I agree that the IATs are a questionable testing method (and address as much in my OP).
That being said, you make the assumption that subconscious reactions aren't tied to bias, attitudes and behaviors. My experience with cognitive behavioral therapy leads me to believe otherwise.
A person who smokes a cigarette every day when they go to work is trying to quit. As they walk out their front door to get into their car, the thought process that walking out the door triggers includes a laundry list of automated responses, from taking steps to the car, to open the door, to putting on a seat belt, to activating the knowledge schematic of operating a vehicle, to pulling out a cigarette and lighting it up. The brain, subconsciously, begins doing work to handle all of these tasks, including prepping nicotine receptors to receive nicotine through that first cigarette. Someone trying to quit will have these receptors fired up through pattern behavior, making them neurologically agitated and making it hard for them to quit. Smokers who participate in cognitive behavioral therapy sessions for smoking identify these triggers and can mentally take steps to remove the thoughts of smoking from their mental behavior list, making it easier to quit because their nicotine receptors aren't being primed (and essentially screaming) for nicotine by their daly behaviors.
Now. Is it too much of a leap that a police officer who sees a brown skinned face and is more likely to draw the association of "bad" or possibly even other adjectives, like "dangerous," would then go through their automated behavioral processes with different outcomes, like more intense scrutiny, higher suspicion (that might result in being more likely to make an arrest) or even using violence in self defense due to a perceived threat?
I'm not saying training to "pass" IATs would make someone less racist. Or even that people who "fail" IATs are racist. But I think it may be very dangerous to underestimate the effects associations have on subconscious thought and subsequent behaviors.
Not to mention the one week retention. People look at that and say "they only tested one week - that's nothing." Which is true, you won't change someone's life by affecting their responses for just one week. But for anyone who has been invovled with training in the workplace, they will tell you - getting people to retain knowledge for a DAY after training is considered a huge success. For an entire week to still see statistically significant effects of the training would be considered highly effective.
CBT is fundamentally different from the IAT. There's no comparison because CBT relies on a different mechanism deriving from a different theory about how cognition, behaviour and emotion interrelate.
Ever since Freud brought the term "subconscious" into being there has been a lot of pop psychology sorrounding the term which I think leads people astray as to its operations. That's not to say that much of cognitive processing happens consciously as we understand it because it does not. But what it means for it to happen subconsciously is complicated and the actual drivers behind different effects have to be mechanically explained before we can start giving weight to different explanatory theories.
CBT operates by effectively co-opting the fact that - contrary to older beliefs about brains - express thoughts exist in a feedback loop with neurobiology whereby your thoughts CAN influence your neurological architecture. This is an oversimplification and the method of CBT is somewhat coarse but the basic idea is that it allows you a way to chemically modify your neurological functions in a way that you would do with drugs (though not at the same magnitude) by taking advantage of the fact that your brain - for the purposes of this goal - is actually quite malleable and plastic.
It is a tremendous leap to go from the general notion that the brain is plastic to the idea that the IAT predicts "attitudes". It's a leap because mental associations of the kind being studied are actually more akin to fundamental premises. Let me give an example.
Let's say someone is an incredibly virulent racist. This person venerates the KKK. The views are so repugnant we would all vomit if we heard them. But at the same time this racist suffers from a series of perceptual deficiencies. So serious, in fact, that this person will not be able to visually distinguish between race except in rare and ideal (from a sensory POV) circumstances. On measure of explicit behaviour you won't see racism. But that's not a reflection of a belief or attitude in any meaningful way. This is an absurd example designed to show how important perception is to behaviour independent of what we consider attitude.
The idea of implicit prejudice is this analytical problem but in reverse. Most reliable cognitive research into associational concept maps indicates that they are fundamental to the way we build frameworks to process information. Our actual perception of the world varies. This is because the world gives of ABSURD volumes of information. We cannot process it all. So we have frames that filter information to create a coherent worldview and process our situation.
A more "prejudicial" concept map by IAT terms means that the actual framework of the world you are operating in - your perceptions of a situation - are not identical to a person with a different score. Now we might debate how different this really is because no one here is an idealist in the philosophical sense. But it is not identical.
So let's go back to the cop. Having different conceptual associations feeds back in a real way into the cops perception. Where someone with a "non" prejudiced map might focus and foreground information not relevant to danger or threat a "prejudiced" might do the opposite. There is a verital wealth of science showing that social perception is equivocal and influenced by our cognitive frame (cf. social cognitive psychology).
The end result is that the IAT is predictive of behaviour we identify as being prejudiced or not. But to say it is an "attitude" is a bit nonsensical.
That's my objection to the IAT. I think it is garbage from a concept POV.
Retour en haut
