Proficiency determination

[IronMike]

(Not sure this counts as practical.)

I'm looking for recommendations for books or articles (in English, Esperanto or Russian) that discuss how to judge (analyze) a passage (written and/or audio) to determine its proficiency level.

I tried several keyword searches in these forums, but didn't have success in finding what I'm looking for.

Thank you in advance for all help you can give.

[Serpent]

Do you mean judging the speaker's proficiency or figuring out the level "required" for learners?

[mcthulhu]

https://vimeo.com/album/139578 has examples of English speech for each of the Interagency Language Roundtable proficiency levels, if that would help.

There's actually a full distance learning course available on how to rate passages by proficiency level, at http://www.nflc.umd.edu/projects/language/pr. There's a free demo video online that only goes up to level 1+, but it might be of interest. There's a link on how to purchase the course, if you're really motivated.

https://ll.mit.edu/mission/cybersec/pub ... Jones2.pdf discusses the ILR levels, and has one short sample level 2 text.

http://fletcher.tufts.edu/Orientation-F ... equirement has level 3 practice tests in a variety of languages, including Russian.

There's probably a lot more out there if you Google ILR language proficiency, and there are also different scales besides ILR.

[IronMike]

Thanks @mcthulhu! That's exactly what I'm looking for.

@serpent, I'm looking on how to determine what level a certain text or utterance is, say if I wanted to compile a list of passages for students to test what level they're at.

[rdearman]

The unix style command will rate a passage for reading levels. It does English, and I think other languages?

https://www.linux.com/news/improve-your ... e-checkers

[Serpent]

Two threads from the old forum

In general I'm sceptical of assigning levels to texts, though. Are you teaching English? There are lots of self-assessment tests online, Dialang presumably being one of the best. I'm hopefully stating the obvious but the same text can be assigned a different level depending on whether the learner is expected to understand it or produce something similar.

[jeff_lindqvist]

I think there's more to proficiency and reading comprehension than meets the eye. I can blaze through a Dan Brown novel, while that is not the case for J.K. Rowling's books (I can't remember if I even finished the second one). I doubt my Illuminati vocabulary is larger than the one for magic. Maybe LIX can give a clue?

[mcthulhu]

Readability scores and language proficiency levels aren't really the same thing, and they serve different purposes. Relative difficulty of vocabulary isn't the determining factor in proficiency levels. A former colleague of mine once commented how strange it was that expressions of opinion or "editorializing," even in words of one syllable, could be considered a level 3 text, while merely factual language like a calculus textbook, for instance, would only be level 2, even though the calculus textbook would be a lot harder to read.

Any broad framework for levels of language difficulty is necessarily going to be arbitrary. It's a matter of practicality and expediency. Maybe you could invent a more perfect system with a thousand gradations along a continuum of increasing difficulty, instead of just a handful of broad categories; but good luck getting anyone to use it or even agree on it. I think systems like the ILR, CEFR, etc. are here to stay, imperfect as they are. Educational systems, employers, etc. need to measure our skills against some sort of standard, whether we enjoy being graded or not. Another former colleague complained about being tested, and argued that our employer ought to just take our proficiency on faith, and not require us to prove it. This was wishful thinking at the time, and I think it still is. Self-assessment has very limited applicability.

"the same text can be assigned a different level depending on whether the learner is expected to understand it or produce something similar" - no, not in any passage rating guidance I've ever seen. Reading and speaking (or writing) skills are usually tested separately and not with the same passages. Your experience may be different.

[Serpent]

"the same text can be assigned a different level depending on whether the learner is expected to understand it or produce something similar" - no, not in any passage rating guidance I've ever seen. Reading and speaking (or writing) skills are usually tested separately and not with the same passages. Your experience may be different.

I simply mean that rating is ambiguous and can refer to evaluating a learner's level (judging them by a written/spoken sample), or to determining the difficulty of a passage in terms of comprehension.

So for example in Esperanto you can use a C2-learner's text at a future B1/B2 exam, with some minor edits maybe