The reason I don't like that study as much is because it's about people's names. But people's names do in fact reflect the level of education of their families, and in an extreme case they can choose a new name if they want. The poster we're reading about, for example, is called "Richard Smith". (Whether by birth or a name that he adopted doesn't really matter.) I like names. If on a dating site I were given, sight unseen, the chance to view an Octavia or a Hermine I would choose it over any Susan or Latisha. Because it seems to me much more likely that Hermine would be educated and interesting.
Does this make Hermine a racial name?
Why couldn't Hermine (or Octavia) be black? In fact they could.
So I would say that the effect of a person's name, and a picture showing them as black, are, as this blog post already shows, somewhat orthogonal. Richard Smith isn't a "black name", but this poster had a black experience.
It's also quite important to point out that parents have almost complete choice over what to name their children. As some linguist pointed out, it's one of the few times that people get to name something (someone) in the world!
So I don't like it as a signal. A picture is much more pure. And also not something anyone can change.
Finally, in my case I'm specifically interested in programmers and tech. So my small "study" would be very targeted.
I'm still very interested if anyone has any comments about the methodology. I've never done any social science research before (well not formally anyway), and I'd be very curious if I'd be making some mistake in the methodology that I could avoid.