Back when Howard Dean called the GOP a “white, Christian party” and said “I hate Republicans,” I created an animated graphic of some of my favorite non-white or non-Christian Republicans and conservatives.

At right, I’ve edited the graphic for use in reference to the latest Washington Post story. The article reports on a study in which a few social psychologists determine that conservatives and President Bush’s supporters “had stronger self-admitted and implicit biases against blacks than liberals did.”

Part of the study’s claim is that “partisans” disregard unwelcome facts and then reward themselves for this “wrong-headed behavior.”

Therefore, if conservatives treat the study with serious and careful consideration, we thereby disprove it.

***

First, let’s take a look at the methodology.

For their study, Nosek, Banaji and social psychologist Erik Thompson culled self-acknowledged views about blacks from nearly 130,000 whites, who volunteered online to participate in a widely used test of racial bias that measures the speed of people’s associations between black or white faces and positive or negative words. The researchers examined correlations between explicit and implicit attitudes and voting behavior in all 435 congressional districts.

Assuming this description of the study is accurate, I have four questions:

1. The researchers offered participants pre-selected words, which they subjectively deemed “positive” and “negative.” Who’s to say the respondants attached the same positive and negative meanings to those words?

2. The researchers selected the faces and categorized them by skin color. Doesn’t that mean the study is inherently discriminatory? Also, are the faces equal in every other way: brightness, contrast, facial expression, age, attire, camera angle, etc? If not, could it be that perhaps participants reacted to “positive” and “negative” attributes of these other differences?

3. The responses are examined by congressional district, but it is not clear if the findings are further broken down by ideology within each district. Perhaps these “self-acknowledged views” are better tied to urban/suburban differences than to political alliances. For instance, are inner city liberals and conservatives more alike in their thinking than would be two partisans of whichever stripe living in different districts?

4. Are 130,000 whites taking an Internet survey representative of America? That number represents 0.04% of the total population, which seems like a pretty insignificant sample, given that it rounds to basically 0%; on the other hand, a sample that large means the margin of error is actually quite low. How reliable is this data?

***

The RNC spokesman quoted in the story lends credence to the study’s hear-and-reject theory by immediately questioning the researchers’ own political biases rather than reflecting on their results. Still, I think it is important to note the biases of any resercher studying a topic of this nature; its the same reason why one views candidate polling a little differently than research done by Gallup, or other third-party sources. Biases tend to influence behavior and outcomes, as this very study suggests.

Therefore, it is valid to bring up the political gifts: Brian Nosek donated $500 to John Kerry and Mahzarin Banaji donated $250 to Howard Dean and another $250 to the abortion activist group Emily’s List.

When faced with this information, Nosek reacts with a more subtle version of hear-and-reject:

Nosek said that though the risk of bias among researchers was “a reasonable question,” the study provided empirical results that could — and would — be tested by other groups: “All we did was compare questions that people could answer any way they wanted,” Nosek said, as he explained why he felt personal views could not have influenced the outcome. “We had no direct contact with participants.”

The very focus of the study, though, contains inherent biases that would be repeated even if the follow-up research is undertaken by social scientists who aren’t biased in favor of Democrats.

Why, for instance, does it focus only on whites, or only on racial attitudes? Why isn’t the study interested in the racial attitude of black partisans, and why doesn’t the study look at the way religious biases affect political decisions?

When you realize that the study is biased, and the researchers are biased, it’s a little easier to imagine that perhaps the study’s results aren’t 100% certain.

***

Based on only what I know now, I wouldn’t immediately reject the findings of this study. It is of course true that Republicans have work to do in order to win over black Americans. In fact, Ken Mehlman has made that his primary goal as chairman of the RNC: link, link and link. But at the time there seem to be more opportunities opening up for black politicians on the conservative side of the aisle. Black Republicans are currently running for Governor in Pennsylvania (Lynn Swann) and Maryland (Michael Steele)*. If Condoleezza Rice or Colin Powell ever decide to run for office, they would both enjoy a groundswell of support. And if we’re lucky, we may witness the swearing in of another black Supreme Court Justice before the President’s second term expires.

There may well be some lingering racism in the hearts of some conservatives, just as there may be similar biases in the hearts of liberals, but I’m not sure this study sheds much light on the subject.

UPDATE: * - and Ohio (Ken Blackwell).

UPDATE II: The story was so good, they printed it twice. Newsbusters finds that this same research was covered by the Washington Post one year ago this week. And here are some other problems with the study.