How easy it is to test for racial bias on a computer but how difficult to actually understand the human subconscious and what it portends in civil society when conditions arise to actually test a person reactions when they encounter different people of different races and conditions.
There are so many different scenarios that predict a persons racial tendencies under so many different condition that will tell a different story each time because of those conditions under which the occur. But to predict a person reaction to racial encounters on computer is nonsense if not dangerous!
There are so many different scenarios that predict a persons racial tendencies under so many different condition that will tell a different story each time because of those conditions under which the occur. But to predict a person reaction to racial encounters on computer is nonsense if not dangerous!
Where is the basic common sense here?
This Test Claims to Show Whether You’re Unconsciously Racist. Here Are Its Problems.
Elizabeth Slattery
/
@EHSlattery
/
Are you secretly a racist? Do you harbor prejudices deep within your subconscious? Could you be bigoted without even knowing it? These are highly charged accusations, but a team of social psychologists developed a test claiming to uncover such unconscious biases. Known as the Implicit Association Test, it measures bias based on how quickly test-takers respond to images and words that flash before them on a computer screen.
Proponents of the test claim the results signal “discriminatory behavior.” But does this test predict real-world behavior? And should lawmakers and decision-makers in academia and corporate America take steps to remedy these unconscious biases?
Here’s how the test works: Test-takers are instructed to press one key when a picture of a black person appears on the screen, and another key when there’s a picture of a white person. Then you do the same with positive and negative words (“magnificent,” “delightful,” “rotten,” “hurtful,” and so on).
You repeat this exercise, but with photos and words (so presumably, black people will be paired with negative words and white people with positive words). Then the test flips the race that is paired with the two types of words. The difference in your response times for associating one race with positive or negative words indicates whether you have a slight, moderate, or strong bias.
You can join the millions of people who have measured their bias, and take the test here. There are tests measuring several types of bias—based on race, sexuality, weight, and religion, among others.
Read more about the Implicit Association Test in this new Heritage Foundation Special Report.
For the record, I took the race test twice and got conflicting results. One time, it concluded that I have a “slight preference” for black people, and a second time, I got the opposite result with a slight preference for white people.
According to the test, my bias depends on whether or not I’ve had my afternoon coffee. My casual and highly unscientific foray into the world of unconscious bias shows the imprecise nature of the test.
While the test promises to identify biases, one of its original proponents, Brian Nosek, has admitted there’s “substantial risk for both falsely identifying people as eventual discriminators and failing to identify people who will discriminate.” Warnings like this, however, haven’t stopped the media, academia, and corporate America from jumping on the unconscious bias train.
The legal community is following suit. The American Bar Association now offers “Elimination of Bias” credits for some of its continuing legal education courses, and there have even been lawsuits based on claims of unconscious bias.
Writing in the Harvard Business Review, one of the test’s creators, Mahzarin Banaji, boldly claimed, “Most of us believe that we are ethical and unbiased. … But more than two decades of research confirms that, in reality, most of us fall woefully short of our inflated self-perception.”
She explained in a Washington Post interview, “The Implicit Association Test measures the thumbprint of the culture on our minds. … [I]t is picking up that aspect of the culture that has gotten into your brain and mind.” And apparently that culture is overwhelmingly anti-black: Data from the online test shows that 88 percent of white test-takers had a “pro-white or anti-black implicit bias.”
Putting aside the images of Orwellian thought police that this calls to mind, there’s a growing consensus among researchers that the Implicit Association Test falls woefully short.
Before it becomes further engrained in hiring, college admissions, law enforcement, banking, government contracting, and many other industries, policymakers and cultural leaders should consider the broader “social and political implications of this research,” as a new Heritage Foundation report argues.
Join us on Thursday at The Heritage Foundation for a public event on the Implicit Association Test.
While the intention behind the test—ensuring that everyone is treated fairly by seeking to root out racism—is a good one, statistician Althea Nagai explains, “[T]here is consensus on neither unconscious racism nor the [Implicit Association Test].”
Though the test’s proponents claim “[it] is to psychology what Galileo’s telescope was to the Copernican Revolution,” Nagai identifies several flaws—including the test’s unreliability, the lack of clarity about what it actually measures, its failure to predict real-world behavior, and the high rate of false positives.
Taken together, these flaws make the case that incorporating the test into public policy, the legal system, or ethics training in corporate America is not only ineffective, but may also exacerbate race relations.
Nagai explains:
False accusations of racism are highly likely, and true instances of racism lose their salience. The real difficulty is the public cynicism and indifference that results when accusations are made, new policies are implemented, and millions of dollars are spent on the problem—with little perceived progress.
She continues:
Although it has been hailed by the media as uncovering a dark, secret side of the American psyche, numerous critics of the [test] have demonstrated that it simply cannot predict how test-takers will act in the real world.
So while the media may enjoy running stories with gimmicky, sensational headlines like “Across America, Whites Are Biased and Don’t Even Know It,” “Is Everybody a Little Bit Racist?” and “You May Be More Racist Than You Think,” there’s mounting evidence that the Implicit Association Test is just that: a gimmick that can’t deliver what it promises.
Are you secretly a racist? Do you harbor prejudices deep within your subconscious? Could you be bigoted without even knowing it? These are highly charged accusations, but a team of social psychologists developed a test claiming to uncover such unconscious biases. Known as the Implicit Association Test, it measures bias based on how quickly test-takers respond to images and words that flash before them on a computer screen.
Proponents of the test claim the results signal “discriminatory behavior.” But does this test predict real-world behavior? And should lawmakers and decision-makers in academia and corporate America take steps to remedy these unconscious biases?
Here’s how the test works: Test-takers are instructed to press one key when a picture of a black person appears on the screen, and another key when there’s a picture of a white person. Then you do the same with positive and negative words (“magnificent,” “delightful,” “rotten,” “hurtful,” and so on).
You repeat this exercise, but with photos and words (so presumably, black people will be paired with negative words and white people with positive words). Then the test flips the race that is paired with the two types of words. The difference in your response times for associating one race with positive or negative words indicates whether you have a slight, moderate, or strong bias.
You can join the millions of people who have measured their bias, and take the test here. There are tests measuring several types of bias—based on race, sexuality, weight, and religion, among others.
Read more about the Implicit Association Test in this new Heritage Foundation Special Report.
For the record, I took the race test twice and got conflicting results. One time, it concluded that I have a “slight preference” for black people, and a second time, I got the opposite result with a slight preference for white people.
According to the test, my bias depends on whether or not I’ve had my afternoon coffee. My casual and highly unscientific foray into the world of unconscious bias shows the imprecise nature of the test.
While the test promises to identify biases, one of its original proponents, Brian Nosek, has admitted there’s “substantial risk for both falsely identifying people as eventual discriminators and failing to identify people who will discriminate.” Warnings like this, however, haven’t stopped the media, academia, and corporate America from jumping on the unconscious bias train.
The legal community is following suit. The American Bar Association now offers “Elimination of Bias” credits for some of its continuing legal education courses, and there have even been lawsuits based on claims of unconscious bias.
Writing in the Harvard Business Review, one of the test’s creators, Mahzarin Banaji, boldly claimed, “Most of us believe that we are ethical and unbiased. … But more than two decades of research confirms that, in reality, most of us fall woefully short of our inflated self-perception.”
She explained in a Washington Post interview, “The Implicit Association Test measures the thumbprint of the culture on our minds. … [I]t is picking up that aspect of the culture that has gotten into your brain and mind.” And apparently that culture is overwhelmingly anti-black: Data from the online test shows that 88 percent of white test-takers had a “pro-white or anti-black implicit bias.”
Putting aside the images of Orwellian thought police that this calls to mind, there’s a growing consensus among researchers that the Implicit Association Test falls woefully short.
Before it becomes further engrained in hiring, college admissions, law enforcement, banking, government contracting, and many other industries, policymakers and cultural leaders should consider the broader “social and political implications of this research,” as a new Heritage Foundation report argues.
Join us on Thursday at The Heritage Foundation for a public event on the Implicit Association Test.
While the intention behind the test—ensuring that everyone is treated fairly by seeking to root out racism—is a good one, statistician Althea Nagai explains, “[T]here is consensus on neither unconscious racism nor the [Implicit Association Test].”
Though the test’s proponents claim “[it] is to psychology what Galileo’s telescope was to the Copernican Revolution,” Nagai identifies several flaws—including the test’s unreliability, the lack of clarity about what it actually measures, its failure to predict real-world behavior, and the high rate of false positives.
Taken together, these flaws make the case that incorporating the test into public policy, the legal system, or ethics training in corporate America is not only ineffective, but may also exacerbate race relations.
Nagai explains:
False accusations of racism are highly likely, and true instances of racism lose their salience. The real difficulty is the public cynicism and indifference that results when accusations are made, new policies are implemented, and millions of dollars are spent on the problem—with little perceived progress.
She continues:
Although it has been hailed by the media as uncovering a dark, secret side of the American psyche, numerous critics of the [test] have demonstrated that it simply cannot predict how test-takers will act in the real world.
So while the media may enjoy running stories with gimmicky, sensational headlines like “Across America, Whites Are Biased and Don’t Even Know It,” “Is Everybody a Little Bit Racist?” and “You May Be More Racist Than You Think,” there’s mounting evidence that the Implicit Association Test is just that: a gimmick that can’t deliver what it promises.
No comments:
Post a Comment