Page 1 of 2 12 LastLast
Results 1 to 10 of 15

Thread: The Implicit Association Test

  1. Top | #1
    Contributor Trausti's Avatar
    Join Date
    Jul 2005
    Location
    Northwest
    Posts
    6,261
    Archived
    372
    Total Posts
    6,633
    Rep Power
    62

    The Implicit Association Test

    Can we please be rid of this voodoo nonsense?

    The problem, as I showed in a lengthy rundown of the many, many problems with the test published this past January, is that there’s very little evidence to support that claim that the IAT meaningfully predicts anything. In fact, the test is riddled with statistical problems — problems severe enough that it’s fair to ask whether it is effectively “misdiagnosing” the millions of people who have taken it, the vast majority of whom are likely unaware of its very serious shortcomings. There’s now solid research published in a top journal strongly suggesting the test cannot even meaningfully predict individual behavior. And if the test can’t predict individual behavior, it’s unclear exactly what it does do or why it should be the center of so many conversations and programs geared at fighting racism.
    The Creators of the Implicit Association Test Should Get Their Story Straight

  2. Top | #2
    Super Moderator
    Join Date
    Sep 2000
    Location
    Nevada
    Posts
    27,148
    Archived
    96,752
    Total Posts
    123,900
    Rep Power
    100
    It shows people are racist, that's all that matters. Whether it tells the truth or not is irrelevant.

  3. Top | #3
    Intergalactic Villainess Angry Floof's Avatar
    Join Date
    Jul 2008
    Location
    Lancre
    Posts
    9,879
    Archived
    14,435
    Total Posts
    24,314
    Rep Power
    62
    It does usefully measure subconscious bias. It's the "predicts real world behaviors" part that is questionable.

    In my mind, the best purpose these timed response tests is personal self reflection. It's a mistake, and possibly a dangerous one, to use it to judge or label people.
    The Authoritarians

    GOP and Trump supporters will not be able to say they didn't know. Vote in numbers too big to manipulate.

  4. Top | #4
    Super Moderator
    Join Date
    Sep 2000
    Location
    Nevada
    Posts
    27,148
    Archived
    96,752
    Total Posts
    123,900
    Rep Power
    100
    Quote Originally Posted by Angry Floof View Post
    It does usefully measure subconscious bias. It's the "predicts real world behaviors" part that is questionable.

    In my mind, the best purpose these timed response tests is personal self reflection. It's a mistake, and possibly a dangerous one, to use it to judge or label people.
    It is categorically not of any use on an individual basis.

    It's easier to learn a pattern than to erase a pattern and learn a new one. Thus one's times on the first part will be better than one's times on the second part.

    Of course you can make people think it's showing racism by doing the white-good pairing before the white-bad pairing.

  5. Top | #5
    Intergalactic Villainess Angry Floof's Avatar
    Join Date
    Jul 2008
    Location
    Lancre
    Posts
    9,879
    Archived
    14,435
    Total Posts
    24,314
    Rep Power
    62
    Whether you erase a pattern or learn a new one or spin one out of dingleberries, someone's got to be paying some conscious attention whatever patterns are there to begin with. The human capacity to observe our own thoughts and feelings and patterns is useful like that.

    You certainly don't need this test to do that. It's just a tool. We're largely invisible to ourselves. Anything that offers new perspective on our own habits is potentially useful. Even feedback from people who are just being assholes to us can be useful. A test that illuminates our own reactions to things in some way doesn't need to be a truth teller or dictator. It's just measuring response time to images and concepts. Whatever else anyone wants to try to add to that, this much is at least true. How is that not useful?
    The Authoritarians

    GOP and Trump supporters will not be able to say they didn't know. Vote in numbers too big to manipulate.

  6. Top | #6
    Mazzie Daius fromderinside's Avatar
    Join Date
    Oct 2008
    Location
    Oregon's westernmost
    Posts
    12,527
    Archived
    18,213
    Total Posts
    30,740
    Rep Power
    57
    Quote Originally Posted by Trausti View Post
    Can we please be rid of this voodoo nonsense?

    The problem, as I showed in a lengthy rundown of the many, many problems with the test published this past January, is that there’s very little evidence to support that claim that the IAT meaningfully predicts anything. In fact, the test is riddled with statistical problems — problems severe enough that it’s fair to ask whether it is effectively “misdiagnosing” the millions of people who have taken it, the vast majority of whom are likely unaware of its very serious shortcomings. There’s now solid research published in a top journal strongly suggesting the test cannot even meaningfully predict individual behavior. And if the test can’t predict individual behavior, it’s unclear exactly what it does do or why it should be the center of so many conversations and programs geared at fighting racism.
    The Creators of the Implicit Association Test Should Get Their Story Straight
    Interesting. Timely, Thanks.

    That 'scientists' would use before replicating or validating is a gross methodological sin in general and an even greater injustice to society and the academic community in particular.

    Why not just go back to gabbing over the fence?

  7. Top | #7
    Super Moderator
    Join Date
    Sep 2000
    Location
    Nevada
    Posts
    27,148
    Archived
    96,752
    Total Posts
    123,900
    Rep Power
    100
    Quote Originally Posted by Angry Floof View Post
    Whether you erase a pattern or learn a new one or spin one out of dingleberries, someone's got to be paying some conscious attention whatever patterns are there to begin with. The human capacity to observe our own thoughts and feelings and patterns is useful like that.

    You certainly don't need this test to do that. It's just a tool. We're largely invisible to ourselves. Anything that offers new perspective on our own habits is potentially useful. Even feedback from people who are just being assholes to us can be useful. A test that illuminates our own reactions to things in some way doesn't need to be a truth teller or dictator. It's just measuring response time to images and concepts. Whatever else anyone wants to try to add to that, this much is at least true. How is that not useful?
    The thing is it's basic human nature that you'll do the first part faster than the second part. Trying to draw conclusions from this about how one feels about the things is stupid.

    Put two cups in the microwave. One holds tap water, the other holds ice water & cubes. Should we conclude the microwave doesn't like ice because it's slower to boil? Drawing an individual conclusion from the IAT is about as intelligent.

  8. Top | #8
    Intergalactic Villainess Angry Floof's Avatar
    Join Date
    Jul 2008
    Location
    Lancre
    Posts
    9,879
    Archived
    14,435
    Total Posts
    24,314
    Rep Power
    62
    Quote Originally Posted by Loren Pechtel View Post
    Quote Originally Posted by Angry Floof View Post
    Whether you erase a pattern or learn a new one or spin one out of dingleberries, someone's got to be paying some conscious attention whatever patterns are there to begin with. The human capacity to observe our own thoughts and feelings and patterns is useful like that.

    You certainly don't need this test to do that. It's just a tool. We're largely invisible to ourselves. Anything that offers new perspective on our own habits is potentially useful. Even feedback from people who are just being assholes to us can be useful. A test that illuminates our own reactions to things in some way doesn't need to be a truth teller or dictator. It's just measuring response time to images and concepts. Whatever else anyone wants to try to add to that, this much is at least true. How is that not useful?
    The thing is it's basic human nature that you'll do the first part faster than the second part. Trying to draw conclusions from this about how one feels about the things is stupid.

    Put two cups in the microwave. One holds tap water, the other holds ice water & cubes. Should we conclude the microwave doesn't like ice because it's slower to boil? Drawing an individual conclusion from the IAT is about as intelligent.
    I think we are disagreeing about the "drawing a conclusion" part.

    Also, self reflection is not high priority for everyone.
    The Authoritarians

    GOP and Trump supporters will not be able to say they didn't know. Vote in numbers too big to manipulate.

  9. Top | #9
    Contributor
    Join Date
    Dec 2014
    Location
    Whale's Vagina
    Posts
    5,188
    Rep Power
    28
    Quote Originally Posted by Angry Floof View Post
    It does usefully measure subconscious bias.
    No it doesn't. It measures miniscule millisecond differences in how long it takes to process the semantic links between stimuli when presented in an artificial context where the stimuli have no meaning and when the processing in done to perform a completely meaningless task of button pressing of zero consequence.
    Nothing beyond that can be concluded.

    Claiming that is a "bias" is an unscientific interpretation and makes "bias" a meaningless string of letters.

    Also, the test directly asks questions about how one feels about black and white people, prior to performing the supposed "implicit association" part. That introduces so many possible sources of variance, making the tiny millisecond differences even less meaningful.
    For example, answering those questions not only could bring to mind your own feelings about different races, or simply your own awareness of how other people feel about it and about cultural level racism, how often you have heard racist remarks in your life or just recently (no matter how you feel about those remarks), or about how concerned you are if others think you are racist, etc.. Each of these are completely different, uncorrelated variables and how much they come to mind will differ greatly for each person.
    Yet, any one of them could easily impact the results of the IAT. That means the exact same score could mean any infinite number of things. The leader of the KKK could get the same score as a frequent victim of racism or an objectively non-racist person who is simply knowledgeable about the extent of racism in society. That makes the results worse than meaningless. It makes them almost certainly misleading and pointing to wrong conclusions about a person.

    Basically, using the IAT as a measure of anything is equal to measuring one's persons height and the length of another person's name, then comparing those two numbers as if they tell you how those people differ.

  10. Top | #10
    Contributor repoman's Avatar
    Join Date
    Aug 2001
    Location
    Seattle, WA
    Posts
    6,921
    Archived
    2,280
    Total Posts
    9,201
    Rep Power
    75
    Have they done these IATs with puppies, happy children and attractive people (all races) versus snarling animals and other scary or ugly people or things?

Similar Threads

  1. Implicit Social Cognition
    By ruby sparks in forum Social Science
    Replies: 2
    Last Post: 12-22-2017, 10:29 AM
  2. Evidence of Implicit Gender Bias
    By lpetrich in forum Political Discussions
    Replies: 28
    Last Post: 10-27-2017, 03:25 AM
  3. Evidence of Implicit Racial and Ethnic Stereotyping
    By lpetrich in forum Political Discussions
    Replies: 5
    Last Post: 09-22-2017, 05:51 PM
  4. Replies: 16
    Last Post: 01-16-2017, 05:51 PM
  5. National Implicit Sexism - Debate One
    By Jimmy Higgins in forum US Presidential Politics
    Replies: 65
    Last Post: 10-07-2016, 10:00 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •