At least some of this paper is relevant to learning a second language though.
I’m still in the introduction, but
Furthermore, several studies have shown learning of word
meanings from written context to be a relatively ineffective
process (e.g., Gibbons, 1940; Sachs, 1943), especially when
compared with intensive direct instruction (Jenkins, Pany &
Schreck, 1978; Johnson, Toms-Bronowski & Pittelman, in press;
Margosein, Pascarella & Pflaum, 1982).
makes a lot of sense if I compare words I’ve learned through verbal instruction/explanation first. For example, 飲み放題 (all you can drink) was thoroughly explained to me after I moved to Japan due to standards of work socialization. When I first saw the word put together on WaniKani, I instantly knew what it was. I’ve had near 100% accuracy (aside from misreading it as 食べ放題). I’ve never purely learned a word (ie from kanji to vocab) on WaniKani and been able to have that level accuracy. I don’t even learn well audibly though and typically learn facts far better through text. I’ve had the same experience for a couple of other words verbally taught to me in situational context (not directly in the context of a sentence typically).
I’m still working through the rest of the research paper, but taking a break.