5 ms·

I've been reading similar anecdotes lately trying to gauge whether I could reach B1 in a Romance language in a few months of study. What I've found is a lot of

I've been reading similar anecdotes lately trying to gauge whether I could reach B1 in a Romance language in a few months of study. What I've found is a lot of pessimism about how well Duolingo teaches you a language, with many arguing it's a waste of time altogether. I'm surprised that could be the case though. They have data from millions of language learners processing through their lessons, and looking at their making-of blog (https://making.duolingo.com https://making.duolingo.com), the product seems data-driven. They do A/B testing to see the efficacy of new features they roll out. Aside from that, they combine reading, writing, speaking, and listening exercises, and have made some innovative features like chat bots. One language tree includes a few thousand words. If anything, I would expect Duolingo to be vastly more effective than other methods.

gibolt 7y ago
Their A/B testing probably isn't testing education efficacy, but retention and actions that lead to payment/miscellaneous business goals.
- abecedarius 7y ago
  Ironically, as a user I'd use it more if it were more effective and less grind-grind-grind. But I suppose you're probably right. An efficient, effective learning process feels more challenging and frustrating, which must drive a lot of people away. It must be a solvable problem, though.
  - wes1350 7y ago
    Author here. I agree that Duolingo is almost certainly not optimizing for language learning effectiveness, which ends up being somewhat of a side-effect of their business model. However, by using the desktop version and turning off the word bank, you can better optimize for effectiveness for yourself. Duolingo is certainly a grind, but I don't think that's 100% a bad thing. It's far from perfectly executed, but I think you need a certain amount of grind to really drill some of the vocab and grammatical concepts into you. As an example, in a classroom you do some homework exercises, and maybe get feedback the next day on however many exercises you do. You don't necessarily get to redo the exercises you miss enough to let the concept sink in, and you might do way too many exercises that do nothing for you as you've already internalized the concept. With Duolingo grinding, you have to repeat your weakest points after you get them wrong and get to drill in all the exercises until they stick. I estimate that each lesson of roughly 20 questions, with redoing exercises I got wrong, takes me something like 3-5 minutes, so in an hour of focused effort you might get 300 exercises in. That's far more exercises than I imagine you do in school for example, and you also get to decide if a unit is too easy and you can afford to skip it, or if it's hard enough where you have to repeat it. Plus, each sentence has a discussion page, many of which (especially the tricky ones) have detailed explanations about the concepts within the sentence. Duolingo is far from perfect, but despite all the criticism it receives (rightly so for what it claims you can do with 5 minutes a day on the mobile app), you can actually make it a surprisingly effective tool.
    - abecedarius 7y ago
      Oh, I agree it's useful! I'm pretty sure it's a great improvement over my high-school Spanish classes or my rudiments of grade-school French in Canada. It's just that I moved on nearly as soon as I could to learning by osmosis from reading real books; which, not being deliberate practice, is quite suboptimal in learning rate too, but is a much more tolerable use of time. If they were as effective as I think they could be, I wouldn't ditch them so soon. I have the unusual issue of being severely hard of hearing and I'm not sure what'll be the effective way to learn spoken conversation if I do move to a French-speaking country. Even hearing English is, uh, problematic in most settings. So this is also an example of tailoring your own approach in view of goals and context. I like reading books, and this most directly got me the thing I like.
gbrown 7y ago
I feel like Duolingo is a good "in": something to give you enough of a foothold in a language to start learning in other ways. The barrier to entry is nonexistent, but it does become hard to progress past a certain point.

gibolt 7y ago

Their A/B testing probably isn't testing education efficacy, but retention and actions that lead to payment/miscellaneous business goals.

abecedarius 7y ago
Ironically, as a user I'd use it more if it were more effective and less grind-grind-grind. But I suppose you're probably right. An efficient, effective learning process feels more challenging and frustrating, which must drive a lot of people away. It must be a solvable problem, though.
- wes1350 7y ago
  Author here. I agree that Duolingo is almost certainly not optimizing for language learning effectiveness, which ends up being somewhat of a side-effect of their business model. However, by using the desktop version and turning off the word bank, you can better optimize for effectiveness for yourself. Duolingo is certainly a grind, but I don't think that's 100% a bad thing. It's far from perfectly executed, but I think you need a certain amount of grind to really drill some of the vocab and grammatical concepts into you. As an example, in a classroom you do some homework exercises, and maybe get feedback the next day on however many exercises you do. You don't necessarily get to redo the exercises you miss enough to let the concept sink in, and you might do way too many exercises that do nothing for you as you've already internalized the concept. With Duolingo grinding, you have to repeat your weakest points after you get them wrong and get to drill in all the exercises until they stick. I estimate that each lesson of roughly 20 questions, with redoing exercises I got wrong, takes me something like 3-5 minutes, so in an hour of focused effort you might get 300 exercises in. That's far more exercises than I imagine you do in school for example, and you also get to decide if a unit is too easy and you can afford to skip it, or if it's hard enough where you have to repeat it. Plus, each sentence has a discussion page, many of which (especially the tricky ones) have detailed explanations about the concepts within the sentence. Duolingo is far from perfect, but despite all the criticism it receives (rightly so for what it claims you can do with 5 minutes a day on the mobile app), you can actually make it a surprisingly effective tool.
  - abecedarius 7y ago
    Oh, I agree it's useful! I'm pretty sure it's a great improvement over my high-school Spanish classes or my rudiments of grade-school French in Canada. It's just that I moved on nearly as soon as I could to learning by osmosis from reading real books; which, not being deliberate practice, is quite suboptimal in learning rate too, but is a much more tolerable use of time. If they were as effective as I think they could be, I wouldn't ditch them so soon. I have the unusual issue of being severely hard of hearing and I'm not sure what'll be the effective way to learn spoken conversation if I do move to a French-speaking country. Even hearing English is, uh, problematic in most settings. So this is also an example of tailoring your own approach in view of goals and context. I like reading books, and this most directly got me the thing I like.

gbrown 7y ago

I feel like Duolingo is a good "in": something to give you enough of a foothold in a language to start learning in other ways. The barrier to entry is nonexistent, but it does become hard to progress past a certain point.