OkCupid's Reply by Race graph, labeled.

[darxus]

This is correct IF the labeling should match the graph before it in OkCupid's blog, on which it is based.



I based these percentage values on the colors of two of the other charts, which matched each other.

54- means 0-55.
64+ means 64-100 (I'm curious why they clipped these in the color values).
A value of, for example, 55.5 means that I didn't have enough data to tell if it was 55 or 56. Fortunately this only occurred in ranges of 1.

Update: http://blog.okcupid.com/index.php/2009/10/05/your-race-affects-whether-people-write-you-back/
My column and row labels were correct, but my percentages were way too high (although probably proportional).

Comments

[kragar00.livejournal.com]

So in general black females are the most likely to respond to anyone? And are more likely to respond to white males than black males? That seems... weird...

Also, the graph seems heavily biased against black males... They have the worst response rate of any race...
I don't know how much stock I can put in that... But maybe it's a demographic distortion from Cupid's membership... I'm not sure what Cupid's membership looks like...

[darxus.livejournal.com]

Yeah.

Black women are more likely to respond to men of all races except for Native Americans responding to Others.

I'm real curious about the composition of Others.

Everyone is least likely to respond to blacks, including blacks, except Pacific Islanders.

Asians, Blacks, Indians, and Middle-Easterners are all least likely to respond to their own race.

[feng-huang.livejournal.com]

I don't think this is about response rate, just about match percentage.

[darxus.livejournal.com]

Nope.

The title "Message Reply Rates By Race: male sender" is copied from the chart that the data was taken from.

The chart before it on the okcupid blog showed match percentage by race. The variance was pretty statistically insignificant.

[feng-huang.livejournal.com]

Ah. I stand corrected.

Interesting.