career in that area. Students that aren't willing to ask questions about worksheets have little interaction with the instructor. Those students are less likely to be inspired by the instructor because they don't get to regularly see how excited he/she is about the material.
  • A flipped class will work well in an evening class. Evening classes with working adults have a high drop-out rate for an introductory computer science course. I think that teaching in a 'flipped' manner will result in higher scores for these students even if they result in lower scores for day students. Meeting four times a day, is better than a flipped class, which itself is better than meeting one time a week. 
  • The 'flipped' class scales well. It will still perform 'well' with an average instructor and will involve little work on the instructor's part if the material is already created. e.g., a teacher that doesn't understand calculus well can still do a good job teaching calculus with help from the Kahn academy. Or a teacher can have more students in a class and still teach them. But neither can match a great teacher in a small class.
At some point I'll do a survey and see if I can't get more information from the class. I'm going back to my traditional methods for now.

I am still reaching a lot of students who have never set foot at Simpson. So I don't think the effort spend on the website is a waste. But I'm not improving the scores of my class doing this.