Total Posts:10|Showing Posts:1-10
Jump to topic:

IAT and implicit bias

ax123man
Posts: 317
Add as Friend
Challenge to a Debate
Send a Message
10/9/2015 11:44:52 AM
Posted: 1 year ago
I was just looking over an faq on the IAT, a test you can take to measure implicit bias. My immediate thought was, "how do you distinguish the desire for freedom of association and the racial "bias"? Take a look at item #8 in the faq:

https://implicit.harvard.edu...

"Is the common preference for White over Black in the Black-White attitude IAT a simple 'ingroup' preference -- for example, the same as liking members of one's family or feeling connected to people who come from one's hometown?

Answer: For White respondents, the automatic White preference may in some sense be an ingroup preference. However, the automatic White preference is more than that -- it is observed with similar strength among Asian Americans, for whom neither Black nor White is an ingroup."

I must be thick. I don't understand that last sentence. Maybe in-group has special meaning here. I just think of it being "people that you live around, hang around with", etc. How can they make blanket statements across race regarding in-groups?
Greyparrot
Posts: 14,325
Add as Friend
Challenge to a Debate
Send a Message
10/9/2015 12:40:29 PM
Posted: 1 year ago
At 10/9/2015 11:44:52 AM, ax123man wrote:
I was just looking over an faq on the IAT, a test you can take to measure implicit bias. My immediate thought was, "how do you distinguish the desire for freedom of association and the racial "bias"? Take a look at item #8 in the faq:

https://implicit.harvard.edu...

"Is the common preference for White over Black in the Black-White attitude IAT a simple 'ingroup' preference -- for example, the same as liking members of one's family or feeling connected to people who come from one's hometown?

Answer: For White respondents, the automatic White preference may in some sense be an ingroup preference. However, the automatic White preference is more than that -- it is observed with similar strength among Asian Americans, for whom neither Black nor White is an ingroup."

I must be thick. I don't understand that last sentence. Maybe in-group has special meaning here. I just think of it being "people that you live around, hang around with", etc. How can they make blanket statements across race regarding in-groups?

Bias is bad. assimilation is good.
TBR
Posts: 9,991
Add as Friend
Challenge to a Debate
Send a Message
10/9/2015 4:22:47 PM
Posted: 1 year ago
At 10/9/2015 11:44:52 AM, ax123man wrote:
I was just looking over an faq on the IAT, a test you can take to measure implicit bias. My immediate thought was, "how do you distinguish the desire for freedom of association and the racial "bias"? Take a look at item #8 in the faq:

https://implicit.harvard.edu...

"Is the common preference for White over Black in the Black-White attitude IAT a simple 'ingroup' preference -- for example, the same as liking members of one's family or feeling connected to people who come from one's hometown?

Answer: For White respondents, the automatic White preference may in some sense be an ingroup preference. However, the automatic White preference is more than that -- it is observed with similar strength among Asian Americans, for whom neither Black nor White is an ingroup."

I must be thick. I don't understand that last sentence. Maybe in-group has special meaning here. I just think of it being "people that you live around, hang around with", etc. How can they make blanket statements across race regarding in-groups?

I think these tests, and the terms used generally should be... less of a concern. That is, when someone says "you are being a raciest (or bigot)" the first reaction should NOT be to become offended. I am not saying you should aspire to be either, but AM saying that the terms have use to both people in the conversation. You can stop and evaluate why the term is being used, how it effects what you are saying or feeling, and respond.

Despite what many say, there are no PC Police. You are NOT going to the reeducation camp. If you are called a bigot in some conversation, you might very well be saying something bigoted. Now, how does that effect your statement or thought? Nothing more.
ax123man
Posts: 317
Add as Friend
Challenge to a Debate
Send a Message
10/9/2015 4:55:41 PM
Posted: 1 year ago
At 10/9/2015 4:22:47 PM, TBR wrote:
At 10/9/2015 11:44:52 AM, ax123man wrote:
I was just looking over an faq on the IAT, a test you can take to measure implicit bias. My immediate thought was, "how do you distinguish the desire for freedom of association and the racial "bias"? Take a look at item #8 in the faq:


Despite what many say, there are no PC Police. You are NOT going to the reeducation camp. If you are called a bigot in some conversation, you might very well be saying something bigoted. Now, how does that effect your statement or thought? Nothing more.

I didn't even take the test. All I'm trying to do is understand what they're doing, or think they're doing, so I can prove/disprove in my own mind whether it's valid. It seems to me that this specific faq item is critical to their study.

AFA there being no PC police, I beg to differ. I've seen them in person and been a victim of their wrath. In one case, this involved receiving a letter in the mail on this particular PC cop's lawyer's company letter head. Fortunately, this was easily resolved when I informed the company of this and that it violated the state bars code of ethics since it had nothing to do with a company legal matter. And I can promise you the guy was most assuredly in the wrong, not only for the letter, but for the event that triggered it.

Perhaps I misunderstand what you mean. If I google "ridiculous political correctness", are all those stories made up? Or is it that they amount to nothing because nobody's being physically harmed?
TBR
Posts: 9,991
Add as Friend
Challenge to a Debate
Send a Message
10/9/2015 5:12:59 PM
Posted: 1 year ago
AFA there being no PC police, I beg to differ. I've seen them in person and been a victim of their wrath. In one case, this involved receiving a letter in the mail on this particular PC cop's lawyer's company letter head. Fortunately, this was easily resolved when I informed the company of this and that it violated the state bars code of ethics since it had nothing to do with a company legal matter. And I can promise you the guy was most assuredly in the wrong, not only for the letter, but for the event that triggered it.

Where? Under what authority do they operate. As you say, they (whomever) have no authority. This is a stupid term that needs to go away more than the term PC to start with.


Perhaps I misunderstand what you mean. If I google "ridiculous political correctness", are all those stories made up? Or is it that they amount to nothing because nobody's being physically harmed?

I am saying - as plan as possible - you may be called a raciest at some point, and whatever instigated the term being used just might be correct. How you use the information is the key. Get all huffy about the term, or evaluate what was said, your choice. If you are fine, internally, with what you said or did, fine. If not, well then the information was good to hear.
ax123man
Posts: 317
Add as Friend
Challenge to a Debate
Send a Message
10/9/2015 6:25:33 PM
Posted: 1 year ago
At 10/9/2015 5:12:59 PM, TBR wrote:
AFA there being no PC police, I beg to differ. I've seen them in person and been a victim of their wrath. In one case, this involved receiving a letter in the mail on this particular PC cop's lawyer's company letter head. Fortunately, this was easily resolved when I informed the company of this and that it violated the state bars code of ethics since it had nothing to do with a company legal matter. And I can promise you the guy was most assuredly in the wrong, not only for the letter, but for the event that triggered it.

Where? Under what authority do they operate. As you say, they (whomever) have no authority. This is a stupid term that needs to go away more than the term PC to start with.

Thugs seem to find a way to get authority anyway they can. Justice doesn't always prevail. "PC police" is just a term to refer to these types: not only overly sensitive but ready to pull the trigger on any offenses within ear shot. Although I don't tend to use the term, I don't see a problem with it.


Perhaps I misunderstand what you mean. If I google "ridiculous political correctness", are all those stories made up? Or is it that they amount to nothing because nobody's being physically harmed?

I am saying - as plan as possible - you may be called a raciest at some point, and whatever instigated the term being used just might be correct. How you use the information is the key. Get all huffy about the term, or evaluate what was said, your choice. If you are fine, internally, with what you said or did, fine. If not, well then the information was good to hear.

I get what your saying. However, I may be fine with my words/actions but when the other party takes some action against me that goes beyond just words, I have a problem.
feverish
Posts: 2,716
Add as Friend
Challenge to a Debate
Send a Message
10/15/2015 9:17:11 PM
Posted: 1 year ago
At 10/9/2015 11:44:52 AM, ax123man wrote:
I was just looking over an faq on the IAT, a test you can take to measure implicit bias. My immediate thought was, "how do you distinguish the desire for freedom of association and the racial "bias"?

I would of thought the two were easily distinguishable and hard to confuse... Perhaps you mean preference of association and racial bias?

Answer: However, the automatic White preference is more than that -- it is observed with similar strength among Asian Americans, for whom neither Black nor White is an ingroup."

I must be thick. I don't understand that last sentence. Maybe in-group has special meaning here. I just think of it being "people that you live around, hang around with", etc. How can they make blanket statements across race regarding in-groups?

The term "in-group" means exactly what it implies, belonging to the same group. Context is crucial; as the context here is race, same race equals same in-group. Depending on context, your own "in-group" might refer to your race, your family, your species, members of this site, people with the same middle name or many other things.

At 10/9/2015 4:22:47 PM, TBR wrote:
I think these tests, and the terms used generally should be... less of a concern.

I find it really interesting how these tests can provoke such a hostile reaction, I remember sharing it on this site a while ago when I first discovered them and several anti racist people seemed quite angered at being informed that they had the typical bias for white faces. The typical reaction seems to be a firm belief that the test itself is deeply flawed, with no coherent explanation of how.

More recently I showed it to a lady friend (who is neither black nor white) who responded in a similar way. We ended up having a massive argument about it!

But yeah, people seem to take it very personally when there is any suggestion that they might have any prejudice or bias, rather than attempting to learn something from it.
ax123man
Posts: 317
Add as Friend
Challenge to a Debate
Send a Message
10/16/2015 12:11:21 AM
Posted: 1 year ago
At 10/15/2015 9:17:11 PM, feverish wrote:
At 10/9/2015 11:44:52 AM, ax123man wrote:
I was just looking over an faq on the IAT, a test you can take to measure implicit bias. My immediate thought was, "how do you distinguish the desire for freedom of association and the racial "bias"?

I would of thought the two were easily distinguishable and hard to confuse... Perhaps you mean preference of association and racial bias?

that's what I meant.


Answer: However, the automatic White preference is more than that -- it is observed with similar strength among Asian Americans, for whom neither Black nor White is an ingroup."

I must be thick. I don't understand that last sentence. Maybe in-group has special meaning here. I just think of it being "people that you live around, hang around with", etc. How can they make blanket statements across race regarding in-groups?

The term "in-group" means exactly what it implies, belonging to the same group. Context is crucial; as the context here is race, same race equals same in-group. Depending on context, your own "in-group" might refer to your race, your family, your species, members of this site, people with the same middle name or many other things.

Yea, I get that. But the way it's worded is strange. There's no reason to add on the "for whom neither Black nor White is an ingroup" unless you are trying to say that some races are an ingroup for other races, and some are not (which I doubt they were trying to say). It would have been much more clear worded as "it is observed with similar strength among Asian Americans, who also appear to prefer whites". IF that is what they are saying. If that IS what they are trying to say, it's a little odd then that you rarely see anything but Asian Americans working at restaurants owned by Asian Americans. So what exactly IS it this test is supposed to be telling us? What does it mean to "prefer" in this context?


At 10/9/2015 4:22:47 PM, TBR wrote:
I think these tests, and the terms used generally should be... less of a concern.

I find it really interesting how these tests can provoke such a hostile reaction, I remember sharing it on this site a while ago when I first discovered them and several anti racist people seemed quite angered at being informed that they had the typical bias for white faces. The typical reaction seems to be a firm belief that the test itself is deeply flawed, with no coherent explanation of how.

More recently I showed it to a lady friend (who is neither black nor white) who responded in a similar way. We ended up having a massive argument about it!

But yeah, people seem to take it very personally when there is any suggestion that they might have any prejudice or bias, rather than attempting to learn something from it.

When I read over the above statements (and correct me if I'm wrong), there seems to an implied belief that the test is probably right and the "complainers" are probably wrong. Seems like the more rational approach would be to wonder about the study (since it is undeniable that a very high percentage of studies are flawed) and want to look at it further before drawing any conclusions:

http://www.nytimes.com...
feverish
Posts: 2,716
Add as Friend
Challenge to a Debate
Send a Message
10/16/2015 8:40:26 AM
Posted: 1 year ago
At 10/16/2015 12:11:21 AM, ax123man wrote:
At 10/15/2015 9:17:11 PM, feverish wrote:
At 10/9/2015 11:44:52 AM, ax123man wrote:
"how do you distinguish the desire for freedom of association and the racial "bias"?

I would of thought the two were easily distinguishable and hard to confuse... Perhaps you mean preference of association and racial bias?

that's what I meant.

Well bias is basically a synonym for preference, so if your preference for association is based on racial grounds then it would indeed be racial bias.


Answer: However, the automatic White preference is more than that -- it is observed with similar strength among Asian Americans, for whom neither Black nor White is an ingroup."

The term "in-group" means exactly what it implies, belonging to the same group. Context is crucial; as the context here is race, same race equals same in-group.

Yea, I get that. But the way it's worded is strange. There's no reason to add on the "for whom neither Black nor White is an ingroup" unless you are trying to say that some races are an ingroup for other races, and some are not (which I doubt they were trying to say).

But it is a rebuttal to the suggestion that the test is merely measuring in-group bias. They are pointing out that not only white people have this bias. In fact even "African Americans" themselves have been known to show the same bias, as in the famous doll tests. https://explorable.com...

It would have been much more clear worded as "it is observed with similar strength among Asian Americans, who also appear to prefer whites".

That would just be a tautology "white preference... is observed with similar strength among Asian Americans, who also appear to prefer whites"

IF that is what they are saying. If that IS what they are trying to say, it's a little odd then that you rarely see anything but Asian Americans working at restaurants owned by Asian Americans. So what exactly IS it this test is supposed to be telling us? What does it mean to "prefer" in this context?

Well firstly they are saying that Asian Americans show a preference for white over black, rather than white over Asian.

Also, the important point to take from this is that the test measures implicit bias, rather than conscious preference, racism or discrimination. I agree with many of the points that your NY times link makes, in that this test should not be used to predict actual discrimination or to label people as racist. I certainly don't consider myself to be a racist yet when I took this test a while ago, I too showed implicit bias.

I find it really interesting how these tests can provoke such a hostile reaction, I remember sharing it on this site a while ago when I first discovered them and several anti racist people seemed quite angered at being informed that they had the typical bias for white faces. The typical reaction seems to be a firm belief that the test itself is deeply flawed, with no coherent explanation of how.

More recently I showed it to a lady friend (who is neither black nor white) who responded in a similar way. We ended up having a massive argument about it!

But yeah, people seem to take it very personally when there is any suggestion that they might have any prejudice or bias, rather than attempting to learn something from it.


When I read over the above statements (and correct me if I'm wrong), there seems to an implied belief that the test is probably right and the "complainers" are probably wrong.

I'm not suggesting the test is perfect but yes, it is my belief that the test is probably showing us something significant. When a vast majority of people appear to demonstrate a bias that clearly fits with a history of oppression and racism, as well as current stereotypes in media and society, then I think it is reasonable to deduce that something is probably going on here.

Seems like the more rational approach would be to wonder about the study (since it is undeniable that a very high percentage of studies are flawed) and want to look at it further before drawing any conclusions:

I agree that it is rational to look in to how the study may be flawed, which you and I have both apparently done by checking out the FAQs and thinking them through for ourselves.
On the other hand, I think it is highly irrational to dismiss the results of a test because you don't like them on a personal level, rather than because you have discovered a legitimate flaw with the test.
ax123man
Posts: 317
Add as Friend
Challenge to a Debate
Send a Message
10/16/2015 11:49:23 AM
Posted: 1 year ago
At 10/16/2015 8:40:26 AM, feverish wrote:
At 10/16/2015 12:11:21 AM, ax123man wrote:
At 10/15/2015 9:17:11 PM, feverish wrote:
At 10/9/2015 11:44:52 AM, ax123man wrote:
"how do you distinguish the desire for freedom of association and the racial "bias"?

I would of thought the two were easily distinguishable and hard to confuse... Perhaps you mean preference of association and racial bias?

that's what I meant.

Well bias is basically a synonym for preference, so if your preference for association is based on racial grounds then it would indeed be racial bias.

I think I need to just be explicit. If someone would rather associate with their own race, let them. It's their business. Forced (or coerced, nudged, etc) integration creates divisiveness among people. I don't have a problem with somebody pointing it out, or studies that show this is happening.That is, as long as there is a reasonable balance - that political correctness doesn't overtake us. Also inevitably, it turns into policy backed by law. I think that's unnecessary. First government oppressed via slavery, then oppressed via segregation, then forced integration, etc. What are the odds they will ever get it right?

So, in theory, I don't have a problem with the test, and others shouldn't have a problem with it being questioned.



Answer: However, the automatic White preference is more than that -- it is observed with similar strength among Asian Americans, for whom neither Black nor White is an ingroup."

The term "in-group" means exactly what it implies, belonging to the same group. Context is crucial; as the context here is race, same race equals same in-group.

Yea, I get that. But the way it's worded is strange. There's no reason to add on the "for whom neither Black nor White is an ingroup" unless you are trying to say that some races are an ingroup for other races, and some are not (which I doubt they were trying to say).

But it is a rebuttal to the suggestion that the test is merely measuring in-group bias. They are pointing out that not only white people have this bias. In fact even "African Americans" themselves have been known to show the same bias, as in the famous doll tests. https://explorable.com...

In this test, black's show an equal preference toward both blacks and whites.


It would have been much more clear worded as "it is observed with similar strength among Asian Americans, who also appear to prefer whites".

That would just be a tautology "white preference... is observed with similar strength among Asian Americans, who also appear to prefer whites"

your right, that was silly of me. Perhaps I struggled with the last part of that sentence because it's so obvious that I was looking for different meaning.


IF that is what they are saying. If that IS what they are trying to say, it's a little odd then that you rarely see anything but Asian Americans working at restaurants owned by Asian Americans. So what exactly IS it this test is supposed to be telling us? What does it mean to "prefer" in this context?

Well firstly they are saying that Asian Americans show a preference for white over black, rather than white over Asian.

Also, the important point to take from this is that the test measures implicit bias, rather than conscious preference, racism or discrimination. I agree with many of the points that your NY times link makes, in that this test should not be used to predict actual discrimination or to label people as racist. I certainly don't consider myself to be a racist yet when I took this test a while ago, I too showed implicit bias.

ok, thanks.


I find it really interesting how these tests can provoke such a hostile reaction, I remember sharing it on this site a while ago when I first discovered them and several anti racist people seemed quite angered at being informed that they had the typical bias for white faces. The typical reaction seems to be a firm belief that the test itself is deeply flawed, with no coherent explanation of how.

More recently I showed it to a lady friend (who is neither black nor white) who responded in a similar way. We ended up having a massive argument about it!

But yeah, people seem to take it very personally when there is any suggestion that they might have any prejudice or bias, rather than attempting to learn something from it.

Faq 13 from the IAT site:

Q "When will implicit attitudes agree with explicit attitudes?"
A Two reasons why they don't match:
1) "a person may be unwilling to accurately report some attitude." or
2) "a person may be unable to accurately report an attitude."

So they are saying that it's because of flaws in measuring explicit bias, implying it isn't OUR test.

FAQ 15 is interesting:

https://implicit.harvard.edu...

it seems like they are dancing a jig around this one.



When I read over the above statements (and correct me if I'm wrong), there seems to an implied belief that the test is probably right and the "complainers" are probably wrong.

I'm not suggesting the test is perfect but yes, it is my belief that the test is probably showing us something significant. When a vast majority of people appear to demonstrate a bias that clearly fits with a history of oppression and racism, as well as current stereotypes in media and society, then I think it is reasonable to deduce that something is probably going on here.

Seems like the more rational approach would be to wonder about the study (since it is undeniable that a very high percentage of studies are flawed) and want to look at it further before drawing any conclusions:

I agree that it is rational to look in to how the study may be flawed, which you and I have both apparently done by checking out the FAQs and thinking them through for ourselves.
On the other hand, I think it is highly irrational to dismiss the results of a test because you don't like them on a personal level, rather than because you have discovered a legitimate flaw with the test.

Yea, deciding what that "something" IS can be problematic. IAT bias could be nothing more than all the beautiful white faces we see on TV all the time. They just look so DARN happy & friendly all the time!

Out of the 200 or so papers at the Implicit project, there are only a couple that are policy oriented. I've been meaning to take a look at them, but I'm always torn in a dozen different directions.

Thank you for taking the time to respond. I really appreciate it.