Presidential Candidate Preferences

[Icey]

I ran into a neat application that my political science professor encouraged the class to look at.

https://implicit.harvard.edu/implicit/demo/featuredtask.html

Basically some words and images will appear on screen and you need to press "i" if that word or image is a "good" word like friend or pleasant or if it is the candidate you are supposed to identify, and you press "e" if it is a bad word like "angry" or "enemy" or if it is not that candidate. So for the first series you press i if it is Barack Obama or a good word, the second series Mike Huckabee or a good word, etc.

At the end of it (it takes like ten minutes) you get your results. On a scale from "more positive" to "more negative" it shows you your views of the candidates, along with their relative displacements. I have seen these things used in polisci research before (I am a polisci major) and while I do doubt the credibility of this stuff usually, the results I got were incredibly accurate.

I thought some of you might want to give it a try and say whether or not the results you got followed your true preferences. You could even post the image of your scale. Plus, you're taking part in a research study at the same time :p

[phattonez]

I had Huckabee all the way up, and McCain all the way down. I don't agree with that in terms of politics, but it does seem that Huckabee is a lot more likable.

[rock_nog]

Well, being ever the fan of psychology, I took your little test. I most preferred Obama, with a dead tie between Hillary and Huckabee, and Mccain was in last. Actually, I was a little surprised - I expected Hillary to be lower and Mccain to be higher.

My $.02 though... It seems to me that potentially, you could get inconsistent results with a presidential candidate IAT. I know I certainly had the feeling, in simply trying to rate them, that I felt ambivalent about most of the candidates. I mean, how do I feel about Huckabee vs. Mccain? Mccain vs. Clinton? Clinton vs. Huckabee? I honestly have no strong feelings, because my feelings about each of the candidates vary from issue to issue.

One final note - IATs are often used to measure implicit racial attitudes. I wonder if that might have any effect on the results? Then again, maybe not - it's certainly true that having a specific person, rather than generic representations, can cancel out the effects of implicit racial attitudes. Eh, just a thought.

[gdorf]

According the the test, I feel equally about each candidate. All of them fell on the same line.

[Icey]

Well, being ever the fan of psychology, I took your little test. I most preferred Obama, with a dead tie between Hillary and Huckabee, and Mccain was in last. Actually, I was a little surprised - I expected Hillary to be lower and Mccain to be higher.

My $.02 though... It seems to me that potentially, you could get inconsistent results with a presidential candidate IAT. I know I certainly had the feeling, in simply trying to rate them, that I felt ambivalent about most of the candidates. I mean, how do I feel about Huckabee vs. Mccain? Mccain vs. Clinton? Clinton vs. Huckabee? I honestly have no strong feelings, because my feelings about each of the candidates vary from issue to issue.

One final note - IATs are often used to measure implicit racial attitudes. I wonder if that might have any effect on the results? Then again, maybe not - it's certainly true that having a specific person, rather than generic representations, can cancel out the effects of implicit racial attitudes. Eh, just a thought.


Yeah I have to agree with you. I think a lot of factors come into play in this kind of thing incuding possible racial biases or other measures that I wouldn't even think of. And like you said, with a presidential candidate, you don't always have a clear scale - if you have a more nuanced view, like you agree with McCain on one issue but not on another, that's bound to complicate it. Any 2D representation is bound to fail to catch that.

At question also is whether this is a valid test at all, or if it is just completely meaningless. I did find it cool though that it put my preferences almost exactly where I thought they should be (Obama very close to the top, Hillary slightly above the middle, McCain below her at the middle, and Huckabee directly below him slightly below the middle).

[rock_nog]

Oh, don't even get me started - you could probably write entire books on IATs. I mean, the most common ones are the ones designed to measure implicit racial attitudes, but what that even means, nobody's quite sure. They certainly seem to be measuring something, in that generally speaking, the results are consistent, but consistent with what, nobody's quite sure. No one can really define what implicit racial attitudes are. And if they don't translate to explicitly racist actions, do they mean anything?

[Pineconn]

It placed McCain at top, Huckabee and Obama in the middle, and Clinton at the bottom. I can't agree with it more.

[The Desperado]

Obama, Clinton, McCain, Huckabee. Exactly like I thought it would be.

[Russ]

I had Huckabee on top, McCain in the middle, and Hilary and Obama at the end. Just how I thought.