Creative Destruction

June 13, 2006

How prejudiced are you really?

Filed under: Science — Daran @ 5:43 pm

Hat tip: The current edition of Scientific American.

Link

So far I’ve done three of their tests:

Race: Moderate automatic preference for Whites.

Gender-Science: Strong automatic association of male with science.

Microsoft vs. Open Source (Available via the ‘featured task link, top right): Moderate preference for Open Source.

Edit: Sexuality: Moderate automatic preference for heterosexuality.

Edit: Presidents: Moderate preference for Other Presidents over Bush.

It’s a shame that there’s no simple test for preference between men and women. Come to think of it, another good test would be of ones association between “alledged victim”, “accuser”, etc., on the one hand and “alledged rapist” “defendent” etc.,on the other against “innocent”, “truthful”, etc., on the one hand and “guilty”, “lying”, on the other.

Advertisements

24 Comments »

  1. So, are you shocked to find out that you’re a bit racist, sexist, and homophobic?

    My results are on the bizarre side: While I have a mild preference for whites, I also have a strong automatic association of weapons with whites. What the frigg is that supposed to mean? I guess I like white people well enough but expect them to break out in paroxisms of violence once in a while. As far as the others, it appears that I couldn’t really care less about gender, sexual preference, or skin tone.

    Comment by Dianne — June 13, 2006 @ 9:27 pm | Reply

  2. Dianne:
    Who said he’s a bit racist, sexist, and homophobic?

    As far as your second paragraph goes, I think the more interesting observation to make is that you explicitly associate weapons with paroxysms of violence.

    Comment by Brandon Berg — June 14, 2006 @ 1:10 am | Reply

  3. Why do I get the feeling that most of these tests are rigged? As Dianne pointed out, anyone that gets test results that point towards the status quo can be labeled as a bigot, a racist, a sexist, a homophobe, an anti-Semite… And have the hard data to back up the label.

    Also known as a PC maven’s wet dream.

    Comment by Off Colfax — June 14, 2006 @ 3:09 am | Reply

  4. Why do I get the feeling that most of these tests are rigged?

    My guess is, because it suits your political preferences to get the feeling they’re rigged.

    Either you’re suggesting that the tests are literally rigged, or you’re suggesting that the theory behind the tests is mistaken. If the former, then I think that’s a worthless suggestion to make unless you can back it up with some evidence. If the latter, then I’d be interested to hear your argument.

    Comment by Ampersand — June 14, 2006 @ 5:40 am | Reply

  5. Who said he’s a bit racist, sexist, and homophobic?

    The test, which shows that he has a preference for whites (aka associates blacks with negative words), strongly believes that only men can do science, and thinks negatively of homosexuality. Also the last paragraph in his post suggests a certain anti-female bitterness. He’s probably still mourning the fact that women can vote and therefore be on juries and convict men who commit rapes. Everyone knows there are no rapists, only women who falsely accuse, right daran?

    I don’t think the test is “rigged” per se (if it were, how would it be possible that every result is represented?), but it can be gamed if you know how it works and are a sufficiently quick test taker. For example, I can make my anti-black preference go away by concious effort without sending the test into “uninterpretable” range. I’m not sure how common this ability is. One might also point out that the test hasn’t AFAIK been validated. Not that I’m sure how it could be validated. What would the “gold standard” test for unconcious racism be anyway?

    I don’t think that it should come as any real suprise to anyone that most people have unconcious prejudices, though. Look at the society we grew up in. Is it any wonder with all the implicit prejudice around that most people believe, on some level, that straight white men are the best people around?

    Comment by Dianne — June 14, 2006 @ 6:23 am | Reply

    • The bit about sexism is not strictly true – it might also mean he thinks only women can do Liberal Arts. And anyway, sexism as it’s usually meant (associating more positive things with one gender) is not at all proven, unless you think either (Hard) Sciences are “better” than Liberal Arts or the other way around. For example, I got a mild association of men with science and women with liberal arts. I am generally more positive about the liberal arts than the sciences and find them more interesting. Does this mean I have a preference for women over men? Not necessarily, in my opinion

      Comment by Jan Smits van Oyen — March 9, 2016 @ 1:09 pm | Reply

  6. Also the last paragraph in his post suggests a certain anti-female bitterness. He’s probably still mourning the fact that women can vote and therefore be on juries and convict men who commit rapes.

    I don’t think it does suggest anti-female bitterness, nor do I think Daran mourns the fact that women can vote or that rapists are convicted.

    Comment by Tuomas — June 14, 2006 @ 6:28 am | Reply

  7. I don’t think the tests are rigged, (and I think that could be tested for fairly easily), but I do have some reservations about the theory.

    Firstly, it’s rather obvious what part of the test you need to do well or badly on in order to achieve a particular result. This means that a person who is anxious about appearing prejudiced against blacks, for example, might feel more anxious about tests which put good/black on one side vs. bad/white. This might affect their performance in that part of the test.

    Secondly, I wonder how much it’s measuring familiarity. Eighty percent or more of the people I meet are white. Most of the rest are Asian. Consequently I’m fairly unfamiliar with African faces.

    Finally, in the case of the sexuality test, it’s not clear how much it’s measuring the test subjects own sexual preference. (I also notice that there were a couple of male-male images, but no female-female images.)

    Comment by Daran — June 14, 2006 @ 7:45 am | Reply

  8. I do have some reservations about the theory.

    I have one final reservation, not with the theory, but with these particular tests: they’re not nearly long enough. (I would expect the researchers would agree with this.) In my case, I found myself losing concentration several times, with the result successive responses would be particularly slow and or inaccurate. Even though I was being fast and accurate throughout the rest of that part of the test, such lapses could probably affect the result.

    Probably not many internet surfers would be willing to spend the several hours doing a single test, which would be necessary to produce reliable results.

    Dianne above made the point about the tests being gamed. Yes they could be, but what would be the point? Nobody’s judging me on my results except perhaps readers of this blog, and if I was so bothered about that, I could just lie about my results, or not report them at all.

    Comment by Daran — June 14, 2006 @ 7:58 am | Reply

  9. Dianne:

    So, are you shocked to find out that you’re a bit racist, sexist, and homophobic?

    Brandon Berg:

    Who said he’s a bit racist, sexist, and homophobic?

    Maybe I’m a lot racist, sexist and homophobic.

    As far as your second paragraph goes, I think the more interesting observation to make is that you explicitly associate weapons with paroxysms of violence.

    Seems a reasonable association to me. Most major paroxysms of violence have been committed using weapons.

    Comment by Daran — June 14, 2006 @ 8:08 am | Reply

  10. Dianne:

    The test, which shows that he has a preference for whites

    Yes

    (aka associates blacks with negative words)

    The test doesn’t show that. It’s possible that I associate both races with positive words, but whites more strongly so than blacks.

    The test is essentially relative; it indicates preference, as you said above, but it tells us nothing about my absolute associations toward either race. That you infer such is more indicative of your bias, than mine.

    strongly believes that only men can do science,

    This is an absurd extrapolation. That I associate science with men doesn’t imply that I think that women can’t do it. I could probably name more female scientists than you could.

    and thinks negatively of homosexuality.

    See my post above. That I find it easier to associate the word ‘pleasure’ with heterosexuality than homosexuality is probably as much or more an indicator that I’m heterosexual than homophobic.

    Also the last paragraph in his post suggests a certain anti-female bitterness. He’s probably still mourning the fact that women can vote and therefore be on juries and convict men who commit rapes. Everyone knows there are no rapists, only women who falsely accuse, right daran?

    Silly and offensive.

    It’s fairly obvious to me that there are any number of posters and commentors here and on related blogs who are in the “he’s a rapist” and “she’s a liar” camps. Others are more circumspect.

    What evidence do you have that I’m in the “she’s a liar” camp, or are you just showing your own prejudice?

    I don’t think that it should come as any real suprise to anyone that most people have unconcious prejudices, though. Look at the society we grew up in. Is it any wonder with all the implicit prejudice around that most people believe, on some level, that straight white men are the best people around?

    I agree that most people probably prefer straight and white. I’m not convinced they prefer men.

    Comment by Daran — June 14, 2006 @ 9:09 am | Reply

  11. I haven’t yet looked at the test. However, I can say unabashedly that I’m already of the opinion that it’s nonsense merely because of its subject matter. In the U.S., we’re awash in the issues of racism, sexism, ageism, etc. — even if our exposure is limited to reporting (which, naturally, carries its own bias). Because of that, it’s impossible to be untainted; the very environment we live in is tainted.

    I also object to the idea that one can be or even should be neutral or transparent about various -isms. Is it optimal to be without bias? Why? Is a bias worse because it’s strongly held? We’re encouraged to be biases favorably toward democracy, even though most of us have no idea how to define or implement it. Is a stronger bias for democracy better?

    Comment by Brutus — June 14, 2006 @ 10:42 am | Reply

  12. That I associate science with men doesn’t imply that I think that women can’t do it. I could probably name more female scientists than you could.

    You almost certainly can’t. I’m a medical researcher and could name more people of either gender in the field than you’ve ever heard of unless you’re also a researcher.

    Dianne above made the point about the tests being gamed. Yes they could be, but what would be the point? Nobody’s judging me on my results except perhaps readers of this blog, and if I was so bothered about that, I could just lie about my results, or not report them at all.

    The tests are primarily looking for subconcious prejudice. So the person you’d most likely want to hide the results from is yourself. Suppose, for example, you were a 60 year old white ACLU lawyer from Alabama*. You might truly want to have no racial prejudices, have no concious prejudices at all, and desperately want to believe you never judged someone based on their race. Yet you grew up in the deep south prior to the civil rights movement. What’s the chances that you are really free of prejudice? Almost nil. If you were such a person, finding out that you have unconcious prejudices might disturb you deeply. Perhaps enough to want to take the test again and get the “right” answer.

    I suppose there’s also the possibility that you might want to run for public office someday. In which case, assuming you’re not running on the Compassionate Facist ticket, a public admission of a preference for whites, even if you do claim that you’re “pro-white not anti-black”, could be damaging.

    Incidently, if anyone has time to waste, I’d be interested to know how easy anyone else finds gaming the system. Take a test where the result is strong preference for X and see if you can get a strong preference for Y or even neutrality by intentional hesitations, etc or if it just comes up uninterpretable.

    *Conditional tense. None of this should be taken to apply to Daran or any other individual personally. It’s just an example.

    Comment by Dianne — June 14, 2006 @ 11:40 am | Reply

  13. I agree that most people probably prefer straight and white. I’m not convinced they prefer men.

    I separated this one out because I’ve got a lot of links to dump on this issue. The short answer is that sexism is alive and well and has been well characterized.

    For example, this study indicating that women in academic medicine are less likely to become full professors and have lower salaries than men with comparable qualifications.

    Or this one showing that women in academia have lower salaries than men even after adjusting for rank.

    Lest you think it’s only academia that’s a problem, this study shows that women in internal medicine practices make less money than men in similar practices, even after adjusting for other factors that are probably also prejudice, such as likelihood of becoming a partner in the practice.

    I’ll stop now before all these links make me look like spam, but let’s just say that I’ve barely scratched the surface of the evidence of continued gender prejudice in academia and other relatively high pay, high prestige area.

    Comment by Dianne — June 14, 2006 @ 11:53 am | Reply

  14. Dianne:
    The test…shows that he has a preference for whites (aka associates blacks with negative words), strongly believes that only men can do science, and thinks negatively of homosexuality.

    Uh…no. You draw much too strong a conclusion. At best, the test measures implict associations, and it may not even do that well:

    You may be giving the test more credit than it deserves! The web versions of these tests are necessarily less accurate than our laboratory versions (and we don’t regard the lab versions as perfectly accurate).

    Your first conclusion may be valid in a weak sense–although probably not enough to call him a racist–but the next two are in no way warranted based on the test results.

    An implicit association between men and science is to be expected in a world where most prominent scientists are men, and in no way indicates a strong belief that only men can do science (though the fact that you think it does provides a data point supporting said belief). Nor does an implicit association of heterosexuality with good things necessarily indicate that one thinks negatively of homosexuality.

    Ampersand:
    I don’t object to the actual theory behind the tests so much as the unwarranted conclusions (see above) that some people form based on what they think is the theory behind the tests.

    Daran:
    Most major paroxysms of violence have been committed using weapons.

    I doubt it, unless we define “weapons” to include body parts. I wouldn’t know where to look for data, but I suspect that unarmed violence is a lot more common than the armed variety. Conversely, the vast majority of weapons are never used violently.

    Comment by Brandon Berg — June 14, 2006 @ 12:47 pm | Reply

  15. Me:

    Most major paroxysms of violence have been committed using weapons.

    Brandon Berg:

    I doubt it, unless we define “weapons” to include body parts. I wouldn’t know where to look for data, but I suspect that unarmed violence is a lot more common than the armed variety.

    For the purpose of this discussion, I’ll define ‘weapon’ to mean a contrivance created or adapted for such use, and ‘violence’ to refer to human actions intend to cause harm to others.

    I don’t doubt that unarmed violence is more common than armed violence, however we’re not talking about all violent acts, but ‘major paroxysms’. The most ‘major’ weapon-free violent incident I can think of is 9/11, which was modest in comparison to the atomic attacks on Japan, and countless acts of warfare using conventional weapons.

    And there’s a huge gap between 9/11 and the next most major weapon-free attacks, which I guess would be several incidents of arson which have claimed a few tens, maybe even hundreds of lives.

    I think you’d have to set the bar of ‘major’-ness awfully low to have weapon free incidents dominate.

    Conversely, the vast majority of weapons are never used violently.

    The vast majority of non-weapons are never used violently. Moreover, I would suggest that the proportion of weapons that have been used violently is vastly greater than the proportion of non-weapons so used.

    Your argument seems akin to claiming that we shouldn’t associate Hiroshima and Nagasaki with the use of nuclear weapons, because the vast majority of such weapons have never been used to attack these cities, and the vast majority of the cities’ inhabitents have never been attacked with such weapons.

    Comment by Daran — June 14, 2006 @ 3:53 pm | Reply

  16. Dianne:

    The short answer is that sexism…

    in favour of males over females

    …is alive and well and has been well characterized.

    Unquestionably.

    Unfortunately sexism in favour of females over males is also alive and well, and much less well characterised.

    [cites snipped, but noted]

    Lest you think it’s only academia that’s a problem…

    I certainly don’t deny the kinds of sexism you’ve refered to, or suggest it’s restricted to academia. However this isn’t the only form of sexism endemic in our society. See this paper for example.

    Note that this neither contradicts nor is contradicted by your cites. It’s quite possible (indeed verifiable) that both forms of sexism are concurrent.

    I’ll stop now before all these links make me look like spam…

    I regularly check the spam posts for false positives. If any of your posts gets caught, either I or one of the other blogger will recover it.

    Comment by Daran — June 14, 2006 @ 6:05 pm | Reply

  17. Dianne:

    You almost certainly can’t [name more female scientists]. I’m a medical researcher and could name more people of either gender in the field than you’ve ever heard of unless you’re also a researcher.

    Whatever. Let’s just take it as read that I could name scores of female scientists, both current and historic, and therefore am well aware that it’s not and never really has been a solely male domain.

    The tests are primarily looking for subconcious prejudice. So the person you’d most likely want to hide the results from is yourself. Suppose, for example, you were a 60 year old white ACLU lawyer from Alabama*. You might truly want to have no racial prejudices, have no concious prejudices at all, and desperately want to believe you never judged someone based on their race. Yet you grew up in the deep south prior to the civil rights movement. What’s the chances that you are really free of prejudice? Almost nil. If you were such a person, finding out that you have unconcious prejudices might disturb you deeply. Perhaps enough to want to take the test again and get the “right” answer.

    Well I’ve never pretended, even to myself, that I have no prejudices, but I’d certainly like that to be so. So I should be motivated to subconsciously game the test in the direction of being less favourable towards whites, heterosexuals, etc. As you saw, if there was such a subconcious gaming process, it was insufficient to eliminate the preference.

    I suppose there’s also the possibility that you might want to run for public office someday. In which case, assuming you’re not running on the Compassionate Facist ticket, a public admission of a preference for whites, even if you do claim that you’re “pro-white not anti-black”, could be damaging.

    I can’t speak for anyone else, but, all other things being equal, I’d support a candidate who acknowledged his biases, but aspired to be bias-free, over one who denied any bias at all, and support either over one who embraced their bias

    Incidently, if anyone has time to waste, I’d be interested to know how easy anyone else finds gaming the system. Take a test where the result is strong preference for X and see if you can get a strong preference for Y or even neutrality by intentional hesitations, etc or if it just comes up uninterpretable.

    OK, I tried that several times with the gender-science test Here are the results for successive gamed attempts (remember my ungamed result was for a strong male-science association.)

    1. Moderate association for male with science.
    2. Indeterminate
    3. Moderate association for female with science.

    And just for comparison, here are the results of two more not consciously gamed tests

    1. Slight association for male with science.
    2. Slight association for female with science.

    I really don’t know what to make of this. One thing I did observe is that ‘science’ and ‘liberal arts’ were always on the same sides – ‘science’ to the left, and ‘liberal arts’ to the right. The genders changed sides. I think this makes the tests easier and less sensitive to ones internal prejudices as you do more of them.

    Comment by Daran — June 14, 2006 @ 7:24 pm | Reply

  18. Brutus:

    I haven’t yet looked at the test. However, I can say unabashedly that I’m already of the opinion that it’s nonsense merely because of its subject matter.

    All that demonstrates is that you’re prejudiced against such tests.

    Comment by Daran — June 14, 2006 @ 7:32 pm | Reply

  19. I took several of the tests.

    Age: a slight automatic preference for Young compared to Old

    Race (Black-White): a strong automatic preference for Black compared to White.

    Gender Science: little or no association between Female and Male with Science and Liberal Arts.

    Sexuality: little to no automatic preference between Gay People and Straight People.

    Gender Career: a slight association of Female with Career and Male with Family compared to Male with Career and Female with Family.

    The results were not really surprising except for the last one. I assume I made that connection because my former foster father is the primary caregiver.

    While the test may be designed to look for subconscious prejudices, all it really demonstrates is preference, which says little. No one should infer anything more from these kinds of results because the format does not get into any specifics.

    Comment by toysoldier — June 15, 2006 @ 7:57 pm | Reply

  20. I can’t speak for anyone else, but, all other things being equal, I’d support a candidate who acknowledged his biases, but aspired to be bias-free, over one who denied any bias at all, and support either over one who embraced their bias

    My apologies for my randomness in responding to your comments. I’m afraid my time for/interest in playing on the net waxes and wanes with how much I’m procrastinating. However, I did want to clarify what I meant here.

    In principle, I agree that, between two candidates, one of whom acknowledged his or her unconcious biases and one who didn’t, the one who did would be the better candidate. However, I was thinking of mudslinging politics, not ideals when I said that it might be detrimental to you to publically acknowledge your biases. For example, suppose you were running for mayor of NYC. Your opponent could canvas Harlem and the Bronx, knocking on people’s doors and asking them if they would be willing to vote for a candidate who said that he preferred whites to blacks and maybe threw out a random quote or two to support your alleged bias. It’d probably hurt your campaign.

    Comment by Dianne — June 16, 2006 @ 3:13 pm | Reply

  21. I certainly don’t deny the kinds of sexism you’ve refered to, or suggest it’s restricted to academia. However this isn’t the only form of sexism endemic in our society. See this paper for example.

    Interesting paper. Didn’t something similar happen in Fallujah? I seem to remember that at one point the US had blocked all exits to Fallujah and were forcing all men and boys over 15 years of age to stay in the city they were about to attack. Yech.

    However, when people go out of their way to attack young men, they do so because they see them as more of a threat because they believe them more competent and capable of causing problems for the invaders than women. Women and children rate a particular mention when they are victims of massacres because it is assumed that they are non-combatants and that it should be clear to anyone that they are non-combatants. So in a twisted sort of way, it’s a bias in men’s favor that is being used against them.

    Suppose a crazy person came into your workplace and started shooting everyone he saw because he hated people in your profession and wanted to kill all of them. But suppose he looked at you, shook his head, and said “naw…this one’s clearly an innocent bystander: he’s far to stupid to be a {whatever it is you do}”. You might be relieved about not getting shot, but I doubt you would claim that your survival indicated that the shooter thought better of you than of his victims. The massacre of men in war is similar: it occurs because the perpetrators believe men, but not women, to be competent enough to threaten them.

    Comment by Dianne — June 16, 2006 @ 3:27 pm | Reply

  22. One thing I did observe is that ’science’ and ‘liberal arts’ were always on the same sides – ’science’ to the left, and ‘liberal arts’ to the right. The genders changed sides. I think this makes the tests easier and less sensitive to ones internal prejudices as you do more of them.

    You should write the authors. I think you’ve discovered a significant and correctable flaw in the test.

    Comment by Dianne — June 16, 2006 @ 3:29 pm | Reply

  23. Argh. These tests drive me crazy.

    They’re trying to be a proxy for whether you associate good or bad things with races, or genders, or whatever. But the tests are in essence a self-report.

    If you want a self-reported proxy for whether someone is a racist, take a self-report from the person. That will be your richest possible data source.

    Comment by Robert — June 17, 2006 @ 3:14 am | Reply


RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Create a free website or blog at WordPress.com.

%d bloggers like this: