Implicit Bias Test

Discussion in 'General Discussion' started by Utumno, Feb 5, 2019.

  1. Utumno

    Utumno Administrator Staff Member

    Post Count:
    40,258
    Can't remember if this was posted here or if someone just linked an article which got me to websurfing that eventually led me here.

    I've only taken one test, and I thought it was pretty interesting. I'll be taking more when I get the time:

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

    Curious if anyone else has seen this or taken tests and have an opinion on whether or not they think it is accurate.
     
  2. Harper

    Harper encrypted account, pls don't sniff my packet

    Post Count:
    5,106
    theres nothing implicit about any of our biases
     
  3. Utumno

    Utumno Administrator Staff Member

    Post Count:
    40,258
    u say that but i think ur biased
     
  4. Kanmuk_Sealclubber

    Kanmuk_Sealclubber Yes

    Post Count:
    8,734
    The IAT measures your performance on the IAT, just like an IQ test measures your performance on an IQ test. Whether or not these constructs correlate strongly to vague conceptions of “intelligence” or “unconscious” bias is debatable. But it doesn’t really matter. What matters is whether or not they predict shit.

    Does the IAT predict shit? Yes, but seemingly not very well. It’s useful in the aggregate, but near useless as a predictor on an individual level. I wouldn’t take your results very seriously.
     
  5. Utumno

    Utumno Administrator Staff Member

    Post Count:
    40,258
    Good to know. I'm curious....

    could you expand on that a bit?

    I mean I for one am relieved because the test indicated I'm moderately racist Lol!
     
  6. Jackpanel

    Jackpanel TZT Abuser

    Post Count:
    6,766
    All Asians are racist. Its what I hate about them.
     
    Utumno likes this.
  7. Kanmuk_Sealclubber

    Kanmuk_Sealclubber Yes

    Post Count:
    8,734
    It predicts discriminatory behaviour fairly weakly. These are not actual numbers, I’m just using them to illustrate a point - someone who scores high on negative associations towards black people might be like 4% more likely to, say, hire a white job candidate over a black job candidate. That can be useful when looking at a million people, but not very useful in predicting how one person is going to behave.

    That’s obviously not quite how it works, but correlates are usually in like the .1 to .3 range iirc. I’ll look up an actual meta analysis in a bit.

    Edit - IAT tests are used for a lot of different shit. I’m talking about race or ethnic discrimination here specifically.

    It may have stronger correlates with other variables.
     
    Last edited: Feb 5, 2019
    Utumno likes this.
  8. Kanmuk_Sealclubber

    Kanmuk_Sealclubber Yes

    Post Count:
    8,734
    This is from somebody else’s essay I just quickly googled:

    https://lnu.se/globalassets/lmdswp201511.pdf
     
    Utumno likes this.
  9. Harper

    Harper encrypted account, pls don't sniff my packet

    Post Count:
    5,106
    what's the difference between pseudo and social science
     
    Samassi Abou likes this.
  10. Velox

    Velox TZT Abuser

    Post Count:
    5,652
    Pseudo science has little internal debate over how well theory matches empirical experiments, so this at least counts in their favor.
     
    Last edited: Feb 6, 2019
  11. Kanmuk_Sealclubber

    Kanmuk_Sealclubber Yes

    Post Count:
    8,734
    Pop reporting of psychology (from both reporters and the researchers themselves) is extremely pseudo scientific. Findings very rarely mean what they are purported to mean in mainstream media. “Take the IAT to find out if you’re a racist” is an example of that.

    That’s true of a lot of fields, but psych is particularly bad because the narratives associated with it are very interesting to the general public.

    But there is nothing inherently wrong with science having limitations. Unless we just don’t want to attempt to study complex things, we have to accept them. We just do a dogshit job of acknowledging them.
     
  12. Velox

    Velox TZT Abuser

    Post Count:
    5,652
    This is currently a big problem in "AI" also. Minor results can easily get blown out of proportion in the media (intentionally or not...). The field feels like 50/50 advertising vs science these days.
     
    Kanmuk_Sealclubber likes this.
  13. AgelessDrifter

    AgelessDrifter TZT Neckbeard Lord

    Post Count:
    43,699
    That feels like a charitable estimate to me
     
    Utumno likes this.
  14. Velox

    Velox TZT Abuser

    Post Count:
    5,652
    I'm a charitable person.
     
  15. Kilinitic

    Kilinitic 6,000 feet beyond man and time

    Post Count:
    16,060
    Not sure I’ve ever seen a fair media portrayal of newly published scientific research, etc
     
  16. Harper

    Harper encrypted account, pls don't sniff my packet

    Post Count:
    5,106
    the only thing i know for certain in life is that breakfast is the most important meal of the day
     
  17. Velox

    Velox TZT Abuser

    Post Count:
    5,652
    There was recently news coverage about a study purportedly claiming that breakfast should be skipped. Nothing is sacred any more.
     
  18. Red

    Red TZT Neckbeard

    Post Count:
    15,500
    Everyday we drift further from God's light
     
  19. Utumno

    Utumno Administrator Staff Member

    Post Count:
    40,258
    This was p informative, thank you.