+- kanji koohii FORUM (http://forum.koohii.com)
+-- Forum: Learning Japanese (http://forum.koohii.com/forum-4.html)
+--- Forum: JLPT, Jobs & College in Japan (http://forum.koohii.com/forum-12.html)
+--- Thread: 2011 JLPT study thread (/thread-6834.html)
2011 JLPT study thread - jankensan - 2012-02-29
http://www.jlpt.jp/e/about/pdf/scaledscore_e.pdf
Here are the details...Seems complicated to me!
2011 JLPT study thread - Zorlee - 2012-02-29
radical_tyro Wrote:congrats... that's damn impressive! how have you been studying?I actually made a thread about my study-method (nothing fancy) a while ago.
Check it out if you're interested: http://forum.koohii.com/showthread.php?tid=8456&page=1
2011 JLPT study thread - kusterdu - 2012-02-29
Zorlee Wrote:Zorlee, did you close your Youtube account?radical_tyro Wrote:congrats... that's damn impressive! how have you been studying?I actually made a thread about my study-method (nothing fancy) a while ago.
Check it out if you're interested: http://forum.koohii.com/showthread.php?tid=8456&page=1
2011 JLPT study thread - Zorlee - 2012-02-29
Oh, yeah, forgot about that. I made a new one.
I'll see if I can re-upload the movie about accent etc.
EDIT: Now it should be viewable
Sorry about that!
2011 JLPT study thread - Splatted - 2012-02-29
thisiskyle Wrote:The point of the scaling does not seem to be comparing the people that took the test in
the same sitting with each other to determine scores.
thisiskyle Wrote:The test should not be a competition and, based on what I can find, it's not. There are not a limited number of "foreigners with good Japanese skills" slots to fill so it would make no sense to compete. I couldn't find anything saying that the tests are curved so that a certain percentage will pass every time.Unfortunately, the PDF that Jankensan posted shows that it is a competition. The maximum number of passes is set based on the percentage that passed the old tests, so you do have to do better than other people taking the test at the same time.
I think that's good. If I am at a certain level of proficiency, my score should reflect that regardless of how good or bad others did.
I couldn't find a list of previous years pass rates so I may be wrong, but I remember reading somewhere that the yearly pass rate was increasing steadily. This makes sense, since new software, multiculturalism and increasingly easy access to Japanese media is making language learning a lot easier than it used to be, and it means that they're effectively increasing the pass mark each year.
2011 JLPT study thread - Guoguodi - 2012-02-29
Finally got my test result here in Australia. Passed N3!
Language knowledge: 60/60
Reading: 49/60
Listening: 51/60
Total: 160/180
2011 JLPT study thread - Zgarbas - 2012-02-29
So anyone else just reading this thread to be proud of how awesome the community results are?
2011 JLPT study thread - thisiskyle - 2012-02-29
Splatted Wrote:Unfortunately, the PDF that Jankensan posted shows that it is a competition. The maximum number of passes is set based on the percentage that passed the old tests, so you do have to do better than other people taking the test at the same time.I read that article and it doesn't anywhere mention having a set percentage of participants passing or failing the test each year. It would make no sense if that were the case. You would expect if a bunch of people with N1 or higher proficiency levels took the N5 test (to the exclusion of all others, this is purely hypothetical), the vast majority of them would pass. If the pass rate was set at, say, 70% however, 30% of those highly proficient individuals could still expect to fail the N5 test.
I couldn't find a list of previous years pass rates so I may be wrong, but I remember reading somewhere that the yearly pass rate was increasing steadily. This makes sense, since new software, multiculturalism and increasingly easy access to Japanese media is making language learning a lot easier than it used to be, and it means that they're effectively increasing the pass mark each year.
Regardless, your second paragraph seems to contradict the first. In the first, you seem to be saying that, by design, only a certain percentage will pass each year but in the second that the pass rate has been increasing.
2011 JLPT study thread - kainzero - 2012-02-29
Zgarbas Wrote:So anyone else just reading this thread to be proud of how awesome the community results are?not really. i'm more disappointed in my own results and i can't really take credit for other people's hard work. heh.
i mean how many of us have really all learned the same way? i don't think any one of us has used the same material. some of us don't even recommend RTK =)
2011 JLPT study thread - Zgarbas - 2012-02-29
Not about getting credit for other's work, I just find it refreshing in comparison to how many learning communities pan out to be. Less complaining about how difficult hiragana is and how you should get a scholarship to Japan for knowing 100 kanjis and more excellent results from people who are willing to share their study methods and resources with us at the bottom, without being holier-than-thou about it. Stuff like that.
That being said, I don't think passing the N2 is anything to be disappointed about, despite the near miss
.
2011 JLPT study thread - radical_tyro - 2012-02-29
Zorlee Wrote:yes, that's exactly what i was hoping existed. thank you for sharing!radical_tyro Wrote:congrats... that's damn impressive! how have you been studying?I actually made a thread about my study-method (nothing fancy) a while ago.
Check it out if you're interested: http://forum.koohii.com/showthread.php?tid=8456&page=1
2011 JLPT study thread - erlog - 2012-02-29
thisiskyle Wrote:I read that article and it doesn't anywhere mention having a set percentage of participants passing or failing the test each year. It would make no sense if that were the case. You would expect if a bunch of people with N1 or higher proficiency levels took the N5 test (to the exclusion of all others, this is purely hypothetical), the vast majority of them would pass. If the pass rate was set at, say, 70% however, 30% of those highly proficient individuals could still expect to fail the N5 test.They don't have a set percentage, but there is scaling. The grade you get back since they changed the test is not a flat score of right questions and wrong questions. They have released the curves for past tests, but not yet for the Dec. 2011 test.
Regardless, your second paragraph seems to contradict the first. In the first, you seem to be saying that, by design, only a certain percentage will pass each year but in the second that the pass rate has been increasing.
This isn't really worth arguing about because it is completely explained in these two pdf's:
http://www.jlpt.jp/statistics/pdf/2011_1_6.pdf
http://www.jlpt.jp/statistics/pdf/2011_1_1.pdf
2011 JLPT study thread - kainzero - 2012-03-01
Zgarbas Wrote:Not about getting credit for other's work, I just find it refreshing in comparison to how many learning communities pan out to be. Less complaining about how difficult hiragana is and how you should get a scholarship to Japan for knowing 100 kanjis and more excellent results from people who are willing to share their study methods and resources with us at the bottom, without being holier-than-thou about it. Stuff like that.ahh. i see.
That being said, I don't think passing the N2 is anything to be disappointed about, despite the near miss
from that perspective it's been great.
i feel like this community, and investing the time to learn japanese on my own was really something new for me. seeing everyone get results because we helped each other is awesome. plus i feel like i triumphed over other people, people who needed classes, non-natives in japan, etc.
from the language perspective it's been pretty horrible.
vocab/grammar, you can choose the right words but you may not even understand the sentence.
reading i did horrible. might've been because of time. on top of that, i really wanted to read the articles again because i really wanted to know what they were saying, not because i wanted to answer some questions.
and listening, honestly, if i passed, it's because i knew which answers were wrong, not because i understood the question. =)
2011 JLPT study thread - tregingigan - 2012-03-01
erlog Wrote:I see. Thanks for the advice! But I'll have to wait the December Test anyway. No July Test in Italy.tregingigan Wrote:This apart, your opinion gentlemen. I study, as average, 1-2 hrs a day. Enough to try N2 next December or i'm aiming too high? N2 grammar, for instance, seems a lot more than N3 and rather complex, too.The grammar section of the test is so tiny now that I wouldn't worry too much about that aspect. 1-2 hours a day for the next year should be fine for you to pass N2 in December. You could probably pass it in July.
Changing subject, congrats to Zorlee. Get to such a good result in 2.5 years is awesome! I immediately checked your study method thread ...
2011 JLPT study thread - Tori-kun - 2012-03-01
@kainzero: Wow.. With me it's the opposite when it comes to listening/reading! I feel like my ears are plugged with something, like dirt or so and my brain is full of porridge when I hear Japanese talking, although I do listen a lot to Japanese podcasts of every kind every day.. at least as much as I read articles on the web.
2011 JLPT study thread - Splatted - 2012-03-03
thisiskyle Wrote:You're right, I misread it, but that's still basically what the new scoring system does. If more people get high scores on one years test than another, they will assume it's because the test was easier and adjust the scores accordingly. This is their stated reason for the change in scoring, so I don't know why you think it's not what they're doing.Splatted Wrote:Unfortunately, the PDF that Jankensan posted shows that it is a competition. The maximum number of passes is set based on the percentage that passed the old tests, so you do have to do better than other people taking the test at the same time.I read that article and it doesn't anywhere mention having a set percentage of participants passing or failing the test each year.
I couldn't find a list of previous years pass rates so I may be wrong, but I remember reading somewhere that the yearly pass rate was increasing steadily. This makes sense, since new software, multiculturalism and increasingly easy access to Japanese media is making language learning a lot easier than it used to be, and it means that they're effectively increasing the pass mark each year.
thisiskyle Wrote:It would make no sense if that were the case. You would expect if a bunch of people with N1 or higher proficiency levels took the N5 test (to the exclusion of all others, this is purely hypothetical), the vast majority of them would pass. If the pass rate was set at, say, 70% however, 30% of those highly proficient individuals could still expect to fail the N5 test.This is exactly the same example I wanted to make. If those N1s took the test they would all get really high scores, so the scaled scores would be set based on the assumption that it was a really easy test, and those few that made stupid mistakes would fail despite being well above what would have been the pass mark if the test was scored conventionally.
thisiskyle Wrote:Regardless, your second paragraph seems to contradict the first. In the first, you seem to be saying that, by design, only a certain percentage will pass each year but in the second that the pass rate has been increasing.Sorry, that was badly written. When I said the pass rate was increasing I meant before they changed the scoring system.
2011 JLPT study thread - zigmonty - 2012-03-03
Splatted Wrote:Most of the adjusting they are doing is on a question by question basis. After everyone's done the test, they decide which questions are worth 3 points, 2 points, and 1 point. But the section has to add up to 60, so they can't just allocate everything as 1 point (all easy questions). Franky, that's better than the examiners guessing what the hard questions are.thisiskyle Wrote:It would make no sense if that were the case. You would expect if a bunch of people with N1 or higher proficiency levels took the N5 test (to the exclusion of all others, this is purely hypothetical), the vast majority of them would pass. If the pass rate was set at, say, 70% however, 30% of those highly proficient individuals could still expect to fail the N5 test.This is exactly the same example I wanted to make. If those N1s took the test they would all get really high scores, so the scaled scores would be set based on the assumption that it was a really easy test, and those few that made stupid mistakes would fail despite being well above what would have been the pass mark if the test was scored conventionally.
They then set the pass mark such that the right number of people pass (are they even changing this year on year or did they just set it once the first year?). It doesn't seem like they do a final tweak to the score, or there's be no need to have such random numbers as the pass marks (and different between all levels). And if they were scaling the final number to achieve a certain pass rate... then a consistent number of people would be passing each year, which they aren't. If they're fiddling with it, they're not fiddling with it to the extent you're imagining.
IMHO, if only N1 people took the test, they would all pass, but their scores would magnify the small mistakes some made. Any questions everyone got right would be scored 1, with 3s being assigned to the ones some made silly mistakes on (such that the section added up to 60). In any case, this is a worst case situation. If you take the test that is appropriate for you, their statistics work and everyone is happy. You're not seriously arguing that ~100,000 applicants isn't statistically significant?
2011 JLPT study thread - erlog - 2012-03-03
To be very very clear. Theoretically if only super fluent people took N1 they would still all pass. The pass/fail score floor is always 100 points and is NOT set as a comparison to other people taking the test. It is consistent with every single test. However, individual questions within the test are scaled in order to balance hard/easy material. This means that in order to fail you still have to get lots of questions wrong and significantly deviate from the mean passing score.
With only 60 points to allocate among the questions it's simply not possible to stratify the scores in the way people are fearing in this thread. Even in a worst case scenario of maybe 10 questions being worth triple with the rest as single points there would still be more than enough points available among those easy questions to receive a passing score on that section of the test. What's more likely is that they aren't even setting questions worth triple, and probably only have a narrow scaling in order to be conservative with their algorithm. This also isn't being done by a human, and so they can be as granular as they want with questions having a scaling of 1.1131321 points or whatever.
The issue being complained about is partly why the passing floor on the exam is so low. The way they have it set up now actually has way way more risk of handing a passing certificate to an unqualified participant.
This is why they tell you your flat score on vocab/grammar with the A/B/C/D grades. I'm also glad they release the test score curves so you can see how well you did as a comparison to other test-takers. Right now it's tough to know what to make of my score sheet other than needing to hammer reading really hard.
2011 JLPT study thread - SomeCallMeChris - 2012-03-03
合格です！
レベルN2
言語知識 34/60 読解 47/60 聴解 31/60 総合得点 112/180
文治・語彙 B 文法 A
I have no idea what to make of my scores... how did I score so low in points in language knowledge if my letter scores are A/B?
How did I score so -high- in Reading if I didn't -finish- the reading section (by a good third of the material!)
Listening at least makes sense - listening comprehension is my weakest skill and after finding that my reading speed -still- wasn't up to par I was in a very poor mental state.
Well, I did feel that my reading comprehension was near 100% for what I did complete so I suppose if I got 40/40 for the what I did complete then getting 7/20 for the rest isn't far from the expected 5/20 from random bubble filling.
Anyway, for anyone curious - this is my second time taking N2; I've never studied to the test - I don't own a single sou-matome or kanken or whatever book; the closest I've come to studying to the test is using vocabulary resources based on the old level... A JLPT 3 anki deck, and the readthekanji website I used for awhile is based on JLPT vocabulary. Oh, well, I did do the online practice test for N2. Most of my actual reading material has been manga, and most of my grammar study has been Tae Kim's guide.
This is not the fast route to passing - the vast majority of the vocabulary used in adventure fiction is not tested by JLPT. It's taken me... many years to get here. (Vocabulary deterioration from various months-long breaks is a factor too... ah the joy of 'I used to know that word' ... I only started using SRS three or four months before the test, too, also a factor in the way I regularly lost heaps of vocabulary.)
Nonetheless, I'm reading ラノベ and 推理小説 now, and will probably challenge JLPT N1 in an equally haphazard and unprepared fashion. I'll probably pick up 謎解きはディーナーの後で and incidentally brush up on my keigo by reading the adventures of that eccentric 刑事お嬢様 and her crazy butler... The drama version is fun, I'll certainly at least finish watching that.
2011 JLPT study thread - kainzero - 2012-03-04
SomeCallMeChris Wrote:I have no idea what to make of my scores... how did I score so low in points in language knowledge if my letter scores are A/B?An A means you got more than 67% right, B means you got more than 34% right. (C means you got less than 34% right.)
2011 JLPT study thread - Tori-kun - 2012-03-04
@SomeCallMeChris: おめでとう～
2011 JLPT study thread - Rina - 2012-03-05
Ｎ２、合格しました！
言語知識：５４・６０
読解：３８・６０
聴解：３９・６０
総合得点：１３１・１８０
2011 JLPT study thread - jettyke - 2012-03-05
CarolinaCG Wrote:Ｎ２、合格しました！おめでとう！！！
言語知識：５４・６０
読解：３８・６０
聴解：３９・６０
総合得点：１３１・１８０
Good Job!
Happy that you did it!
2011 JLPT study thread - Rina - 2012-03-05
jettyke Wrote:ありがとうございました！Ａ／Ａできました。CarolinaCG Wrote:Ｎ２、合格しました！おめでとう！！！
言語知識：５４・６０
読解：３８・６０
聴解：３９・６０
総合得点：１３１・１８０
Good Job!
Happy that you did it!
2011 JLPT study thread - jettyke - 2012-03-06
CarolinaCG Wrote:Super!jettyke Wrote:ありがとうございました！Ａ／Ａできました。CarolinaCG Wrote:Ｎ２、合格しました！おめでとう！！！
言語知識：５４・６０
読解：３８・６０
聴解：３９・６０
総合得点：１３１・１８０
Good Job!
Happy that you did it!