Took N4 today, after passing N5 in December.
I wonder what i got. I really wish they would release the test quetions properly after exams so you could learn frkm them.
Havent come up with a plan yet but its just so nice to have some motivation to learn again.
N2: Anyone else had the feeling that this test was about twice as hard as any previous test?
I took N2 3 times now and this is how I felt. Or maybe I’m more rusty than I imagined.
Or maybe the test difficulty just varies quite a bit.
Also, the audio was echo-y as always, maybe slightly more this time.
(first time I just took it for fun to see where I’m at, second time I had improved a lot and actually almost passed it. This time I feel like I’ll get a 25% score maybe. Both in text and listening, they were using lots of infrequent words I never encountered in the past few months in reading.
But I’m not upset about it, I’m just taking the test for fun and motivation, if I only pass it in a few years that’s fine too.)
I didn’t take the test (hoping to take N2 in December) but I have seen a ton of people saying N2 seemed particularly difficult this July so you’re not alone, it seems to be the most complained-about level this time. Apparently the reading section was especially tough for a lot of people. Hopefully your results are better than you expect anyway!
If this year’s test really was harder, the scoring system is supposed to account for this: it will bump up the points you get for questions when they’re harder (as measured by fewer people getting them right).
The proctor at my mock exam also said that if too few people get an answer right, they will remove it entirely from the scoring too.
Wouldn’t it make more sense to actually give LESS points for harder questions though (and more for easier ones)? I mean, if they wanted to keep the passing rate similar to some easier years that would balance things out, wouldn’t it?
Took N1 this weekend- I definitely guessed a lot, but there were very few pure guesses where I didn’t understand the question or answers. I feel about as confident as I did after N2- that is to say, not very confident at all, but I passed N2 by a decent margin last year, so I feel pretty decent. I’ll be very frustrated if I failed and have to keep studying until December, though.
I felt pretty good on listening and reading. If I did fail, it’s almost certainly because I bombed the vocab and grammar.
So, I took the N3 this July in São Paulo, I think for the first time I left the test with the feeling that I passed, when I took the N5 and the N4 I left the exam room feeling like a maybe at best, so it was a nice feeling, even though I didn’t read as much as I wanted, I still increased my reading by a lot (Thanks Satori Reader) and also did a lot of listening practice with those japanese romantic dramas on Netflix haha, they are very good for listening practice.
One other thing that helped with being confident on the exam day was doing a mock exam one day earlier from the actual exam. It was good not only because I got a good score and because of that I felt more confident throughout the test, but also because it showed me the weakest spots that I needed to pay attention to during the actual test. For example, I spent a lotttt of time during the reading section on the mock exam, and I realized that I was running out of time, so on the actual exam day I made an effort to do some extra focus and to not let my mind wander off during the reading section.
I took the N3 last Sunday. It was better than my first N3 attempt in 2023, but it was definitely far from easy and I was still stumped on a lot of items. It didn’t help that I wasn’t feeling very well in the morning.
Kanji was tough, I know I am weak at kanji but at least now I can make educated guesses. It still sucks to not being able to find the definitive answer and know you are right. I checked some items afterwards and I got some simple kanji wrong lol.
I ran out of time for the reading section and was just answering randomly for a few questions because I don’t have any time to read the last two passages. The listening section feels a bit harder than the first but it’s been two years so I might be inaccurate it might just be the nostalgia talking. All in all it was pretty.. neutral I guess. Hopefully I can pass, but it’s still a long time until the scores will be out. At least now I know how much I need to grind still!
That’s a good point, but I don’t think that necessarily works to my advantage, at least for me personally. If I constantly encounter rare words and difficult grammar (“harder question”), then it takes more time for me to read the section, and I don’t know beforehand whether this section is supposed to be very hard or not. If it is, I could skip it or skim to take less time reading it, but I don’t know that beforehand.
And I’m already strapped for time, this is my problem, I’m running out of time in N2. I think this time I ran out of time with about 10 questions remaining. My reading speed just isn’t fast enough. Gotta read 100 more books (or tests) probably.
And yes, I think the system works by giving LESS points to harder questions and more points to easier ones, which also makes sense from a multiple choice design perspective, because if you’re just random guessing, it really shows when you get super easy questions wrong, so you’ll lose a lot of points for that. (as elbatrofmoc said:)
But as I said, I don’t really mind that much, I’m just doing this for fun, and I’m clearly not ready for N2 yet. I’ll do some N2 test exams in the future and try to see if this N2 exam really was much harder than average. I’ll never give up 
I attempted the N3 in London for the second time. It felt much harder than last year and I found the reading section the hardest, especially with the time pressure. So I really need to develop my reading and comprehension speed for next time.
I took N2 in Leicester for the first time. When I signed up I definitely didn’t expect to pass, just wanted to check in on what I’m missing and set a motivational deadline for myself.
Doing mock tests I started thinking I have a shot at (barely) passing with enough time to go through everything. The reality was much more brutal - I started with the longest reading tasks and they took me so much time I had to miss out on a bunch of mid-test questions. Happy with the progress I made in listening. It used to be my weakest side and now feels the strongest.
I definitely need to read a lot and improve my vocabulary which should in turn improve the reading speed as well. Glad to have had this experience and excited for the December one 
The JLPT uses a complicated system underpinned by an academic theory called Item Response Theory. The essential idea as I understand it is that the maths of how people answer a particular question gives you a function that shows how hard the question is and how good that question is at identifying better vs worse candidates. Then you can score the candidates by looking at how they did on all these questions. So it looks both at whether questions were easy or hard, and also at how good they are at distinguishing good candidates from bad ones. If getting a difficult question right means you’re very likely to be in say the top 20% of candidates then the algorithm is going to weight that quite heavily in how you score. I think the converse may also be true: if you get a question wrong that only the bottom 20% of candidates tend to get wrong, that also is heavily weighted in your getting a lower score. (Questions that are so easy everyone gets them right or so hard everyone gets them wrong are equally useless for identifying good candidates and provide no information.) The models do incorporate handling of “answer was guessed”, incidentally.
I’m sure the JLPT website used to have a document explaining this, though I can’t find it now. In any case the maths of it has always been rather opaque, unfortunately, and I don’t claim to have a solid understanding of the detail. But at any rate it isn’t quite as simple as “a certain number of points are assigned to each question”.
Yeah, I’ve heard about Item Response Theory and its usage in the JLPT before.
That’s not entirely true though as I said, it filters out people who guess randomly, pretty efficiently. And of course there are probably no or few questions so trivial that everyone gets them correct in the test (e.g. people who are actually N5 or N4 but taking N2. these will be filtered out), at least if you don’t take time pressure / reading speed into account, which is also being tested.
Yeah, it’s not like there aren’t copies of all the questions up before even the end of the test. I don’t know why they bother to try to keep them secret.
It’s too bad that the japanese government doesn’t invest more into the japanese language industry. The jlpt could have a speaking interview and even writing. Create a whole industry for japanese teaching and make more fluent speakers.
