Cool graphs of geographic variation in attitudes, based on non-random samples but still interesting

Ben Hyde pointed me to this data-based dating site. I have no comments on how it works for dates, but they have a lot of fun maps, for example this:

Are some human lives worth more than others?


Scale
268,864 people have answered

And this:

If you knew for sure you would not get caught,
would you commit murder for any reason?


Scale
359,761 people have answered

This is great; I can’t resist giving a couple more:

Rate Your Self-Confidence


Scale
581,443 people have answered

OK, OK, just one more . . .

Would you date someone just for the sex?


Just For the Sex

Scale

data set: 448,000 people answered


Where did the data come from?

Assuming these guys are for real, their data come from online surveys of participants in their free dating service: the idea seems to be that you answer a bunch of questions and then they use this in finding a match for you. Which suggests they have lots more background data on age, not just location. I don’t quite see why they have zero responses in Alaska and Hawaii: if this is an online site, I’d think they’d get responses from everywhere, no?

All the warnings about sampling apply, sure, but, hey! This is pretty cool stuff.

You can go to the site to see their discussions of the above maps. I’m sure they’ll make more fun maps for you, if you just ask. . . .

P.S.

The other reason I posted these maps is to get all the commenters here fired up about their red-green color scheme. I thought this would take the heat off me. Seriously, though, setting the color choices aside, I like how they show a continuous color bar at the bottom to map the colors to the responses. I think I’ll steal that for my graphs.

11 thoughts on “Cool graphs of geographic variation in attitudes, based on non-random samples but still interesting

  1. Setting aside the problem of people who can't tell red from green, I think you've got the colours backwards. Red should be "yes".

    The reason is that red stands out, and what you want to stand out is the positive response, since otherwise you have to think in double negatives.

    Analogies with traffic lights are not a good guide to this. Not naively, at least. Non-naively, they are, since it's as I said above – red stands out, so you want it to mean the thing that you're trying most to draw attention to. For traffic lights, that's "STOP". For survey responses, that's "YES".

  2. Radford:

    I think at this point we can all agree that red-green is not a good choice. Just to be clear, though, I didn't make these maps; I'm just linking to them.

  3. Given graph 4, we can say with a relatively high degree of confidence that people in many states lie, either to themselves or others. Or they don't define sex with someone in a not serious relationship as dating.

  4. Their recent analyses examine racism on the site. Also some interesting stuff on how particular phrases in messages affect the probability of reply. (Turns out that "physics" and "zombie" are good words to use… but I imagine they have to control for a few extra variables…)

Comments are closed.