Characters — in Chinese or Japanese (or Korean Hanja) — can be learnt in various ways, and it is plausible that paying attention to radicals helps memorization and retention. But is there any hard evidence for this?
Are there any studies on the effectiveness of character learning or teaching methods that pay attention to semantic radicals? (By "semantic radicals", I mean those that carry meaning, as opposed to phonetic parts or even single strokes.)
PS: Wikipedia has a few lists and articles that are related to radicals:
- Radical (Chinese characters): see also the corresponding entry in the Japanese version of Wikipedia: 部首 (ぶしゅ).
- List of kanji radicals by frequency.
- On Japanese SE, a question about learning radicals was closed in October 2014. The only available answer there (as of August 2017) is based on experience rather than clear evidence.
I previously asked a similar question about Chinese: Evidence that learning radicals speeds up character learning in Chinese? Comments and answers went on to focus exclusively on Chinese, and the situation for Japanese may be different for the following reasons:
- The Japanese strongly reduced the number of kanji after World War II. (See Tōyō kanji on Wikipedia. This list of Kanji from 1946 was replaced by the Jōyō kanji in 1981, which was revised again in 2010.)
- Kanji have more different readings than hanzi (see e.g. Why do kanji have several different pronunciations?).
- Kanji have on average a higher number of different meanings than hanzi.