Updated: Further comments on reading code at end.
The testing sequence features five lectures on testing and one lecture on exception handling, for a total of one hour. I don’t actually see much opportunity to improve them. The presentation follows the guidelines recommended in the cognitive load of multimedia paper, with the voice-over covering the points presented on the screen. The slides feature progressive reveals, with the current point boldfaced. The slide text is often annotated to emphasize key points.
Several times the viewer is challenged with an exercise, the narration stopping to give the viewer room to think.
The videos also exploited a unique possibility of video lecturing (or lecturing with a writable surface, such as a blackboard): A statement such as, “You can raise exceptions yourself”, was presented and explained, then it was revised, crossing out “can” and replacing it with “should”. Although counterpart techniques exist for purely-oral and purely-written presentation, they are not quite the same. I found this technique effective.
There was one way in which these videos—and nearly all other computing-related instructional videos—violated the cognitive load guidelines. As soon as code was presented, I stopped the video and read the code, then restarted it. The way the presentations explained the code wasn’t the way that I read code for understanding, so the narration overloaded me. I found it easiest to come to my own understanding of the code first, then resume the video and follow the narration as it highlighted how the code exemplified the point being made.
Although the the videos were competently made and the topic interested me, it seemed like a very long hour. This was partly due to a lack of affect in the narration but mostly due to the difficulty of paying sustained attention to video lectures. I don’t know how to solve this problem.
Update: Reading code in the light of cognitive load theory:
Mayer and Moreno’s Type 2 overload describes my problems reading code while narration is going on. Type 2 overload occurs when the learner’s cognitive capacities are overloaded by essential aspects of the material. When reading code (or results presented as graphs or tables), I have to stop the narration, scan and integrate the code to form my own understanding, then restart the narration, at which point I can follow the narration as it highlights features of the code.
This response is an implementation of both the strategies Mayer and Moreno recommend for Type 2 overload: (i) Segment the lesson, allowing the learner to assimilate material at their own pace during pauses, and (ii) pretrain the learner so that they already recognize the components and the narration can show the relationships between those components.