Suppose you start to investigate something, like an investment. The phrase imagines it as similar to walking into a building. You usually know _something_ going in. You enter into a learning process and learn more.
Maybe investor X knows very little going in: "Tesla makes electric cars."
Maybe investor Y knows a lot going in: "Tesla's stock has the symbol TSLA. It is run by Elon Musk. It is the biggest electric automaker in the world. The stock has grown a lot. If you had put $1,000 into Tesla stock in 2010, it would be worth $200,000 today."
The study showed that all investors tend to exaggerate how much they know. And that, oddly enough, knowing more doesn't change that tendency. Investors X and Y may be equally overconfident, and equally foolish.