(2016-06-01, 2:28 pm)pm215 Wrote: I kind of feel like for that kind of thing reading isn't a very efficient way to attack it, because it's quite easy to just glaze over the detail (for instance for 間 vs 間に you can generally read quite happily without knowing the distinction, because the meaning in any particular sentence will be obvious from context), but JLPT questions love to check if you know the distinctions and which verb form goes with what, and so on.
It's certainly not efficient in many ways - you have to read the same expression dozens, maybe hundreds of times before you develop a sense of what 'sounds right'. On the other hand, every page you read is effectively reviewing dozens of grammar and vocabulary points.... from the recognition side only, but still. It gets the job done eventually, and you *do* have to read a *lot* to pass the JLPTN2 and N1 (not so much the N3 I think, although I never took it), because if your reading speed isn't up to par you won't be able to finish the test. It doesn't help to score very highly on only the first 60% of the test and get a zero* on the other 40%.
Still, I agree that you'll improve faster if you make a point of figuring out the grammar points. I don't know about 'drilling' them in the sense of Anki, but still. I realize having a lot of missed questions can seem overwhelming, but you can just ask …here or on another learning site... or just search the web. Just take the points one at a time.
The first result on googling "difference between 間 and 間に" has a pretty good answer, for instance.
http://japanese.stackexchange.com/questi...3%E3%81%AB
*(Pedantic note... Of course you shouldn't get a zero for any section of the JLPT... if you run out of time, fill in random bubbles. Or just a single column of bubbles, or whatever bubble pattern you like... you'll average 25% instead of 0% on questions that you 'didn't answer' that way.)