Vocabulary Tests?

[rdearman]

I was recently reading through the logs and came across a vocabulary size tester. This particular test had a built in failure because if you know the words they are going to use, you can guess the meaning. OK, that seems confusing, but what I mean is, that if you took this particular test in your native language, then take it again in the target language you'll score much higher because all they are doing is using the same words translated. So I took the test once in French and they estimated my vocabulary at 5000+ words, then I took the test in English where I scored in top 3%, then having noticed the pattern took the test again and scored 6000+ on the French and 6000+ on Italian. Not because I got any better at either, but just because I knew what words they were going to use in advance.

So probably the real score is the one I took first which was just over 5k. This took me down a rabbit hole looking to see how vocabulary level might correlate to CEFR. This estimate is probably around B2. A google search brought me back here to the forum with a couple of threads about vocabulary size, and some interesting insights from s_allard and Casava.
https://forum.language-learners.org/vie ... =14&t=3424
https://forum.language-learners.org/vie ... =14&t=1990

What I was looking for was for another online test which would give an estimated vocabulary size in French. I quickly looked through reineke's nice list of proficiency tests
https://forum.language-learners.org/vie ... php?t=2897

But I couldn't find anything which is just a vocabulary size test. There are a lot of these in English, anyone know any for French. I'm just looking for estimated vocabulary size, not CEFR level, etc. Ideally I'd like to take 3-4 vocabulary size tests in order to get an average.

[tiia]

Here's one from a German company: Vocabulary French

Just mark the words you know (If you slide over the black box the German translation is shown, so really useful for you ) and then click "Weiter" (=next) and do it again until you're getting a result.

[rdearman]

Here's one from a German company: Vocabulary French

Just mark the words you know (If you slide over the black box the German translation is shown, so really useful for you ) and then click "Weiter" (=next) and do it again until you're getting a result.

OK, I can't read the result, but I'm guessing B2.

result_in_german.PNG

But what I'm really wanting is something which estimates the amount of vocabulary you know. For example in English (according to the tests I've taken) I know 30,000+ words.

[DaveBee]

The "Test de la Taille du Vocabulaire" (TTV). Scroll to bottom to the page linked to below.

https://www.lextutor.ca/tests/

EDIT
Sorry, that's not an online test. You download a PDF/excel file.

[smallwhite]

OK, I can't read the result, but I'm guessing B2.

But what I'm really wanting is something which estimates the amount of vocabulary you know.

Didn't believe the bolded "3900" would be your result, eh? This site is more stingy than Xmmm's one.

[Iversen]

I trust my own dictionary based word counts more than the internet tests, which apparently all work on the premise that certain words are difficult, and the more you know of those the larger your vocabulary is. But the presumably 'difficult' words are mostly learned words based on Latin, quaint old words and words denoting technical/scientific things - and for someone like me those are actually the easy ones, and therefore I tend to get inflated test scores. If you had tested me on artisans' tools or sports or modern slang I would score lower.

[s_allard]

I won't rehash all my writings on vocabulary size from HTLAL lest somebody complains again of me repeating myself but I want to make two points about how vocabulary tests work and why I think they are useless.

I'll skip the very fundamental question of what is a word for counting purposes and recall the notion of word family that I discussed in posts referred to above.

First of all, it is important to note that no test, at least that I know of, measures a person's vocabulary, i.e. actually counting the words they know. This is probably impossible. What all the tests do is test the knowledge of a small sample of words and extrapolate a figure for total words known.

The starting point of the methodology behind all this is a numbered frequency list of the words of the language. Then a key concept is that if you know a word of a certain rank you must know all the words below it. For example, if you know the word at rank 500, you must know all the 499 words below that. Similarly, if you know the word in position 10,000 you must have a vocabulary of 10,000 words. Things can be made a bit more sophisticated of course but that's the basic idea. So the tests will start with more common words and work up the list until the candidate can no longer continue. This way a test can estimate a person's vocabulary size using only, let's say, 30 questions.

There is also a difference between a test of receptive vocabulary where one is asked to recognize the definition of a word and a test of productive vocabulary where one may be asked to actually write the correct word in the phrase.

The key point here is that at no point is actual individual vocabulary being counted. Of course, one can say that this is what sampling theory is all about. This conventional approach does give us some idea at least for comparative purposes of relative vocabulary size. I can see the logic of this.

But the big problem with all these tests is that they are incapable of measuring how well the person can really understand (receptively) and use (productively) words. We know that many grammar and content words can have multiple meanings in different contexts and can be used in different ways. To say that one knows 10,000 words in French does not mean much more than one got a score of 10,000 words on an Internet test. The real question is what can one do with 10,000 words.

The only thing we can say with some certainty is that a bigger vocabulary is better than a smaller one.

Secondly, I should also point out that in all the years of discussing and bickering about this question at HTLAL only two persons, Iversen and myself, have actually counted real vocabulary size as measured by the different words used in HTLAL posts. This is a long and very tedious task. Iversen and myself used different definitions of a word so our figures may not be exactly comparable.

I don't have the figures before me. Maybe Iversen will provide his figures but as for myself I think I came up with a figure of a productive vocabulary of less than 1600 word families over a six-month period. I think my speaking vocabulary is even smaller. So what does this mean? It says that regardless of those test scores, the actual vocabulary that comes out of my month or from my fingers is actually quite small and is a far cry from all those fanciful figures we see concerning vocabulary size.

[Deinonysus]

I can attest to the inaccuracy of that first vocabulary test that's been going around. My wife and I both took the test and got the exact same result for French, a bit over 5,000 words, equivalent to an 8-year-old French child.

What's the problem with that, you might ask? Well, I'm a solid B1 in French, and I only got so many words because there were a lot of direct cognates to words I knew in English. My greatest French accomplishments include successfully ordering fried chicken at St-Hubert and managing to have a 30-second conversation with a 2-year-old French child with only one major grammatical error that was pointed out to me.

My wife is a professional French teacher and is often mistaken for a native speaker. She has aced graduate-level courses that were taught in French. She's at least a solid C1. She definitely has a vocabulary of well over 5,000 French words. She just happened to miss some of the random, obscure words on this test.

If a test says that she and I have the same vocabulary in French, then the test is wrong.

[Iversen]

Those who followed the link to my five-part language learning guide in my previous message may have noticed this illustration:

a2.jpg

In 2009 I did a first count of the words I actually had used in my old log at HTLAL, and then I did a second count in 2014, where I divided the corpus into two sections (called 2014.1 and 2014.2). The total number of words written by me in the corpus from 2014 was roughly 73.000, and the total number of unique headwords reached 5433 - a figure which of course depends on my criteria for reducing wordforms to headwords (in principle the same as dictionary words). But it may be more interesting that the overlap between the headwords from 2014.1 and 2014.2 (amounting to 3498 resp. 3914 words) was no less than 1979 words. Which simply means that if I had taken a third sample of around 37.000 and counted headwords in it, it is likely that up to half of the words would be overlap, and the rest would be new words.

But of course this couldn't continue forever: if I had produced ten samples of 35-40.000 words each (all written by me) then the proportion of totally new words in no. 10 would almost certainly be far less than half the words. But even the test with just two corpora suggests that I only use a small part of my total number of active words in English and even less of my passive vocabulary. With ten corpora or more I would of course be closer to emptying my complete stock of known words, but even message 11 would almost certainly contain some not hitherto represented words.

And that's where the word counts from dictionaries come in handy insofar they give an idea about the size of my passive vocabulary in English. These counts have yielded figures between 30.000 and 40.000. The internet tests I have done have generally resulted figures at the same magnitude as my own counts so some amount of sophistication must have gone into their construction, but as S_Allard I remain fairly sceptical about the idea that there is a simple link between the rarest words you know and the total size of your vocabulary. If I had done the same tests in Romanian (where I know a higher percentage of the numerous loanwords from French and Latin than of the indigenous words) there would probably be a larger gap between the two kinds of test.

[smallwhite]

I plan my L2 vocabulary acquisition very carefully, taking both frequency distribution and CEFR scales into account and aiming for B2~C1 and real life needs. I feel that my vocabulary is distributed as planned and has few glaring gaps.

I took 3 arealme tests (from Xmmm) and 2 sprachenlernen24 tests (from tiia), and I found all 5 results reasonable and consistent.

Greek: I have 5xxx flashcards which based on my experience should be able to take me to B2, but I studied these cards only casually, only L2->L1, and I have ignored most of them for quite a while now. Meanwhile Dialang has been giving me consistent and reasonable scores. I haven't taken the vocab paper lately but based on past results I believe I'd score B1 now. Tiia's test placed me at B1 and I agree.

French: Tiia's test placed me at C2. I was officially C1 with full marks in vocab six years ago. I haven't touched French much since but I read a novel once in a while, looking up vocab and revising them also once in a while, and I've been studying Spanish and Italian. C2 now sounds reasonable.


Note that the CEFR scale is not a frequency scale. A1 topics don't necessarily involve more frequent words than B2 topics.