Khan Academy.page

Neal:

> http://singularityhub.com/2011/02/13/yes-the-khan-academy-is-the-future-of-education-video/
> Looks like they've got some pretty sophisticated features, but still no SRS. Watch the video for full experience.

> Some questions: how do we position ourselves and/or differentiate ourselves from the khan academy? They've got vastly more funds, and a lot of similar ideas.

> However, it seems we have the potential to add some value on top of the core content features-- specifically, synced notes to the video, and spaced repetition! Right now, they measure "proficiency" in modules based on a 10-question correct streak, but it's not clear you ever go back and review from previous modules. So in that situation, some spaced repetition algorithm could add a lot if benefit-- looks like they collect a lot of analytics on time to answer questions, etc. Which could be used to assign an "ease factor" to that module in an SRS algorithm!

Ryan: 

> They have the exercise and video stuff down really really nice. Yeah, they don't have SRS (they could easily add it were it seen as an advantage of a competitor), but I think there's a more important issue. They and a whole lot of other competitors stop at the stage of providing exercises for stuff that is extremely easy to make into exercises. So while it's cool to earn badges and shit, I don't think there will be a huge fundamental difference if little Johnny uses Khan or one of 10,000 other arithmetic exercise programs. Of course the video content and being web accessible and decently useable is really nice. 

> Something that we are struggling with now -- not solved yet obviously-- is having arbitrary content and using collaboration to solve the problem of generating neat little exercises for all of human knowledge. I do think KA has a leg up in the fact that Khan is a really smart guy and can probably figure out how do that for a wide range of subject matter. But I think the collaboration stuff potentially has a much greater power of the masses. Just extremely hard to do right.

> Or... if you can't beat 'em, join 'em? It's an open source project with a strong community involvement: https://sites.google.com/a/khanacademy.org/forge/for-developers/getting-started-with-the-code
Maybe Gates can give us an inside connection to open up a research arm... (But who knows what conflict of interest it'd have with CMU or GT's stuff....)

Ryan:

> http://www.wired.com/magazine/2011/07/ff_khan/all/1

> Best part is Bill Gates' quote arguing against constructionists. Won't spoil it for you.

> Very interesting link from the comments: https://fnoschese.wordpress.com/2011/02/21/pt-pseudoteaching-mit-physics/

Neal: 

> Some takeaway thoughts after reading the article + comments + some of the outbound links: I'm still not sure what to think about Khan. In a way, the fnoschese posts do seem a bit curmudgeon-y and reactionary against a potentially disruptive mode of instruction, but at the same time the pseudo-teaching evidence is hard to ignore. But again, at the same time, if the progress of students using Khan academy is as big as they say it is... (although i have a hard time understanding what "advancing an astonishing 366%" really means). My tentative conclusion: having video is better than no video. Example-- Su's lectures. A dedicated student with video recording of a lecture will likely get WAY more out of that class than a similarly dedicated student who is madly scribbling notes to keep up.

> There are undoubtedly better ways to present this material, but the Khan lectures at least put the material into students' own hands whereas in the past they had to rely on teacher-as-keeper-of-knowledge, lacking immediate access when they felt motivated.

Ryan:

> https://github.com/khan/khan-exercises - might be worth checking this out... avoid not invented here syndrome!

> Looks like they are also Google docs fans: https://spreadsheets.google.com/spreadsheet/ccc?key=0AsgWawUKHSJldGlvX3RUX2FyMEpMdzdRRWlOLXg3TVE&hl=en_US&authkey=CJWi-LMM#gid=8

> I also keep going back and forth on Khan, and I agree that basically things are so bad that a little bit goes a long way. I also think there must be something to the way Khan explains things that he is so popular and effective. The fact that he hasn't invested any of those millions into other teachers means he's either very arrogant, or he and others working with him are convinced that he's actually uniquely good. One point that Prof. Benjamin made (on our car trip to Magic Castle) is that the number one problem we have in math education is that teachers aren't good enough at what they are teaching. Clearly, Khan is. 

> I might go into more about this on the LS+ post, but I'll state my position on constructionism (and Meyer's stuff seems a little less extreme; let's call it "contextualism" for kicks), which is a little different after reading Mindstorms: any institution of day-in-day-out education needs a dose of constructionism and contextualism to make education meaningful and tolerable for students. However, for any serious learning, there is going to have to be some grind (AKA "deliberate practice"). It can be fun, gamified grind, maybe with some pretty pictures attached, but the actual learning comes from the grind, and with little variation on whether it's enhanced with web 2.0, fun stories, LaTeX, social features, Britney Spears, etc. That stuff can certainly help motivate to start doing the grind, and that's extremely important as well, but when you are trying to teach your brain a fact or procedure, the presentation matters less. (Something like the [Heisig Method]() is a little different though because the presentation is actually critical to how the memory is formed. You can try to argue that is true Meyer's or Papert's stuff as well, but it's much less direct and studies have _not_ verified any causation.)