I am going to follow up my last post with a topic on the same theme: academic tools.
Video lectures are the future of learning physics.a The ability to speed up the rate of playbackb , or go back over a point you missed without wasting everyone else’s time, is just so much more efficient than a live lecture. Of course, getting answers to questions is crucial and video lectures lack this mechanism of feedback, but this can easily be supplemented with office hours and by taking courses alongside others.
But the real power of video lectures is the ability for everyone to view the very best lectures on any given topic, and moreover for individuals to find lectures that are tailored to their particular expertise. This is just so much better than what is possible when you are restricted to hear lectures from people in the same building. When each course has 100 different versions taught by different lecturers ranked by popularity, and when the best lecturers can improve through targeted feedback from thousands of students, the quality is going to skyrocket.
Perimeter is one of the places leading the way on this. They record all their major lectures (and many of their minor, unimportant talks) and put them all up on the website, for free and fully searchable.c Once lots of institutions start doing this and once they (hopefully) collaborate on a large, public, and well-indexed database (a “video ArXiv”), the process of learning physics will take a huge leap forward.
Of course, the technological infrastructure to do mass, reliable recording was presumably not cheap, and PI benefits from the fact that it is well-funded and young, and that it shrewdly invested in this infrastructure when it was constructing its building. Retrofitting older buildings (and older administrative minds…) will be challenging, but hopefully unstoppable momentum will build once people see how useful this stuff is.
There are a lot of good lectures out there, and a tremendous number of bad ones. I am looking forward to a future that fixes this.
Footnotes
(↵ returns to text)
- …at least for the lecture medium. The other key component for the future of learning physics will be massive collaboratively edited textbooks and review articles (i.e., Wikipedia+GitHub for physics). I have more thoughts on this for another post. (EDIT: Now posted here.)↵
- Download a lecture and play it in VLC. Use the ‘[‘ and ‘]’ keys to change playback speed. This is very satisfying, as your whims literally control the pace at which the lecturer goes through the material.↵
- Currently they don’t have a direct link to the mp4 video files for technical reasons, but you can use savefrom.net to save the file to your computer and then playback with VLC.↵
As a physics student, I absolutely agree. Do you have any thoughts on how a shift to freely accessible video lectures and collaboratively edited textbooks will impact higher education as a whole? Over at her MathBabe blog, Cathy O’Neil expressed concern that Math research is a byproduct of Calculus teaching, and that a shift towards online courses will reduce the number of math research positions. Are office hours enough to keep physics departments intact?
Here’s a link to the post I mentioned: http://mathbabe.org/2014/01/20/the-coming-calculus-mooc-revolution-and-the-end-of-math-research/
Right, so this is basically the process of “unbundling” (in Econ speak). I’m familiar with the basic concepts I don’t know enough to say anything about effect size with confidence. It’s definitely true that there are times when certain goods that would otherwise be under-produced (such as non-excludable public goods) can be beneficially subsidized through bundling, and we should be careful anytime a long-standing beneficial subsidy is destroyed through unbundling. But in general unbundling is good because it allows products to be specialized, and people to buy only what they need. (Think of the benefit of unbundling smart phone software from the cellular service provider: Smart phones would be undeniably worse if — for technical, monopolistic, or regulatory reasons — only Verizon could provide phone apps.)
Furthermore, note that the calculus-research subsidies are rather crude currently, in the sense that they are accidents of reputation systems and the demand for undergraduate education, neither of which need to be well-correlated with the research needs of society. When these subsidies are eliminated, it’s perfectly possible for government or philanthropic foundations to step in and directly subsidize research in a more sensible way. For example, it may make more sense to concentrate the researchers in fewer institutions where they may benefit from interacting with each other, rather than spreading them out in tiny math departments for every small college just so they can teach a few large calculus classes.
As O’Neil points out, this will probably not have a big impact on the math and physics research done at elite schools, since those professors have grants from places like the NSF to support their research. Given the fact that the impact of research (both as measured by citations, and as measured by conventional wisdom) is rather concentrated in elite institutions (both schools and independent places like IAS), we shouldn’t expect the impact on the *research* produced for society to be drastic. As you can tell from O’Neil’s emphasis, this is perhaps more of a jobs issue than a research issue. Note also that many schools have been increasingly hiring adjuncts and lecturers to teach service courses like calculus for years, and that this process would continue even in the absence of MOOCs.
However, I think there is an even bigger effect: higher education is more about hard-to-fake signaling of intrinsic human characteristics than it is about teaching, a position I have been convinced of by Bryan Caplan. (See however a qualification from his colleague GMU.) As Caplan notes, colleges won’t feel compelled to move to MOOC’s if they don’t really care if their students learn, and unusual teaching styles would work against the (claimed important) signal of conformity.
My hunch is that if you compare the negative effects of removing the calculus-research subsidy (which will be muted, per above) with the positive benefits to students from taking great courses (which can occur *regardless* of whether the colleges officially incorporate them into the courses), the later would dominate by one or more orders of magnitude. However, I really would need to take a look at some numbers to make this more than a guess. And in any case, I agree that there will be some serious shake-ups in the jobs market that young researchers should carefully consider.