To study vocab I create study sessions based around a theme. Usually that theme is something like “sentences which contain kanji X.”

I built a sentence deck from materials on smart.fm, and I use the Anki browser to find matching sentences and create a cram session. I study each sentence and keep notes of all the distinct words I find that use the kanji I am focusing on, and how the words are read. The cards have the sentence written on the front, and the audio and translation on the back.

Because the study session has a theme, I feel it is easier to make connections between words, recognize usage patterns, and get a sense for which meanings and readings are common and which are obscure. Dictionaries give you lots of information on readings and meanings of kanji, but they don’t tell you what is common and what isn’t, so you don’t know what you need to focus on now, and what you can ignore until later.

Before doing this, it is a good idea to get an idea of which kanji are common in your deck and which are not. To put together a good cram session, you need 50-100 sentences, so you can’t cram based on kanji that only show up once or twice, or kanji that show up too often. For example, today I studied 切, which is in 85 sentences, and I found 23 distinct words. I wrote about how to compute kanji frequencies using Python in another post. If you aren’t a hacker, you might try just using general information about kanji frequency to give you some good ideas of kanji to focus on.

The idea is to find kanji that are common enough to be useful, but not so common that you have to slog through hundreds of sentences with dozens of repetitions of identical usage. You will learn those while working through your cram sessions for other kanji.