JLPT 2022 thread(Results out now!)

I used the Core10K deck as the vocab part of my N1 study, and I passed N1 fairly easily. Vocab/kanji was not my best area, certainly, but I did well enough not to be in danger of failing because of a low score in that area. My personal feeling is that general reading ability matters more – including both speed and comfort with comprehension of texts where you don’t recognize all the vocab.

4 Likes

That’s definitely my experience after taking the N2 this year as well. It’s less about knowing all the words and more about being able to navigate through text. It’s more problematic with listening passages with lots of keigo, of course.

That being said, if 5-6k words is the recommendation for N2 and one knows only that much in general, it’s extremely easy to come across words one doesn’t know. I think if one specifically learns common words with a high chance of appearing at the JLPT, one might have a better chance of scoring high.

6 Likes

yeah, probably if you learn “the most common 6k words in JLPT exams”, then it might be enough for N2, but who does that? :wink:

1 Like

How do you count your vocabulary knowledge btw?
I would like to know how much vocabulary I know.

Does a vocabulary count as known if you can read it and understand what it means or do you have to be able to use it as well?

Can confirm. Knew all the words, still flunked (N1) :sweat_smile:.

4 Likes

There’s the just-for-fun vocab estimator described in this article. It rated me at 15,300, which doesn’t seem like it’s too massively wrong. It’s only testing recognition.

5 Likes

Either the estimator (didn’t know there was one for Japanese, neat. Only used one for English before),
or in my case I counted the ~6500 vocab items on Wanikani (learned all the vocab), + ~2000 (flash-)cards in Anki, plus miscellaneous words that I know but didn’t make cards for. And that’s not including Bunpro grammar points (which you could often see as extended vocabulary flashcards).

In my case I count recognition/understanding (at least one reading and meaning, ideally a couple meanings if they differ a lot), because reading and listening is what I learn for the most.

1 Like

The estimator unfortunately didn’t work for me. First I had to repeat step 1 (50 words), because 100 words in step 2 was too much for me, then I chose 0 words for step 2, and after doing 50 words again, the result was NaN (not a number, well-known programming error). It shouldn’t let me choose 0 if that doesn’t work.

Also, it has super obvious katakana words like ロボット and チャンピオンシップ, which might skew the results or make you get more or less lucky depending on how many you get.

So, a tip for others: At least set step 2 to 50 or something and click Apply before you start. If you change the options during the selection, you’ll have to start over again. But you’ll still have to check for each of the 100 words whether you know them, can take quite a while.
Maybe I’ll do it again some time.

1 Like

Yeah, ideally you should probably do it a few times and kind of average the results or something. But the nice thing is that the randomness means it can be meaningfully repeated, unlike many other ones that always use the same set of words. I got 24000 the first time, and 31000 the next time. So there can be a decent amount of fluctuation.

3 Likes

Definitely, it’s a good test, it’s just a bit of a chore on my level to check 100 or 150 words in Japanese when there are so many more or less rare ones included (by design) where you have to think for a while.

Yeah, some of the words are pretty ridiculous, and I feel they would never show up on a curated vocab test. But maybe it’s intentional like you said.

The first word in my list of 100 the second time was , which after checking should be read サンチ, an obsolete alternative for センチ.

1 Like

Yeah, I think that’s how the test works, it shows you sets of random words from multiple frequency ranges, and estimates from there. If you know 2 of 10 obscure words, it’s likely you know a bit of rare vocab.

But yeah, the fact that you have to manually check each one to make sure you’re not being lulled into a false sense of security is a little annoying. With 珊, I was like, oh that’s 珊瑚’s 珊, surely it’s just a shortening of that or means that on its own as well, which is common for kanji use in difficult compounds. So I easily might have just checked it and moved on.

3 Likes

Ok I was too curious and did a 50+50, result: 14539 (rounding = 1)
Pretty sure I got lucky and it’s overestimated, quite a few obvious katakana words, etc.
Though I think there are always words you’ll know the reading and meaning of even if you haven’t seen them before, like composite words. I don’t know.

Well that’s a bit of a relieve, because the 6k+ words from wk don’t match the JLPT very well from what I know :thinking:

Yeah, the katakana-word problem is tricky – but on the other hand they are still words and they do bump your effective vocabulary size up quite a bit in the real world too. Anyway, this kind of thing is why I said “just for fun” :slight_smile:

a lot of them don’t, WK has some rare words, but it also has a lot of common words you’ll find in the JLPT. And that’s why I’ve prioritized the most common words with Anki. I’ve written a tool that inserts the frequency of words into your Anki cards and mostly learned the most common ones first.

1 Like

Well, I finished the inscription process a couple of days ago, so I’m oficially doing N2 this december!

I’m gonna have to start doing some mock tests so I can see what my weak points are (spoilers: it’s still gonna be listening)

6 Likes

I took the test five times with these results:

26.900
16.300
16.600
31.700
30.400

When I got the lowest points I was super strict and didn’t check Hiragana words as known, which seemed to have been more grammar items that vocabulary, like られる.
I found that out the third time because three of these happened to show up at the beginning and none made any sense as stand alone word.

Don’t know how much this reflects reality though. :sweat_smile:

1 Like

has N2 registrations opened? I thought they start 3 months before the exam

1 Like