Study on Language Improvement with Native Materials

[MamaPata]

This really interesting- I'm intrigued.

You mention that people would do a test before they start, do oyu have a plan about what kind of test you will ask us to do? Presumably it varies a lot by language.

[rdearman]

This really interesting- I'm intrigued.

You mention that people would do a test before they start, do oyu have a plan about what kind of test you will ask us to do? Presumably it varies a lot by language.

If this was a well funded study by a university or something then I would have everyone get a certification test. However, it isn't so testing will be done like this.

• Everyone will do a self-certification test.
• A reputable online test will be given if available for the TL
• If possible a brief skype assessment by a native speaker (very unlikely however)

So if you're doing a minority or endangered language or even a conlang for which there is no objective testing available then I'll need to depend on the results of the self-assesments.

[smallwhite]

The subs2srs deck isn't obligatory, but I would give everyone assistance and instructions to help them do it.

How about you create decks for some of the participants? If it takes 2 hours for a first-time subs2srs-user to create a deck then you create it but they enter 2 hours onto their timesheet?

So the test group restrictions are:
[*]Use subs2srs to create at least 1 deck to review and learn the dialogue of a film or TV show.

The subs2srs deck isn't obligatory

I'm confused.

[rdearman]

The subs2srs deck isn't obligatory, but I would give everyone assistance and instructions to help them do it.

How about you create decks for some of the participants? If it takes 2 hours for a first-time subs2srs-user to create a deck then you create it but they enter 2 hours onto their timesheet?

So the test group restrictions are:
[*]Use subs2srs to create at least 1 deck to review and learn the dialogue of a film or TV show.

The subs2srs deck isn't obligatory

I'm confused.

I will not create your decks for you, and for copyright reasons couldn't. But I'll give instructions and answer any questions and try to help as much as I can. Because some people may have to pay for anki, I don't want to make it obligatory, or a reason not to participate, but it is highly recommended and if you can then you should. I don't want anyone to have to fork out cash just to participate in the study. But if you can get it free, and you have the DVD resources and computer to install the software, then I want you to use the subs2srs.

[rdearman]

Just wanted to lay out a timetable so people are aware.

17th-27th --- nothing, I'm on holiday.
28th June - 17th July --- I'm going to write up the instructions for the subs2srs, write the questionnaire, create the tracking spreadsheet, any other admin work - like creating a google drive, mailing list, etc.
29th-30st July --- send out list of tests to participants and collect the initial data. (a weekend)
31st of July --- Start
31st January 2018 --- Get everyone to do the second test, and collect and collate all the data.
February 2018, --- publish results

[blaurebell]

I could give you an Excel sheet with my entire French journey, which is mostly native material with minimal grammar. It's been almost a year and the only grammar I've done was half of Assimil and 90% of Duolingo, and a bit of pronunciation training at the beginning. Amounts to a total of 84h of formal stuff and so far 772h of native material, 304h of audio and I should be done with 10,000 pages by July with about 470h of reading so far. The test at the beginning would have been unnecessary, I started out at A0, but with B2 in Spanish and A2 in Italian up front. If you want I can do the test for evaluation after I hit 10,000 pages, should be in 2 weeks at the latest, which would give you some idea of what kind of results you can expect. I assume you're thinking of Dialang as the online test? That said, I started off with LWT though, no anki, and I'm not sure that's what you had in mind? My progression was intensive reading for the first 5000 pages, then starting to watch series with subtitles for the first 2 seasons of dubbed series, then extensive reading / kindle and watching/listening. I'm right now following the same progression for Russian and I can participate in the native material group for that if you want, I'm already beyond my initial grammar boost and won't do any more grammar until I hit 10,000 pages at least.

If I were you I wouldn't expect much from A0 beginners who start entirely without grammar though. With some languages it's impossible to decipher native material without grammar - Russian or German for example because of the cases and word order. Even after 150h of grammar in Russian I still get sentences where I know all the words and should theoretically be able to figure out the grammar of the sentence, but simply can't parse the damn thing. In my experience half of a Duolingo tree usually is enough of an overview to start reading though, even with a language like Russian. One thing I can tell you already from the comparison between French and Russian is that I seem to need more than double the time for Russian, because I have way less cognates for that one and Russian has way more vocabulary and word forms in general. You will probably have to keep cognates and closeness of languages in mind when comparing the data.

[trui]

This sounds more like a study on language improvement with anki rather than on language improvement with native materials.

Would someone who just wants to use native materials (and no SRS) ideally be a part of the control group?

[Cavesa]

Yes, if I happen to be in the test group, I would go without any SRS too, like trui suggests.

I think it may be wise if we all used a few weeks to not be at complete point 0 at the beginning of the experiment (yes, I am crazy and considering a new language for the purpose of the study, somebody stop me please ). But perhaps the truly starting learners are an important part of it?

So, with your permission, I would like to write a short reddit post with link to this thread, perhaps we'll catch a few participants there.

I am still quite curious about the distribution of people into the individual groups, as it will be difficult to divide people evenly, given the very diverse background.

This is going to be awesome!

[rdearman]

Short post before I head off to airport, so last you'll hear from me for awhile. But although Anki & Subs2srs isn't required, you'll need to create spaced repetition / review for the words you're learning while watching or reading material. Spaced repetition is required for word review. I don't care if you use wordlists, goldlists, paper flashcards, anki, excel, etc. But SRS is required.

[Cavesa]

you'll need to create spaced repetition / review for the words you're learning while watching or reading material. Spaced repetition is required for word review. I don't care if you use wordlists, goldlists, paper flashcards, anki, excel, etc. But SRS is required.

Ah, I'll need a bit more information on that. Is that going to be part of the testing or in what way is it important? You know, I usually don't review words at all while using native input. In these cases, my SRS is simply reading/watching/listening on as the words will come up again. If I have to use a way to review the words, it undermines the whole principle of learning with extensive reading/listening, I'd say, and it complicates the matters for me. It doesn't mean I wouldn't participate, but perhaps more in the control group as I always fail at trying to SRS for several months in a row and as having to read/listen intensively, and review it, would change the whole point and perhaps add workload I cannot afford.