3 ms·
This particular instance of correlation and causation has always been fascinating to me. Is it possible to have a non causal correlation that is impossible to
by geebee 8y ago
This particular instance of correlation and causation has always been fascinating to me.
Is it possible to have a non causal correlation that is impossible to bring about without increasing (causing) the effect.
For instance suppose you tried an experiment of getting 500 books into a household with a new infant and zero books. Let’s say you did this with no special attempt to produce the cause, but made no attempt to ensure the process of getting 500 books over there had no influence either.
How would you go about this? Ask for donations? Collect books? How would you find donors, how would you find families to donate to?
My guess, which I make purely on a hunch, is that you would have to make dramatic (perhaps unconscionable) efforts to ensure that this process would not bring about the effect of higher literacy on average.
Not sure by how much but I think (just a thesis) that it would be meaningfully measurable
- bluGill 8y agoBooks are cheap. I'd get 500 books from a used bookstore and put them in a hut in some primitive village. Note that books would not be in the native language, and nobody in the village can read anyway. My hypothesis is that books will make no difference. An interesting and useful experiment would be to place books in the homes of poor people who can read but don't. Turns out cities all over the US are doing this experiment on a large scale right now (often paid for by United Way). While the experiment isn't completely controlled anyone who cares to can collect data and analyse it.
- geebee 8y agoSounds like they’re running the experiment. The first scenario you described is a good illustration of the kind of effort it would take to ensure no effect. The second? I’d be interested in knowing how they find families and decide what books to include. I’d expect (and I admit I hope) it will have a measurable efect. I’ll search for it but please post a link if you have one.
- bluGill 8y agohttps://1000booksbeforekindergarten.org/ https://1000booksbeforekindergarten.org/ They claim studies support the idea, but I'm not able to figure out which