Back

Chart of Ability Levels and Assesments

#1
Hello!

I'm trying to gather feedback on a chart showing different levels of Japanese ability, various indicators of that level, and some convertible assessment results, and then give a rough idea of how many hours of studying (all forms) that would take.  After getting some feedback elsewhere, I figured it was probably time to make a koohii account and ask y'all!

The chart assumes a "level wear" (meaning no strengths, and no weaknesses) and an average level of aptitude.  Questions of study efficiency and knowledge decay are certainly relevant, but I'm starting with average levels of efficiency, and minimal decay.  Basically, everything being improbabilistically average, how much studying should get you where, and how will you know?


[Image: iqVEfsX.png]

So what do you think?  Do these numbers jive with your experience?  Do you have contemporaneous J-CAT and JLPT scores that disagree with my comparisons?  Do you think the hours or vocab are off-base?  What's your impression overall?  Thanks!
Reply
#2
I'm unable to answer your question but I'm sure other people here will be able to give you a good answer Smile
Are you Italian? I'm asking because your nick sounds Italian
Reply
#3
Hello. ^_^

The name is Latin, but I'm an American.
Reply
August Sale (14th - 25th): 30% OFF Premium PLUS - 25% OFF Premium
JapanesePod101
#4
I know the US State Department has a language proficiency rating rubric that's pretty well established. The one for Japanese could be aligned with JLPT scores maybe?
Reply
#5
Here's my suggestions (note, I'm nowhere near N2 so it's an educated guess).

Replace "ignorance" with "beginner" 

N5 is  J-CAT 100, takes about 150 hours, 1000 vocabulary and ~150 kanji. Equivalent is finishing Genki I.

N4 is J-CAT 150, takes about 300 hours, 2000 vocabulary and ~350 kanji. Equivalent is finishing Genki II.

N3 is J-CAT 200, takes about 600 hours, 4000 vocabulary, ~700 kanji. Equivalent to finishing Tobira.

N2 is J-CAT 250, takes about 900 hours, 6000 vocabulary, ~1000 kanji. Equivalent to finishing Shinkanzen Master N2 series.

N1 is J-CAT 300, takes about 1500 hours, 10000 vocabulary, ~2000 kanji. Equivalent to finishing the SKM N1 series.

Biggest change is the J-CAT score as they tend to be 1 level higher than they should be.  The time is assuming steady learning of kanji, vocabulary and grammar without the necessary reading/listening to native material that should be happening but is not actually studying.
Edited: 2017-08-12, 12:47 am
Reply
#6
Wow N1 in 1500 hours. That seems pretty doable. Of course you have to allocate the hours correctly and not have any long interruptions along the way.
Reply
#7
My first thought is: "This kind of stuff is really, really hard to do."

The JLPT has an official "can do self-evaluation" that can probably help you. It covers the four main language competencies (listening, speaking, reading and writing). And I think that the survey is probably done contemporaneously with people who actually take the test. So it is probably the best source out there for "What does JLPT mean for real life?". Here's the URL: http://www.jlpt.jp/e/about/pdf/cdslist_e_all.pdf.

The JLPT also surveys people for how many hours they've studied, and things like that. Every time I take the test I kinda guess for those questions though! Those questions are hard for people who self-study - especially because lots of people put down their studies for a few years and then pick it up.

JCAT also tries to map to JLPT. But they are on the old system (1-kyuu - 4-kyuu, as opposed to N1-N5).

As someone else mentioned, the US government has at least a few organizations that measure language performance. When I was interested in understanding the broader context of my language ability, I found reading the Interagency Language Roundtable's self-assessments to be interesting. You can read them here: http://govtilr.org/. (The "self assessment" section is in the lower left of the page). I liked this system because it specifically addresses speaking, which the JLPT does not.
Reply
#8
(2017-08-12, 12:44 am)Nukemarine Wrote: Biggest change is the J-CAT score as they tend to be 1 level higher than they should be.  The time is assuming steady learning of kanji, vocabulary and grammar without the necessary reading/listening to native material that should be happening but is not actually studying.

I beg to differ. My JCAT score was 256 when I took the N1 and scored 132/180 on it.

Oh and OP, Murakami's language is simpler than light novels. Maybe replace it with something like Mishima or Kawabata?
Edited: 2017-08-12, 10:55 pm
Reply
#9
To Nuke and Risu,

I initially did have all of the JLPT levels up one rank, but someone in my conversation group with an N1 and roughly contemporaneous JCAT scores convinced me that if the JLPT scores were made explicitly the "just barely passing" scores, that that would be most accurate.  That said, both you, Risu, and my IRL friend made their argument based on N1 scores... it seems obvious to use 50 point increments on the JCAT, but now I'm questioning if some of the JLPT levels might not be too close to one-another to do that...

Does anyone have JLPT and JCAT scores from within a month of one-another for levels other than N1?  I took the N4 in December and got exactly 90 points, but then I didn't take the JCAT until mid March, and got a 176 (after I'd been aggressively studying all that time).  I'm a bit concerned about that as, according the chart, that would be an increase of ~700 hours over the course of 103 days, or an average of nearly 7 hours of studying per day which seems quite a bit higher than my estimate of 4 hours of studying per day.

So yeah, I'm concerned that N5, N4 and maybe N3 are lower, even if N2 and N1 are right on.

(2017-08-12, 11:07 am)ChestnutMouse Wrote: Wow N1 in 1500 hours.  That seems pretty doable.  Of course you have to allocate the hours correctly and not have any long interruptions along the way.

Yeah, those were the two caveats about skill decay and efficiency.  This assumes average efficiency and minimal decay.
Edited: 2017-08-13, 1:05 am
Reply
#10
The vocab size way too low while the 4,500 and 6,000 kanji values are a tad ridiculous.
Reply
#11
It's worth considering adding the Kanken levels into there.
Reply
#12
(2017-08-13, 2:14 am)mintou Wrote: The vocab size way too low while the 4,500 and 6,000 kanji values are a tad ridiculous.

Yes, if 6,500 kanji is your standard for perfection, than only <1% of Japanese people have "perfect" Japanese.
Reply
#13
(2017-08-13, 2:14 am)mintou Wrote: The vocab size way too low while the 4,500 and 6,000 kanji values are a tad ridiculous.

Which of the vocab sizes are you referring to?  All of them?  I know native speakers are generally said to top out around 30k, and I based the earlier ones on the "core" decks above the JLPT suggested vocabulary sizes, and then just sort of filled in the gaps.

As for the high end of kanji, yeah, the idea was that most Japanese people would be at "Full Fluency", but only 国語 college majors and the like would be at "Academic" and nobody who wasn't a lifelong scholar would ever achieve "Perfect".  That was my thinking anyway.  What do you think is more appropriate?

(2017-08-13, 2:23 am)Nukemarine Wrote: It's worth considering adding the Kanken levels into there.

Good idea!  I'll look into it, thanks!

(2017-08-13, 5:03 am)kusogaijin Wrote:
(2017-08-13, 2:14 am)mintou Wrote: The vocab size way too low while the 4,500 and 6,000 kanji values are a tad ridiculous.

Yes, if 6,500 kanji is your standard for perfection, than only <1% of Japanese people have "perfect" Japanese.

Honestly I was looking for something like 0.1%, but yeah.  I mean... would you just rather the last line wasn't on there?  I don't know where to go beyond "Academic" that isn't a vanishingly small minority.
Edited: 2017-08-13, 9:45 am
Reply
#14
I've been studying for almost 6 months and maybe about 4-6 hours a day. I took the J-CAT about 20 days ago and scored 303. I would say my vocab size is about 15000-20000 (have at least finished 10k) and kanji ~3000. I haven't taken N1 but I'll be taking it end of this year.

I think the listening/speech capacity can't really be aligned with the other categories because it's kind of a totally different skill. While I can read quite a lot, I don't have fast natural speech at all.

The chart is overall a bit too arbitrary (particularly the J-CAT scores don't line up I think) as well as the hours. It depends how effectively you spend the hours not how many hours kinda. The chart only gives a very rough idea of Japanese ability I suppose.
Reply
#15
(2017-08-15, 2:36 am)doingnot Wrote: I've been studying for almost 6 months and maybe about 4-6 hours a day. I took the J-CAT about 20 days ago and scored 303. I would say my vocab size is about 15000-20000 (have at least finished 10k) and kanji ~3000. I haven't taken N1 but I'll be taking it end of this year.

I think the listening/speech capacity can't really be aligned with the other categories because it's kind of a totally different skill. While I can read quite a lot, I don't have fast natural speech at all.

The chart is overall a bit too arbitrary (particularly the J-CAT scores don't line up I think) as well as the hours. It depends how effectively you spend the hours not how many hours kinda. The chart only gives a very rough idea of Japanese ability I suppose.

That is superb progress in very little time @doingnot.

Quick questions - is your vocab both reading and listening? Writing? Do you speak other languages?
Reply
#16
(2017-08-15, 12:13 pm)scooter1 Wrote:
(2017-08-15, 2:36 am)doingnot Wrote: I've been studying for almost 6 months and maybe about 4-6 hours a day. I took the J-CAT about 20 days ago and scored 303. I would say my vocab size is about 15000-20000 (have at least finished 10k) and kanji ~3000. I haven't taken N1 but I'll be taking it end of this year.

I think the listening/speech capacity can't really be aligned with the other categories because it's kind of a totally different skill. While I can read quite a lot, I don't have fast natural speech at all.

The chart is overall a bit too arbitrary (particularly the J-CAT scores don't line up I think) as well as the hours. It depends how effectively you spend the hours not how many hours kinda. The chart only gives a very rough idea of Japanese ability I suppose.

That is superb progress in very little time @doingnot.

Quick questions - is your vocab both reading and listening? Writing? Do you speak other languages?

I think it's in both reading and listening though sometimes I'll see a word and not be able to read it but if given the reading then I would know the meaning.. By writing if you mean like usable words maybe like 5000+ and if you mean actual physical writing then maybe only like 3000+... Just english ^ ^
Reply
#17
(2017-08-13, 2:14 am)mintou Wrote: The vocab size way too low while the 4,500 and 6,000 kanji values are a tad ridiculous.

I agree that 4500-6000 kanji seems a bit disconnected from reality.
Reply
#18
(2017-08-18, 11:00 am)ChestnutMouse Wrote:
(2017-08-13, 2:14 am)mintou Wrote: The vocab size way too low while the 4,500 and 6,000 kanji values are a tad ridiculous.

I agree that 4500-6000 kanji seems a bit disconnected from reality.

It's an elite accomplishment for natives and something needed for those wanting to pass Kanken level 1.
Reply
#19
What's the motivation to make a chart like this?
I don't think the realities of language learning or the meaning of words like fluent or functional are really accurately expressed here.
Reply
#20
The number of kanji at each level seems beyond the corresponding speaking ability.

Also the gap from 1150 to 2150 kanji seems big to me. That range probably includes 90% of the people who ever get past 1150, so for practical purposes it might be better to spread that over more slices.
Reply
#21
(2017-08-19, 4:07 pm)Furikake Wrote: The number of kanji at each level seems beyond the corresponding speaking ability.

Also the gap from 1150 to 2150 kanji seems big to me. That range probably includes 90% of the people who ever get past 1150, so for practical purposes it might be better to spread that over more slices.

A stop gap could be inserted, but that gap still represents the difference between N2 and N1. Interestingly enough, while 1100 kanji represent ~90% use by frequency, adding another 1000 kanji raises it up to 98% frequency use (an 8% increase). Diminishing returns and the Zipf principle can be a pain in the butt.

(2017-08-19, 4:02 am)m8719705030 Wrote: What's the motivation to make a chart like this?
I don't think the realities of language learning or the meaning of words like fluent or functional are really accurately expressed here.

One can think of the chart as useful targets for balanced growth. If you're learning the language and get up to 6000 vocabulary, perhaps ensure a grasp of 1000 or so kanji and around 400 grammar points. That's in addition to the hundreds of hours of listening to Japanese dramas and hundreds of pages reading Japanese novels.
Edited: 2017-08-19, 10:15 pm
Reply
#22
(2017-08-19, 10:13 pm)Nukemarine Wrote:

Yeah, that's why so few people get past 2150. Past 2500 or so, I just don't see the point. Pass a useless kanken exam? It just seems like the pursuit of vanity more than anything else. We're better off using our Japanese ability for something beneficial to the community, not pursuing more kanji trivia.
Reply
#23
(2017-08-20, 9:02 pm)Furikake Wrote: Yeah, that's why so few people get past 2150. Past 2500 or so, I just don't see the point. Pass a useless kanken exam? It just seems like the pursuit of vanity more than anything else. We're better off using our Japanese ability for something beneficial to the community, not pursuing more kanji trivia.

Umm, you lost me on that one. Everyone has their own reasons for studying and what they find enjoyable. If somebody finds pleasure in studying 3000 to 6000 kanji and 20 to 30 thousand vocabulary words, more power to them. They have just as much reason as those that spend hundreds of hours toward online gaming.

I'd just caution beginners to have a balanced approach of grammar, vocabulary and kanji instead of doing what I did which was spend 4 months at an hour or so a day learning 2000 kanji in English before starting on actual Japanese. That's where the chart actually helps.
Reply
#24
(2017-08-20, 9:23 pm)Nukemarine Wrote:
(2017-08-20, 9:02 pm)Furikake Wrote: Yeah, that's why so few people get past 2150. Past 2500 or so, I just don't see the point. Pass a useless kanken exam? It just seems like the pursuit of vanity more than anything else. We're better off using our Japanese ability for something beneficial to the community, not pursuing more kanji trivia.

Umm, you lost me on that one. Everyone has their own reasons for studying and what they find enjoyable. If somebody finds pleasure in studying 3000 to 6000 kanji and 20 to 30 thousand vocabulary words, more power to them. They have just as much reason as those that spend hundreds of hours toward online gaming.

I'd just caution beginners to have a balanced approach of grammar, vocabulary and kanji instead of doing what I did which was spend 4 months at an hour or so a day learning 2000 kanji in English before starting on actual Japanese. That's where the chart actually helps.
""They have just as much reason as those that spend hundreds of hours toward online gaming."

That's exactly what I mean. 

If they're doing it just as a pastime then yeah, we can point out some other equally useless ways to spend one's time. But if their goal is to learn Japanese, they should spend that time another way, like reading, speaking with Japanese people, etc.

I'm not trying to change your mind or anything, but I think it's important to have strong pushback lest the forum set people up with goals that are really not practical.
Reply
#25
(2017-08-20, 9:02 pm)Furikake Wrote: Yeah, that's why so few people get past 2150.
I think also that once you get past intermediate or so a lot of people stop counting kanji; at least I did. The concept of what it means to "know" a kanji is pretty fuzzy anyway, and there's no well defined finishing line, so why count?
Reply