Cloze, reading, learning, life

While working on a little research paper a couple weeks ago, I came across cloze procedure. A cloze test is used to measure the difficulty of a text. In a cloze test, you take a text and replace every fifth word with a blank space. The reader, who has never seen the passage before, reads it and fills in the blanks. It’s kind of like mad libs, but the goal is to choose the correct words instead of just having fun with it.
What’s cool about cloze tests is what they can tell you about learning. By comparing how well readers complete the passage vs. how well they answer questions given a complete text, you can find where the optimal difficulty is. It turns out that there is an optimal difficulty level if you’re looking to maximize information gain. Right around a 35-40% cloze success rate is best if you’ve got an instructor available when needed, and around 50-60% if you’re learning independently.

You tend to acquire the most information with texts at those particular difficulty levels. You bring enough context and prior knowledge, but just enough to get a handle on the new stuff. What’s crazy, if I can stretch it a bit, is that the most efficient learning takes place when you’re stumbling roughly 40-60% of the time.

So it kind of woke me up to thinking, if the goal is to learn and grow, how can I pick and choose the best experiences? I don’t mean it in a snobby sense—“that is below me”—but in the sense of growth and challenge—“this is difficult and worth it.” If you’ve got perfectionism issues (like I do sometimes), sometimes you get stuck doing things you’re great at, because you’re great and being great feels good. But there’s no growth there. So the cloze thing comes into play. Try something where you know you’ll only be partially successful. See what happens.