Not by all mechanisms, at least. I practiced with IATs a few years ago and figured out how to mostly get whichever scores I wanted by priming myself. For example, I normally test as moderately prejudiced against African-Americans, but if I fix in my mind at the outset a full emotional picture of what I dislike about myself (I’m white), and tie this into a negative stereotype of white culture, and then think about a couple of particular African-Americans with whom I have positive associations, and tie this into a positive stereotype, and if I kind of keep these associations in mind through the test… I test as prejudiced in favor of African-Americans.
Isn’t the priming effect something inherently useful in the IAT? I could see advertisers wanting to test the effectiveness of their ads based on this. You could also test to see which factors most greatly influenced a change in bias based on how the subjects were primed couldn’t you? I mean I have at least one study I would want to do where priming vs. controls would be involved.
Wait, if you do that repeatedly, would you then eventually still find that you have to explicitly prime yourself?
Or what if without deliberately priming? I mean, I’d expect that to actually become faster, one has to build up the right associations in the first place, right? As one optimizes oneself toward this task...
Well, it would be an interesting bit of research to try, imho.
Good point; I wasn’t thinking. Keeping such emotional associations in one’s mind might plausibly cause the associations to stick over time, and the IAT might give you a yardstick to notice when you’d got the associations evened out.
Also, if there were a particular context where you particularly wanted to avoid a given prejudice (e.g., while you were interviewing job candidates), you might be able to first calibrate with an IAT to know how much priming evened out the associations, and then keep a similar set of primed concepts in your mind while you did the task for which lack of prejudice was important.
Yeah, though actually I was thinking that specifically training oneself toward doing “well” on the IAT would more or less retrain the brain to associate various things, since when taking the IAT, well, you have to make (or break) certain patters, so training oneself to do faster on the IAT (I’d suspect) may shift the various patterns.
ie, over and over training your brain to do that task quickly, well, the actual way your brain might respond to speed it up would be to shift the associations to, well, speed it up. Maybe. (at least, given my profoundly limited knowledge of cogsci, I don’t see any obvious implausibility here, but...)
As I said, I’d like to see some experimentation along those lines...
Actually, also I wonder if the IAT can be generalized to the more abstract type of bias talked about here and on OB. ie, can a variation be set up that sets up/activates the various cog biases, and to do well at it one would have to break those patterns somehow?
Actually, you know… “does it work in reverse?” sounds like the type of thing that others may have already thought of. Anyone here know if there’s already been some investigation into that?
Not by all mechanisms, at least. I practiced with IATs a few years ago and figured out how to mostly get whichever scores I wanted by priming myself. For example, I normally test as moderately prejudiced against African-Americans, but if I fix in my mind at the outset a full emotional picture of what I dislike about myself (I’m white), and tie this into a negative stereotype of white culture, and then think about a couple of particular African-Americans with whom I have positive associations, and tie this into a positive stereotype, and if I kind of keep these associations in mind through the test… I test as prejudiced in favor of African-Americans.
But the effects of priming wear off quickly.
Isn’t the priming effect something inherently useful in the IAT? I could see advertisers wanting to test the effectiveness of their ads based on this. You could also test to see which factors most greatly influenced a change in bias based on how the subjects were primed couldn’t you? I mean I have at least one study I would want to do where priming vs. controls would be involved.
Hrm, oh well...
Wait, if you do that repeatedly, would you then eventually still find that you have to explicitly prime yourself?
Or what if without deliberately priming? I mean, I’d expect that to actually become faster, one has to build up the right associations in the first place, right? As one optimizes oneself toward this task...
Well, it would be an interesting bit of research to try, imho.
Good point; I wasn’t thinking. Keeping such emotional associations in one’s mind might plausibly cause the associations to stick over time, and the IAT might give you a yardstick to notice when you’d got the associations evened out.
Also, if there were a particular context where you particularly wanted to avoid a given prejudice (e.g., while you were interviewing job candidates), you might be able to first calibrate with an IAT to know how much priming evened out the associations, and then keep a similar set of primed concepts in your mind while you did the task for which lack of prejudice was important.
Yeah, though actually I was thinking that specifically training oneself toward doing “well” on the IAT would more or less retrain the brain to associate various things, since when taking the IAT, well, you have to make (or break) certain patters, so training oneself to do faster on the IAT (I’d suspect) may shift the various patterns.
ie, over and over training your brain to do that task quickly, well, the actual way your brain might respond to speed it up would be to shift the associations to, well, speed it up. Maybe. (at least, given my profoundly limited knowledge of cogsci, I don’t see any obvious implausibility here, but...)
As I said, I’d like to see some experimentation along those lines...
Actually, also I wonder if the IAT can be generalized to the more abstract type of bias talked about here and on OB. ie, can a variation be set up that sets up/activates the various cog biases, and to do well at it one would have to break those patterns somehow?
Actually, you know… “does it work in reverse?” sounds like the type of thing that others may have already thought of. Anyone here know if there’s already been some investigation into that?