https://implicit.harvard.edu/implici...lectatest.html
Want to find out if you have negative associations for various groups? Try whichever one interests you. Were you surprised? What can you actually conclude from your result? The poll is anonymous.
Printable View
https://implicit.harvard.edu/implici...lectatest.html
Want to find out if you have negative associations for various groups? Try whichever one interests you. Were you surprised? What can you actually conclude from your result? The poll is anonymous.
I have not clicked the link but I voted that I'm not surprised that I'm biased. I base this on the fact that my brain is working. Most of the time, anyways.
Judgments are there for a reason. The goal of absolving the ego or whatever you want to call it is not to remove your own judgments or subjective reactions to your environment. It's simply not to cling to them or contain the world in your judgments. You realize your judgments are only there to base decisions upon, not to define reality.
I tried the sexuality related test, said I prefered heterosexuals a little bit over homosexuals. I think it has more to do with how the test was presented though, muscle memory builds fast, seems like some extra conscious effort would be needed, when things get switched around. Even then, the result does not really surprise me that much. I think society has done its part, in skewing me towards that result.
For the Age IAT, I scored "moderate automatic preference for young compared to old".
For the Sexuality IAT, I scored "slight automatic preference for straight people compared to gay people".
For the Weight IAT, I scored "moderate automatic preference for thin people compared to fat people".
For the Countries IAT, I scored "little to no automatic preference between the United Kingdom and the United States".
For the Race IAT, I scored "slight automatic preference for black people compared to white people".
For the Skin-tone IAT, I scored "little to no preference between light skinned people and dark skinned people".
For the Gender IAT, I scored "strong association of male with science and female with liberal arts".
Some of those results surprised me a little, especially the Race IAT results. Being white, I would have expected my bias to be skewed in favour of white people, but for some reason that wasn't the case. I see myself as quite liberal and progressive. It just seems like common sense to me that things like gender, sexuality and skin colour shouldn't be all that important to us. I'm sure the general views of society and my personal connections must have an affect on my own feelings. I'm close to quite a few racists and bigots in my personal life, and whenever they go on a rant about how the foreigners are taking over, I usually just roll my eyes and wish they would grow the fuck up. Maybe my results in the Race IAT reflect my opposition to their paranoid bullshit.
Your data suggest a moderate association of Black Americans with Weapons compared to White Americans.
Hm. Interesting. I'll probably try a few more of these later.
I have actually done some of those before, so the results are not surprising. Everyone will probably find they have some biases in many areas, because we are bombarded by biased images in the media, and in our schools and every where we look. It is natural that people pick up those biases subconsciously, even if they personally disagree with them. People often act on their unconscious biased without even realizing it. The trick is to realize when you are doing it, so you can correct yourself.
Being largely white myself I scored a slight preference in favor of white people. Of course I was faster at the task by the time I had a chance to associate "good" with "white".Quote:
Originally Posted by Heavy Sleeper
I think that the test is flawed in that while they randomly present the order, they can only correct for the speed with which the mechanical task is learned based on statistical inferences which are themselves subject to assumptions. Hence without the correction, the bias detected will be skewed in favor of the last race to be associated with "good". Some people learn these things faster than others. I perhaps learn slightly faster than average and so the correction was insufficient for me.
They should give you a system to play around with that mirrors the mechanical skills required to eliminate this built in bias directly as opposed to mucking around with statistical assumptions.
This is yet another reason I have a hard time taking the conclusions of psychology seriously. How do you control for these things?
I have negative associations with allegedly well defined groups in general.
I took the first test first... the preference of light-skinned vs. dark skin. I was laughing by the time I fumbled through the picture/word sets. I agree with Philosopher though that I was more used to it the second time around and was able to respond faster. The results said I have a strong preference for light skin over dark skin, but when I clicked on the details, it's actually Moderate.
But either way, I'm very surprised.
Maybe I'm wired that way sub-consciously, but that's not how I act.
Now I'm going to go take some more....
I wonder if taking many of these tests, with different subjects, ruins the integrity of the test, or makes it more accurate.
If you don't think its accurate, take it a second time if they results are similar then it was probably accurate. Should also try a few different ones as the results are likely different with different areas.
A persons first reaction is to always think that it can't be true and there must be something wrong, but that is because most of us are good people and we don't like seeing we have a bias.
I took a bunch of them a while ago, and the results didn't seem effected by how fast they were done, or in which order they were presented.
If they didn't try to correct for the fact that there is a skill being learned then I imagine that the earlier tests would serve as practice and alllow the later ones to be more accurate. I bet almost anything that it would be best to just give people practice for (say) ten minutes first.Quote:
Originally Posted by Marvo
Test seems like bullshit. No offense. :P
Being faster doesn't really matter, since you would be faster at pressing the key for both group. If you think about it, if you learn to associate one group with another faster than you do with the other group, that doesn't throw it off. Since that is part of what is being tested.
A couple responses concerning methodology:
- Alric pretty much has it on the head with his comment above. What is of interest is the differential response times to stereotype-congruent pairs of stimuli vs. stereotype-incongruent pairs. While learned familiarity with the task certainly increases over the course of the session, this learning should result in increased speed and accuracy for all responses, not preferential learning for some pairings. In fact, if learning did occur at a different rate for stereotype-congruent pairs compared to stereotype-incongruent pairs, this would if anything indicate precisely the response bias that the researchers are interested in, albeit in a slightly different form (i.e, evidence that it is easier to learn associations between stimuli that have a prior established association due to cultural stereotypes).
- As mentioned above, learning occurs in the course of the task. Such learning curves typically are well described by negative exponential functions (e.g., rate of mistakes often conforms well to a function of the form e^(-p*r), with p = presentation number and r = a learning rate parameter to be estimated from the data). In other words, learning occurs rapidly in the beginning but then quickly slows down (although the learning rate is constant with respect to what is left to be learned--which is interesting from a psychological perspective even if not obviously relevant to this discussion). This is why a series of practice trials are included--only the data from the final blocks are used, where the effects of learning should be minimal.
- The point about results being slightly biased in favor of the last stimulus-pair type to be presented, since this is when maximum learning has occurred, is an interesting observation, but not quite right. What could bias the results in favor of one category over another would not be the final presentation per se, but more precisely the average position of the categories in the presentation sequence. It will be obvious on reflection that a stimulus category could receive the last response in time despite occurring earlier on average throughout the series of presentations. This should also make clear that it is actually not too hard to control for this confound since it is not too hard to hold constant the average position within the presentation sequence. Of course, even if this were for some reason not achieved for a particular subject, counter-balancing the final presentation category across subjects (i.e., have half of participants end on a congruent pair while the other half end on an incongruent pair) would be a simple way to alleviate bias in the aggregate. And finally, all of this is assuming that there are in fact substantial learning effects in these final trials in the first place--which point #2 above suggests is doubtful.
That isn't what this test reflects. I'm sure nobody taking it has some kind of logical reason that blacks are inferior to whites. It is testing more deeply ingrained, irrational stereotypes. Either you're not being entirely honest or you're quite unusual, because even black Americans generally have a negative association towards black people on this test. It's obviously not about your higher cognitive views, then.Quote:
Originally Posted by Heavy Sleeper
Well, I'm not lying, if that's what you're saying. Those are the results I got. I haven't tried the test a second time to see if it would make any difference. If it is accurate, then it certainly is unusual and I have no good explanation for it.Quote:
Originally Posted by Xei
I wouldn't doubt it. It is certainly possible to have your unconscious reactions move to be more aligned with your conscious thoughts of the issue. It is just very difficult. It is also possible that Heavy sleeper just never got exposed to the stuff in the first place, or was more aware of it at a younger age or did a good job and keeping any of it from sinking in.
Also everyone is different. I would suspect that most people have strong or moderate biases in some areas, and little or no bias in others.
Is it prejudiced to have slightly negative feelings towards a group when those feelings can be justified by practical reasons?
For example, I'm going to be negatively inclined towards overweight people because they can have negative impacts on me, such as cutting into my personal space in a cramped environment (e.g. on a plane), or making it more difficult for me to navigate enclosed spaces.
I'm sure I am prejudiced at least a little, no matter how hard I try not to be. It's extremely difficult to remove all traces of personal bias.
Oh man I have to try the fat people's test, I hate fat people.
I am taking another test (the fat people one) because It should detect a fairly strong bias. I'm not even into the real part yet (so I don't mind stepping away from it) and I'm already enraged. "How much control do people have over their weight?" Seriously? Some people have a lot of control some people have almost none. How am I supposed to respond to this?
I hope to be able to write an adequate response to DuB's post later. It's illuminating but it sounds like he's describing a different test from the 'black-white' one that I took.
I just took the fat one and I have a slight automatic preference to fat people over thin people :D
Other than my siblings, the rest of my family is very obese, including one of my my kids. I guess I can empathize more.
I'm gay and got a 'slight preference to straight people.' XD
Your data suggest a strong automatic preference for Thin People compared to Fat People.
Haha oh man
Your data suggest little to no automatic preference between Fat People and Thin People. Interesting... I would have thought I would have been biased against fat people.
Your data suggest a slight automatic preference for Straight People compared to Gay People. I wasn't expecting anything too biased here.
I also preferred Hinduism and Islam the most, Christianity less, and Judaism even less...