The data coming out from edX experience is intriguing, but there are a lot of facets in video preparation; so probably my original question in my mind “perfect duration for video eLearning content” was flawed. That data from edX findings indicated that the shorter video is (maxed out in 6 minutes duration) the more likely students stay watching the video (engagement?) until the end of the video; and anything longer than 6 minutes, students will only watch the first 6 minutes or less; hence wasted efforts.
Well first of all, it is a median value, so we don’t know if the actual data are homogenous or has large variance; I think the problem with what data we have is, we can't have identical video (since they already different length to begin with) to compare, both the long-video and short-video group are watching different videos; and the time factor here is inherently corresponds with risk of the video going bad. In layman term, it's probably relatively easier to make short good quality video, than to make long equally good quality video; even if we put aside the matter about the viewer perception, because the longer the video is, the more chance for the producer to make mistakes.
If that's the case, then a short video will have a tendency to consistently be perceived of having higher quality than longer video, although the quality is not necessarily the function of the length of the video per se. Thus it could be not accurate to say, if we want to make a better video, then make a short one; but instead it could be more accurate to say, if you want a higher chance to make better video, then make a short one so you have less time to screw up.
We might want to bring in another vocabulary at this point, which is probably the word "concise"; it's not the matter of the length, but then density of information within that length of video. That is what I suspect have a greater impact to engagement than the length per se. maybe name it Conciseness Index, and it’s like a measure of quality content per time duration. If we have two videos of significantly differing length, but consistently have the same conciseness index throughout the length, then see if they have different in-video drop out, if there is, then we can regard the length of the video as a contributing factor in itself to student engagement.
When students faced with bad or very bad short video, and they see that the video progress bar is already pass half-way mark and close to the end, they might think, well I’ll just sit tight until the end of this video, the video is playing but their mind is disengaged. Maybe, maybe not, we don’t know.
But my comment here is neither to defend nor to attack short or long video; which seems to be an interesting research project in itself; but taking the practical educator’s hat, we need to shift focus to the conciseness of our videos. We are making 6 minutes video, or 16 or 60 minutes video, good, how do we use every second in that video? Are we wasting the time? Ours to produce and students to watch? Have we made a good use of it?
Before we record our videos, either a full blown video production, or simply voice/video over presentation, have we prepare and make us of the time to have enough conciseness that does not over load the students, but also not putting them to sleep.