Thread by @lymanstoneky

I agree with this, but it's worth noting one reason quant people and qual people don't get along is a LOT of serious qual people FIRMLY REJECT the idea that qual work is most useful in a supporting/hypothesis generating role for quant work to test.

https://twitter.com/abhishekn/status/1659287646424952832

I find that there's actually tons of openness among quant people to the argument that WE undervalue qualitative work, and it is super useful for elaborating mechanisms, generating hypotheses, identifying blind spots, identifying "interesting topics to study," etc.

The people who usually chafe at this argument aren't quant people. It's the qual people who dislike this argument, because it suggests their entire work is basically a good Chapter 2 of a dissertation.

Imagine the reverse. You do sometimes hear qualitative people saying something like, "The best use of quantitative work is to establish some broad empirical contours and interpretative limits for the serious work of carefully describing meanings, mechanisms, and rhetoric."

Actually you can find something like that take on page 112 of the UNFPA's 2023 SWP report, for example, and I have to say, I think it's a really, really, really bad take. Because of course I think that; I'm a quant nerd. unfpa.org/sites/default/…

Upshot is, that even if we all agree that Mixed Methods Are Good, there remains a considerable disagreement about what the actual roles are. Which method is the hypothesis generator and which the tester?

Now, I have to say, there are some very good pieces of comparative and historical sociology that use this "quant to generate qual to test" framework that I do really like and admire! Here's one example I think is really excellent work: press.uchicago.edu/ucp/books/book….

But my point is that while there absolutely are researchers who are comfortable in a wide range of mix ratios in mixed methods, there are also a lot of people who have genuine philosophical commitments against mixing or at least certain kinds of mixing.

*In particular* there are meaningful cadres of researchers who make what I think are cogent arguments (if not entirely persuasive) in favor of the extreme corner solutions: that nearly-pure quant and/or nearly-pure qual in fact make distinctive and irreplaceable contributions.

My personal resolution to all this is that I am, on a certain level, basically just a nihilist. Research is stumbling mostly-blind in the dark room of the world, making occasionally-useful guesses about what the objects you can't really see might be, and most methods work-ish.

That we find anything useful, predictable, replicable; that anything at all rises to something we might call "knowledge" is genuinely miraculous. And I do mean that in the spiritual sense. That our hunter-gatherer brains handle abstraction so well and so often arrive at useful...

... theories about the world that seem to offer such a compelling account of real ground truth is, in my view, only credible if a rational mind gave us rationality.

That other animals with similar synaptic density (elephants) or neuron counts (cats) or that are vastly more competent at discrete tasks (judging jump distances like goats) are sooooo much worse at abstraction is genuinely pretty wild!

Like elephants actually have more synapses than we have and they DO have the ability to engage in some meaningful abstractions (they can group unfamiliar humans by ethnicity for example), their ability is extremely rudimentary compared to us.

do you like how "side comments on mixed methods" turned into "radical arguments about human cognition and the divine"? i bless you people with the best #content