1. Forum
  2. >
  3. Topic: Duolingo
  4. >
  5. Ratio of translations from le…

https://www.duolingo.com/profile/sebastian.wl

Ratio of translations from learning language to native language (and vice versa)

Hello,

something that really bothers me is that when I learn language X, the by far greater part of the exercises is translating from the language you're trying to learn to your native language, which I find much easier than the other way round. Usually I storm through the translations from X to native and only have to pause and think (sometimes real hard) when a couple of native to X questions come at the end of an exercise. I believe the learning effect would be greater if there were more tasks to translate from the native language to the language you are trying to learn. Maybe adjusted to your fluency level since this would make things harder in the beginning but becomes increasingly more interesting as your command of the language improves.

What is the intention behind this if there is any? Any way to change this (maybe it already does based on your fluency level or any other statistic?)? Any plan to change this in the future?

January 17, 2018

5 Comments


https://www.duolingo.com/profile/Peter594672

Right... but the fact is that this is actually one of the two most efficient approaches to learning: first you learn to understand, second you try to actively use what you've understood. (The other is throwing immersing you into the X language environment and watching you try to swim out).

Duo supplies basic courses for beginners, and apparently tries to do so in a way that keeps a beginner's life somewhat tolerable.

While the Skill Levels feature is on its way, you might try to find a reverse tree where people try to learn your native language (or a language you already know) from your target language. The to/from ratio will be lopsided there just the way you might like ;)


https://www.duolingo.com/profile/sebastian.wl

I agree - a steep learning curve will discourage beginners. Ideally you would be able to set the ratio yourself because everyone has their own ambitions and pain threshold ;)

As simple as your suggestion is, I hadn't thought at all about just reversing native and learning language in duolingo :) Thanks for the idea, I will try it out!


https://www.duolingo.com/profile/pentaan

Duolingo uses a very good teaching method for beginners and for people, who want to brush up their school knowledge.

  1. In the course "French for English speakers" you are learning the grammar and the pronunciation. You will mostly translate from French to English and the user interface is in English.

  2. In the "reverse tree", the course "English for French speakers", you will mostly translate from English to French. The user interface is in French. And you can start to read (and write) in the French forums (set the microphone and sound to OFF in Duolingo's settings)

  3. In the "laddering trees" you can do "Foreign language 2" from "Foreign language 1" and reverse (if available in Duolingo).

For instance:

  • "Spanish for French speakers" and the reverse tree
  • "German for French speakers" and the reverse tree
  • "Spanish for German speakers" and the reverse tree

Laddering courses are my favourite courses, because .....

  • you will always translate to your target languages
  • in each course you will learn a lot of new words and sentences, because they are made by different teams
  • you can read and write in your target languages in Duolingo's discussion forums

https://www.duolingo.com/profile/sebastian.wl

I do not doubt that the teaching method of Duolingo is good, especially for beginners. The "reverse tree" method as you call it is a workaround though, as viable and sensible as it may be!


https://www.duolingo.com/profile/VjeranC

Given the way things work, the features are not tailored toward individuals but towards all of the users together. So if having more translations to target language affects the time spent on learning negatively when looked at all of the users, feature will not be implemented. So, even if you, an individual spent more time learning with more translations to target language, because on the whole, people were learning less, you won't get them.

If you wanted to make the service more tailored to the individual you either add feature flags and similar, but then you leave the risk to the user. User could self-sabotage oneself and not realize that and quit learning, or could make it too easy and not learn at all.

There are automatic ways of figuring things out on an individual level but I'm not sure it's simple enough to make it work.

I personally enjoy translations to target language and timed practice. I believe that if you have to do something fast and correctly, you learn better than if you allow yourself to do it slow.

I'd also like more features towards timing like seeing how fast I translate sentences X words long (to target language) over time, but I'm not sure if anything like that is ever going to appear.

Learn a language in just 5 minutes a day. For free.