Monday, July 2, 2018

Interesting to me: people are bad at this sort of conditional probability.

Interesting to me: people are bad at this sort of conditional probability.

That is: to know the percent of Democrats who are white is hard. You need to either have good sampling of Democrats, or you need to solve a fairly complicated distribution, as you need to think about the case rate and how race affects party affiliation. That's practically baysean.

Instead, ask: what percent of whites are Democrats?

That is a tremendously easier question.
https://fivethirtyeight.com/features/democrats-are-wrong-about-republicans-republicans-are-wrong-about-democrats/

9 comments:

  1. The take-away, here, is that each party overestimates the representation, in the other party, of whatever qualities they despise.

    That Democrats are so far wrong about Republicans, though, amazes me. After a year or more of disassembling the election results, why would any Democrat not understand that Republicans are mostly people of modest means?

    ReplyDelete
  2. This is interesting. And makes sense in a certain way — if you don’t like black women in positions of power, or atheists, and you think the Democratic Party is largely that, it’s easier to drum up war against the Democratic Party.

    If you think the Republicans are mostly “evangelical Christians” and can recognize how toxic that theology is when living in a diverse city, it’s easier to drum up war on Republicans.

    In both cases, it turns into opposing “group disliked for deeply entrenched reasons” and “their enablers.”

    ReplyDelete
  3. John Jainschigg

    This is speculation, but on the left I get to see and hear about a bunch of big rollers banking the Republicans and dictating policy. It’s easy to imagine there’s lots of Mercer’s, Kochs, and the like.

    Also: income and net worth are two different ways of measuring wealth.

    Oh the left, I’ve heard the debunking of the “economic anxiety” concept — that trump voters weren’t poorer than democrats, just whiter. I’d like to know what percentage of democrats earn over 250k/annum, and what republicans and democrats thought it was.

    ReplyDelete
  4. the real interesting thing about the queerness numbers in particular is that it shows both parties overestimate the numbers of those people /in the country/- that is to say, there's just not enough people who identify as queer to be more than 6% of /any/ political party. Which is probably why the question you ask above is more generally used, on the whole, though the one asked here is extremely useful to highlight just how much minorities are, well- minorities.

    ReplyDelete
  5. Again: my assessment differs from 538. I think this type of assessment is difficult.

    It'd be interesting to ask the opposite question: Republicans, tell us what percent of LGBT folks are also Democrats?

    This style of question is found in done Phil 101 courses indicating how bad people are at this sort of estimation.

    ReplyDelete
  6. That is: I'd wager this relies on the conjunction fallacy.

    ReplyDelete
  7. Try that a different way, John Jainschigg

    ReplyDelete
  8. John Jainschigg - Trump supporters are not poor. We may overestimate the percentage of Republicans who are rich, but they're richer by a lot than Democrats.

    But also the stats in this are pretty stupid stats.
    theatlantic.com - People Voted for Trump Because They Were Anxious, Not Poor

    ReplyDelete
  9. Robert Bohl Not the thread for it. Let's wind this down.

    ReplyDelete