“a 3-paragraph blog post describing how you would tell if live real-time tutoring was better than, worse than, or just as effective as (a) recorded video and (b) live tutoring.”
I think the assignment as stated has a fair amount of misdirection. If the question is “which is more effective” how you measure that can and should be essentially decoupled from how the instruction method. In other words, if I’m measuring how well a student has learned Python’s list comprehensions, my assessment of that knowledge shouldn’t depend on whether they learned from reading a tutorial, watching a screencast, online-live tutored, or tutored in person. The “how I would tell” part would be which method leads to greatest measured results on my assessment. So I could talk about assessment design in general, but instead I want to focus on the video vs live distinction. I’ll talk about a subjective set of pros for each method, and then how you could perhaps isolate some of those experimentally.
For live teaching some of the advantages are:
Can quickly survey the student(s) as to prior knowledge and experience and adapt material in response
Can pause and check to see if people have any questions along the way
Can ask people about what they do, and then use authentic examples in exercises
Can be available if someone gets stuck in an exercise and needs help
The main advantage I see to video recording is that sometimes teaching is a little like performing — I might teach the same thing 10 times, and they might all vary in quality. With a recorded video, you can capture “the best” of the performance, and editing enhances this result. Note I’m not saying this makes recordings “better” than live teaching, only that it is an advantage.
I’m going to start with the assumption that live teaching is more effective, and ponder how you could test the relative contribution of its advantages listed above toward this effectiveness. Each experiment would use the same hypothetical assessment instrument mentioned at the beginning
You could survey students, and then present one of 3-4 versions of a video targeted to different levels.
You could have the video broken into several logical chunks, watched over time, with live Q&A in between
I don’t think there is any practical way you could record videos for different authentic uses, but you could do that for the sake of this experiment with 3-4 groups.
You could have the instruction be static video, with the exercises done live with help.
Looking at the results from these experiments would begin to clarify WHY live teaching is more effective.