Hacker News Reader

Comment by latch | Hacker News Reader

This article is a critique of the traditional aspect of teaching (having a teacher lecture students). This is obviously important, so it's good to have professional educators provide feedback.

However, I thought the aspects that was supposed to be "revolutionary" was detailed tracking and analysis of students and that feedback loop (either back into the system, to the student themselves or to teachers). They seem to completely ignore this.

Khan Academy isn't disruptive because of Salman Khan's lessons. It's disruptive because of the underlying platform, which any teacher (some possibly better than Salman) or any student could take advantage of.

View on HN

I think one of the most groundbreaking ideas they have is "flipping the classroom" [1], where students watch Sal's videos at home before class for homework, then in class they work on problems with the teacher and other students. That's where the ability for teachers to track and target deficiencies really becomes powerful.

[1]: http://www.usatoday.com/life/people/story/2012-05-30/sal-kha...

I have seen this firsthand in Massachusetts and it was very powerful. The kids were all clustered around iPads or Macbooks, helping each other, doing problems, and reviewing what they learned for homework. The teacher was rotating around the room, helping stuck kids, encouraging others to do peer-to-peer coaching, etc. It was so absurdly different than how I "did school," yet so logical. The teacher felt she could now really differentiate her instruction, and the kids were impressively engaged. Gone are the days of 40 minute lectures and 10 minutes of practice— you know, where the kids who got the lecture blaze through worksheets and the rest sit dumbfounded and then take all their problems & questions home for the evening.

All that to say: yes, a dude sketching some problems on a tablet is not the "revolutionary" part (if there is one at all) of Khan Academy.

I agree 200%. It also has a number of other nice side effects such as teaching teamwork/collaboration, etc.

I found the examples the article gave about students' common misconceptions interesting. I'm wondering whether the tracking and analysis by Khan Academy picks up these misconceptions as well. For example, in one of the examples (8 + 4 = _ + 5) many students answer 12 or 17. Khan Academy should be ideally placed to find such misconceptions if only because their sample size is much bigger than the average classroom.

"Khan Academy should be ideally placed..." I would say 'yes' and 'no' to this. Yes if the questions ask, and no if they don't. And that's the point of the piece. Khan won't pick up these misconceptions because he doesn't ask. He doesn't ask about decimals with different numbers of decimal places; he doesn't ask about the meaning of the equal sign. Maybe it's because he doesn't know that he needs to ask these things. That is a kind of knowledge he didn't need in his old profession and that he does in his new one. Maybe as KA grows, he will improve on this point and he will bring in people with this knowledge. But given Khan's dismissive ness of education research in the Harvard EdCast interview, I am skeptical.

Thousands (millions?) of teachers prepare mostly the same lessons every day. Whilst some preparation is obviously necessary, I see huge benefits for students and teachers from centralising this.

I am looking forward to the future when a video lesson distributed to 20% of children on the Monday can have its feedback aggregated and analysed, with a patched version of the lesson pushed out on Tuesday.

For example, imagine teaching a maths lesson, and seeing that a significant number of children paused and rewatched a particular 20-second section. You could re-film that and splice it in ready for testing the next day. Imagine the possibilities if you had schools offset their holidays by a few weeks!

  > teachers prepare mostly the same lessons every day

That "mostly" hides a lot. A good teacher will prepare this lesson in a context: context of his pupils, their abilities, their knowledge of other material, etc.

Centralized version will have none of that.

So who is making the site that will deliver more personalized instruction? Where is the research that site will use, telling all about which kinds of personalization are proven and how much effect they will have?

I don't see how an individual teacher has the advantage. Most classes have a huge range in ability and any lesson must be suitable for all. A centralised system could (in my dreams) deliver more than one version of each lesson and determine (from their enormous amounts of data) which lesson each child will benefit most from.

Of course if I'm dreaming that much I should point out that with each child able to learn at their own pace a single physical classroom could hold many different levels of the same subject at once, with children from different schools working together...

It is no coincidence that the biggest proponents of this style of learning have strong AI backgrounds. Their vision is a centralized system that can deliver that highly customized material on a per-student basis. It is still early days, but the technology has a lot of room to evolve over time.

Reply on news.ycombinator.com