A first experience with video-based flipped-classroom teaching

A first experience with video-based flipped-classroom teaching

Background

In the spring semester of 2016 I taught 5 lectures on Mass Transfer to Master-level students. This was the first time I'd taught this course, and for reasons that I may expand on in the future, I wanted to teach this using a flipped-classroom approach, having students view material online in advance. The format of the videos was based upon the successful Khan-academy-style format (writing live while narrating the video, much like writing on a blackboard). The intention was that students could gain a deeper knowledge of the material than is possible in the class by being able to pause, rewind and review the lecture. Class time would then be devoted to working on problems related to concepts covered in the videos, with the idea that having myself and their peers immediately on-hand to correct misunderstanding would enhance learning of the material.

Equipment

I used a medium-size Intuos Pro tablet (30,800 yen) and a Blue Yeti USB microphone (19,800 yen) to record sound. Sketchbook Pro (3,600 yen / year) was used for drawing while screen recording and editing was done using Camtasia for Mac (9,583 yen). Lectures were assigned using the scalable-learning platform (free, for now).

To enable in-class feedback, I used voting machines ("clickers") to obtain anonymous feedback during the lecture, asking questions about concepts covered in the videos and expanding upon these when an insufficient number of students voted for the correct answer.

Recording videos

It takes a long time to create videos. Recording the first videos was difficult, but eventually I got into a rhythm for producing them. I found the following method worked well. The timings in brackets below are for a 10-minute video that I recorded:

  1. Test run, without recording (40 minutes)

    Thinking about what to say (without writing a script), practising writing on screen while talking, planning use of the space on-screen, including how to lay things out and determining if I need to scroll down.

  2. Recording (23 minutes)

    Doing the recording in a single take makes editing simpler and avoids the risk of the on-screen material being inadvertently moved (by screen-scrolling or otherwise) and thus needing to redo the entire video for consistency. I also found it good to get into the habit of taking the mouse-pointer off the screen whenever not writing anything, to minimise jumps in the mouse position when cutting and editing the material. Finally, it was useful to put "firebreaks" into the video. That is, to pause after explaining one stage (with the mouse already off-screen), and think quietly about exactly what I was going to say in the next stage. The pause during the firebreak can easily be removed during post-processing. Also, if I make a mistake, I can just erase everything up to the firebreak and start again without needing to stop the recording.

  3. Editing down to 10 minutes (44 minutes)

    If the recording has been done in a single take, then it's just a case of cutting out pauses, mistakes, coughs and "um"s (the latter not entirely necessary, but makes the video smoother and is not difficult to do). It's also the opportunity to add in text if I've made a minor mistake that was not caught during recording, and to add a title or finishing page.

  4. Upload (13 minutes)

    Uploading to Youtube and creating the page on the scalable-learning platform.

http://www.jamescannon.net/images/video_lecture_timing.png

So in this example, a 10-minute video took almost 2 hours to prepare. This was for roughly the 9th video I made, so I'd practised already with 8 previous videos.

Information density is high. I did an anonymous survey in-class, and students claimed to take 2-3x the time of the video to study it, accounting for pauses and taking notes. So a 10-minute video is maybe equivalent to 20-30 minutes of class time, if the class was held at the pace that students were able to fully absorb the material. Thus the ratio of video preparation time to class time is about 4-6 times. Interestingly, scalable-learning suggests it should take no-longer than 1.1 to 1.5 times normal class time to record the lectures. I would love to hear from anyone reading this who has achieved such a ratio, how they managed that.

In-class

Using voting machines, I asked questions about concepts I wanted them to have picked up during the videos, and went over those concepts that they had trouble with.

Some students will diligently watch the videos and take notes. Others will not. The first time, the students received no credit for watching the videos, and therefore only 10 of the 14 students did. After I awarded a little credit for each video watched, all the students watched the videos. Here, the scalable-learning platform was helpful since I could see which students had watched the videos and assign marks accordingly. But even then, 2-3 students didn't bother taking written notes (the same students also had trouble with traditional homework too, so this is not specifically a video-flipped-classroom problem). This created difficulty back in the classroom since it would slow the pace of the class down as I had to go over concepts that I was hoping they all would have picked up in the videos. How to motivate such students when the lecturer has limited time to focus on individual students, is an issue. In contrast, the top-achieving students took detailed notes and I think they got a lot from this approach (very high marks in the final exam).

It's worth noting that in MOOC's (where this video-based approach is often applied), students who watch the videos are motivated to do so, since it is of their own free will, and if they drop out it doesn't matter. In a classroom environment, students don't necessarily want to be there, and it matters if they give up.

Summary

Make sure you are sufficiently experienced in teaching the subject before moving to a video-based flipped-classroom approach. This is for two reasons:

  1. Teaching anything requires you to learn the material in sufficient depth to teach it. Doing this on top of learning how to make the videos requires a significant investment of time, and I recommend to introduce a video element once time permits.
  2. It is necessary to have a feel for where students are going to have problems, and how long it will take to get through material. There is the risk of assigning topics for study by video before the class, only to not have time to discuss that topic because of student difficulties in previous topics.

It's necessary to balance calculation-style problems and concept-style problems in class. Watching videos for homework and applying their knowledge to answer questions about concepts in class, they students did miss more traditional calculation-style problems. I brought these in after the first few classes when I moved to a more traditional teaching approach again.

My own understanding benefitted. The need to break up the teaching into 5-10 minute topics really helped me build hierarchal understanding of knowledge required to master the subject, so this contributed to my own understanding too.

My recommendation: There are challenges concerning wider adoption by teaching staff, given the time, equipment and motivation required, however I can see this video-based flipped-classroom approach being an effective method of delivery for someone with reasonably deep experience of teaching the subject, when introduced gradually over a number of iterations.