#implicit association test

By: Mahzarin Banaji and Frank Dobbin

Published: Sep 17, 2023

At least 30 states are considering legislation to defund DEI initiatives in public universities and state agencies. At the same time, conservative activists, emboldened by the Supreme Court’s ruling against affirmative action in college admissions, are suing companies to stop DEI initiatives. These challenges come on the heels of the growth of corporate DEI programs after the murder of George Floyd in May of 2020.  

Meanwhile, advocates for DEI—which stands for diversity, equity and inclusion—have bemoaned the fact that after decades of diversity training, many university faculties, state agencies and corporations have made little progress on diversifying the workforce.

Are the right and the left on the same page here—is diversity training a hopeless cause?

We are a psychologist and a sociologist who have been studying bias and organizational diversity programs, respectively, for decades. The research makes it clear that Americans desperately need education about bias, because even people who value fairness and equality hold biases—without being aware of it. They need to understand that bias operates systemically and must be addressed at the individual, institutional and societal levels.

Education offered on these matters is very much in the national spirit. As Alexis de Tocqueville noted in “Democracy in America” in 1835: “The greatness of America lies not in being more enlightened than any other nation, but rather in her ability to repair her faults.”

What research shows

The social and behavioral sciences have developed strong evidence about conscious prejudice and implicit bias. Three lines of research, together, are pertinent. One provides good news. As our colleague Larry Bobo has documented, conscious unabashed racial prejudice has fallen consistently since the 1960s. White Americans today largely believe in racial equality.

This isn’t to say that explicit expressions of prejudice have evaporated; in fact they pop up with surprising regularity. The pandemic witnessed precipitous increases in anti-Asian hatred, and according to the Anti-Defamation League, instances of antisemitism are at a record high. 

A second line of research shows that less conscious, or implicit, bias has declined more slowly. Bias against some groups has barely budged. If only explicit values and biases drove discrimination, unfair treatment of, say, Black workers would be low. But implicit bias taints employer behavior and decisions. Our colleague Mandy Palais and collaborators find, for instance, that implicit racial bias in grocery-store managers still influences worker performance.

A third line of research uses audit studies, in which matched Black and white people, for instance, apply to the same job. Who gets called in for an interview, or hired? Scores of studies show discrimination by race, ethnicity, gender and disability. These studies show, among other things, that white applicants are about 50% more likely than identical Black applicants to be called back for an interview or offered a job. Other audit studies show discrimination in real-world access to financial resources, healthcare and treatment by the law and law enforcement.    

Research by Lincoln Quillian and colleagues compares the results of audit studies over time, finding that discrimination against Black job applicants is virtually unchanged from a generation ago. And the economist Raj Chetty and colleagues not only show a shocking drop in American upward mobility over time, but also show that in regions with high levels of implicit bias, Black Americans are less likely than white Americans to move up the economic ladder.

Research by Lincoln Quillian and colleagues compares the results of audit studies over time, finding that discrimination against Black job applicants is virtually unchanged from a generation ago. And the economist Raj Chetty and colleagues not only show a shocking drop in American upward mobility over time, but also show that in regions with high levels of implicit bias, Black Americans are less likely than white Americans to move up the economic ladder. 

Research from one of us, Frank Dobbin (with Alexandra Kalev), meanwhile, shows how likely a worker in a U.S. firm is to have a management job, by group. Women and people of color see increases until the mid-1980s. But progress stalls for Black and Hispanic workers after that. Men from those groups make no progress between then and 2021, and women make almost no progress. We clearly have more work to do to equalize opportunity.

Falling short

It’s not hard to conclude from all these studies that we are not the land of opportunity for everyone we claim to be. An enlightened society should see that education about the prevalence of discrimination is imperative. In fact, it would be downright dumb not to educate people. 

But, as Dobbin and Kalev have shown, the typical DEI training doesn’t educate people about bias and may even do harm.

Most training programs fall short on two fronts. First, they use implicit-bias education to shame trainees for holding stereotypes. Trainers play gotcha, sending trainees to take an online test co-developed by one of us, Mahzarin Banaji, for education and research. Instead of training people about research that finds that bias is pervasive, trainers use the test to prove to trainees that they are morally flawed. People leave feeling guilty for holding biases that conflict with American values.

“Gotcha” isn’t going to win people over. The approach is disrespectful, and misses the main takeaway from implicit bias research: Everyone holds biases they don’t control as a consequence of a lifetime of exposure to societal inequality, the media and the arts. Trainers should introduce these ideas with humility, for trainers themselves can’t help but hold these very biases. They could easily educate themselves about the implicit bias research with resources at outsmartingimplicitbias.org.

The second problem with most trainings is that they seek to solve the problem of bias by invoking the law to scare people about the risk of letting bias go unchecked. Trainers recount stories of big companies brought to their heels by discrimination suits. They detail rigid do’s and don’ts for hiring, disciplining and firing people. They require trainees to pass tests on what the law forbids. All of this makes it clear that the CEO approved the training solely to avoid litigation. Trainees leave scared that they will be punished for a simple mistake that may land their company in court.  

Trainings with this one-two punch—you are biased and the law will get you—backfire. The research shows that this kind of training leads to reductions in women and people of color in management.

Why would diversity training actually make things worse? Making people feel ashamed can lead them to reject the message. Thus people often leave diversity training feeling angry and with greater animosity toward other groups (“There’s no way I’m biased!”). And threats of punishment, by the law in this case, typically lead to psychological “reactance” whereby people reject the desired behavior (“Nobody’s telling me what I can’t say!”). This kind of training can turn off even supporters of equal-opportunity programs. 

A better way

It doesn’t have to be this way, and Dobbin and Kalev’s research on training points to a better alternative. Instead of using legal scare tactics, training programs should give managers a way to counter biases—namely, training in strategies for cultural inclusion. This kind of training teaches skills in listening, observation and intervention. It thus helps managers to hear employee concerns, notice when workers are feeling shunned or dissed, and intervene. It also offers skills for starting tough conversations about how to treat colleagues at work.

Those are skills from Management 101, but managers often don’t want to hear bad news, so they don’t ask employees about troubles, watch teams for signs of bullying, or speak up when they sense a problem. Reminding managers that they can use these tools to suss out problems and nip them in the bud helps them to feel capable of managing biases and microaggressions. When managers use these skills, they retain women and people of color for long enough to come up for promotion. That’s how good diversity training can boost diversity. Unfortunately, only about a quarter of diversity trainings emphasize cultural inclusion. 

Moreover, if training succeeds in conveying the findings from bias research—that bias is unseen but pervasive—it can build support for wider systemic changes designed to tear down obstacles to equal opportunity. In that sense, training isn’t designed to blame people for their moral failings. Instead, it’s galvanizing them to support organizational change by arming them with knowledge.

In the end, DEI training can’t squelch implicit bias; nothing short of changing people’s life experiences can do that. But when done right, implicit-bias education can alert students to the fact that people committed to equality nonetheless hold biases. And that knowledge can, in turn, motivate them to reshape their workplaces to counter discrimination by democratizing key parts of the career system.

That means extending recruitment visits from Harvard to Howard; offering mentors to each and every worker; and inviting all employees to nominate themselves for skill and management training programs. It means offering work-life supports to people up and down the ladder. Each of these changes has been shown to produce significant increases in managerial diversity.

The lesson here, the one that should be at the core of DEI training, is that implicit bias resides in individuals, but it resides in organizational career systems as well. And fixing those systems is as simple as democratizing them.

Mahzarin Banaji is a professor of psychology at Harvard University and co-author of “Blindspot: Hidden Biases of Good People.” Frank Dobbin is a professor of sociology at Harvard University and co-author of “Getting to Diversity: What Works and What Doesn’t.” They can be reached at [email protected].

[ Via: https://archive.is/0D4kV ]

==

--

Virtuous Lie: 1619 Project (subthread):

==

I've posted on most if not all of these before, and yet I wonder how many people still take the "truth" of these as articles of faith, regardless of what evidence is prevented.

By: Te-Ping Chen and  Lauren Weber

Published: Jul 21, 2023

Two years ago chief diversity officers were some of the hottest hires into executive ranks. Now, they increasingly feel left out in the cold.

Companies including Netflix, Disney and Warner Bros. Discovery have recently said that high-profile diversity, equity and inclusion executives will be leaving their jobs. Thousands of diversity-focused workers have been laid off since last year, and some companies are scaling back racial justice commitments.

Diversity, equity and inclusion—or DEI—jobs were put in the crosshairs after many companies started re-examining their executive ranks during the tech sector’s shake out last fall. Some chief diversity officers say their work is facing additional scrutiny since the Supreme Court struck down affirmative action in college admissions and companies brace for potential legal challenges. DEI work has also become a political target.

“There’s a combination of grief, being very tired, and being, in some cases, overwhelmed,” says Miriam Warren, chief diversity officer for Yelp, of the challenges facing executives in the field.

In interviews, current and former chief diversity officers said company executives at times didn’t want to change hiring or promotion processes, despite initially telling CDOs they were hired to improve the talent pipeline. The quick about-face shows company enthusiasm for diversity initiatives hasn’t always proved durable, leaving some diversity officers now questioning their career path. 

In the wake of George Floyd’s murder in police custody in May 2020, companies scrambled to hire chief diversity officers, changing the face of the C-suite. In 2018, less than half the companies in the S&P 500 employed someone in the role, and by 2022 three out four companies had created a position, according to a study from Russell Reynolds, an executive search firm. 

Once mostly tasked with HR matters, today’s diversity leaders are expected to weigh in on new product development, marketing efforts and current events that have an impact on how workers and consumers are feeling. Warren and other CDOs said the expanded remit is playing out in a politically divided environment where corporate diversity efforts are the subject of frequent social-media firestorms. 

Falling demand

New analysis from employment data provider Live Data Technologies shows that chief diversity officers have been more vulnerable to layoffs than their human resources counterparts, experiencing 40% higher turnover. Their job searches are also taking longer. 

“I got to 300 applications and then I stopped tracking,” says Stephanie Lubin, who was laid off from her role as diversity head at Drizly, an online alcohol marketplace, in May following the company’s acquisition by Uber. In one case, Lubin says she went through 16 rounds of interviews for a role she didn’t get, and says she is now planning to pivot out of DEI work.

The number of CDO searches is down 75% in the past year, says Jason Hanold, chief executive of Hanold Associates Executive Search, which works with Fortune 100 companies to recruit HR and DEI executives, among other roles. Demand is the lowest he has seen in his 30 years of recruiting.

At the same time, he says, more executives are feeling skittish about taking on diversity roles.

“They’re telling us, the only way I want to go into another role with DEI is if it includes something else,” he says of the requests for broader titles that offer more responsibilities and resources. He estimates that 60% of diversity roles he is currently filling combine the title with another position, such as chief human resources officer, up from about 10% five years ago.

During the pandemic, some companies moved people into diversity leadership if they were an ethnic minority, says Dani Monroe, even when they weren’t qualified. Monroe served as CDO for Mass General Brigham, a Boston-based hospital system and one of the largest employers in the state, until 2021 and convenes a yearly gathering of more than 100 CDOs.

“These were knee-jerk reactions,” she says of the hurried CDO hires, adding that some of those elevations didn’t create much impact, leaving both sides feeling disillusioned.

On-the-job obstruction 

American workers are split on the importance of a diverse workforce, surveys find. 

Diversity chiefs also encounter obstruction from top executives, says Melinda Starbird, a human resources and diversity executive who has worked at AT&T, Starbucks and OfferUp, an online marketplace. Leaders sometimes associate diversity efforts with mandates, such as the equal-employment rules that apply to federal contractors. Those requirements for compliance can create executive resistance that bleeds over into other cultural or policy shifts, such as adding Juneteenth as a company holiday, she says. 

“Even if you report to the CEO, it’s still a battle and it’s a smaller budget,” says Starbird, who was laid off from OfferUp in November during a broader restructuring. 

Many diversity executives feel a lack of buy-in from their colleagues. In a survey of 138 diversity executives conducted this spring by World 50 Group, a networking organization for corporate leaders, 82% said they had sufficient influence to do their job, down 6 percentage points from 2022. Asked if they felt supported by middle managers, 41% said yes, an 8-percentage-point drop.

Since the Supreme Court overturned affirmative action in June, companies are anticipating spillover legal action could have an impact on them. Those that are still hiring CDOs want people who can help the board navigate the political and legal landscape of diversity work and figure out how to take defensive moves to shield them from litigation, says Tina Shah Paikeday, global leader of Russell Reynolds’s diversity, equity and inclusion practice. 

“They recognize it would be smart to get ahead of that.”

People are more resistant to company-backed efforts to advance diversity when they are worried about their own jobs, whether because of impending layoffs or disruptions from AI, says David Kenny, chief executive of Nielsen, the media-ratings company. 

Kenny was both CEO and CDO for a time, taking on the diversity role to emphasize how important it was to the future of the business. Even as CEO, it could be a tough sell. Efforts to restructure compensation to make it more equitable created a backlash.

“A lot of it is, ‘I’m losing my slice of the pie,’ ” he says.

[ Via: https://archive.vn/jHRFo ]

==

The grift is over.

There seems to be a built-in implication that much of the movements around DIE in the last few years have been performative: organizations making the approved signals to keep the puritans at bay. Perhaps they've now figured out that these measures are, at best, unable to demonstrate their efficacy, or at worst, anti-productive. The number of DIE programs that can or even will quantify or demonstrate their effectiveness with metrics and data can be counted on one hand; the truly fanatical ones will scold you for even suggesting that you should. Or more likely, perhaps they've figured out that as an insurance policy, the impact to the bottom line is no longer worth the investment; throwing buckets of money to purchase indulgences during a moral-religious panic might have made sense in 2020, but not so much in 2023.

Study after study reveals that none of this social snakeoil - from the phrenology of "implicit bias training" to the Maoist struggle sessions of "white fragility training" - actually help, and reliably make things worse by making everyone fixate on identity politics rather than doing anything productive. Meaning DIE is nothing but expensive and destructive virtue signaling. If you want to destroy an organization from the inside, there's no better way than embracing DIE.

You're far better off sticking to your core telos, supported by liberal ethics like equal opportunity, colorblindness and the ideal of meritocracy. Or more formally, Merit, Fairness and Equality (MFE). Whatever results you get from a fair process are inherently fair.

"Diversity" in particular is always about superficiality and thinly-veiled racism, while "equity" requires someone in authority to artificially create preferred outcomes (establishing the perfect conditions for an authoritarian), rather than a system of fairly and consistently applied rules (equality).

I can name five people, men and women, where I work who have different ethic ancestry, who grew up within 40 miles of each other and have the same local accent.

And I can name five white men who grew up on four different continents with three different first languages, who have worked for over a dozen different organizations, from multi-national companies to military to non-profits to education institutions before immigrating.

By: Marilyn Singleton

Published: Feb 22, 2023

Marilyn Singleton is a board-certified anesthesiologist and a visiting fellow at the medical advocacy organization Do No Harm.

I’ll never forget my parents’ reaction when I was accepted to the University of California at San Francisco’s medical school. Having attended segregated schools, my mother and father were thrilled that their daughter would attend a fully integrated, top-tier institution.

When I graduated with a medical degree in 1973, a Black woman in a class of mostly White men, there was a real sense that the days of obsessing over skin color and making race-based assumptions about our fellow human beings was finally fading — and, hopefully, soon gone for good.

Apparently not. That racial obsession has come rushing back — in academia, politics, business and even in my beloved medical profession. But now it’s coming from the opposite direction. The malignant false assumption that Black people are inherently inferior intellectually has been traded in for the malignant false assumption that White people are inherently racist.

That is the basic message conveyed by “implicit bias training,” which is now mandatory for California physicians; it is a message that I believe is harmful both to physicians and patients. There is a sad irony in all this, because the misguided focus on racism is intended to improve the health and well-being of Black patients in particular.

The law, which took effect last year, includes other bias targets, including gender identity, age and disability. But in practice, such training — a mainstay of the diversity and inclusion industry, worth an estimated $3.4 billion in 2020 — is overwhelmingly about race.

In California, where I’ve been licensed since 1974, every physician is required by law to participate in this racially regressive practice. Doctors must take implicit bias training not just once but as part of the curriculum of “continuing medical education,” for at least 50 hours every two years, required for their medical license renewal.

The training’s focus is on exactly what the name suggests: deeply ingrained prejudice toward people of different races. There is no room for debate, for the law states baldly: “Implicit bias, meaning the attitudes or internalized stereotypes that affect our perceptions, actions, and decisions in an unconscious manner, exists.”

And the law asserts as fact that implicit bias is responsible for “racial and ethnic disparities in health care,” particularly for Black women.

I am so disturbed by the state’s mandate that, so far, I have balked at the training. But I know that I must comply before the end of 2023 if I wish to have my medical license renewed.

Many of my friends and colleagues ask why I’m so upset by the law. Clearly, implicit bias training isn’t meant for me. It’s aimed at White people, who are far and away the biggest share of the medical profession. My answer is simple. I reject the unscientific accusation that people are defined by their race, not by their individual beliefs and choices. It is little consolation that studies are finding implicit bias training has no effect on its intended targets, and might even make matters worse.

Think about the message this mandate sends to Black physicians. It suggests that I should be wary of my White colleagues because, after all, they’re biased against people like me. Sure, they can undergo frequent training, but their bias is always going to be there, beneath the surface, threatening to rear its ugly, racist head. Collegiality and collaboration — two essential components of high-quality medical care — are targeted by this mandate. Call that an implicit bias.

Since I became a physician, I have seen exactly one instance of racism in health care — and it was from a patient, not a fellow physician. As for my colleagues, I have been consistently impressed with the conscientious, individualized care they have provided to patients of every race and culture. When we all took our oath to “first, do no harm,” we meant it, and we live it. I can’t imagine spending my entire career thinking my peers can’t uphold that oath without constant racial reeducation.

The message to physicians is bad enough, but the message to patients is much worse. Black people are, in effect, being told that White physicians are likely to quite literally damage our health. If that’s the case, why on earth would you seek medical care, unless you could be absolutely certain of not being treated by a White physician? And if you do seek medical care, why wouldn’t you doubt every word from a White doctor who is inherently prejudiced against you?

The whole point of implicit bias training is to create better health outcomes for Black patients and others who might be the target of discrimination, but the opposite seems more likely. It fosters a climate of distrust and resentment that threatens to undermine the medical and moral progress I’ve seen over the decades. When I graduated from medical school, we were moving past the era of racial obsession and anger. Why are we going back to the days when race defined so many lives and dimmed so many futures?

[ Via: https://archive.is/dAu17 ]

==

Reminder: The virtue theater of “implicit bias training” is as pseudoscientific as homeopathy, phrenology, thetan cleansing and demon possession. Even the creators admit that it is:

https://faculty.washington.edu/agg/pdf/Greenwald,Banaji&Nosek.JPSP.2015.pdf#page=5

“problematic to use [it] to classify persons as likely to engage in discrimination,” and “attempts to diagnostically use such measures for individuals risk undesirably high rates of erroneous classifications.”

https://musaalgharbi.com/2020/09/16/diversity-important-related-training-terrible/

https://dailynous.com/2017/01/12/reconsidering-implicit-bias/

https://www.freedomcast.us/iopod/episode1

https://www.thecut.com/2017/01/psychologys-racism-measuring-tool-isnt-up-to-the-job.html

https://www.psychologytoday.com/us/blog/rabble-rouser/202203/12-reasons-be-skeptical-common-claims-about-implicit-bias

https://static1.squarespace.com/static/5186d08fe4b065e39b45b91e/t/601766eb09286e6fda49d962/1612146411820/PaluckPoratClarkGreen_2020.pdf

https://xlab.berkeley.edu/connect/Intersectional%20Implicit%20Bias%20Pre-Print.pdf

https://nymag.com/intelligencer/2017/12/iat-behavior-problem.html

https://pubmed.ncbi.nlm.nih.gov/25402677/

The only things it reliably does is signals virtue, creates an obsessive fixation on categories, divides co-workers, and makes snake-oil sellers billions of dollars in an industry that, like the church, sells an imaginary product.

By: Jesse Singal

Published: Dec 5, 2017

At the moment, you may have heard, the field of psychology is grappling with a so-called “replication crisis.” That is, certain findings that everyone had assumed to be true can’t be replicated in follow-up experiments, suggesting the original findings were the result not of actual psychological phenomena, but of various flawed methodologies and biases that have crept into the scientific process.

One of the major contributing factors to the replication crisis, which is centered mostly on social psychology, is human nature. Humans, being humans, do not like hearing that ideas they’ve worked on for a long time might have to get tossed in the bin, or at the very least revised significantly. That’s why some researchers — though by no means all of them — have responded to good-faith critiques of their work by attempting to derail the conversation, calling their critics crazy or mean or attributing to them dark ulterior motives. The researchers who attempt such derailings tend to be established, well-respected ones who have benefited from the old regime — the regime that led the field into its current, precarious situation, and which is now threatened by a growing reform movement.

The implicit association test, co-created by Harvard University psychology chair Mahzarin Banaji and University of Washington researcher Anthony Greenwald, is an excellent example. Banaji and Greenwald claim that the IAT, a brief exercise in which one sits down at a computer and responds to various stimuli, measures unconscious bias and therefore real-world behavior. If you score highly on a so-called black-white IAT, for example, that suggests you will act in a more biased manner toward a black person than a white person. Many social psychologists view the IAT, which you can take on Harvard University’s website, as a revolutionary achievement, and in the 20 years since its introduction it has become both the focal point of an entire subfield of research and a mainstay of diversity trainings all over the country. That’s partly because Banaji, Greenwald, and the test’s other proponents have made a series of outsize claims about its importance for fighting racism and inequality.

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.

One striking thing about the process of reporting that article was the extent to which Banaji tried to smear her critics, suggesting to me in an email she believed that critiques of the test could be explained by the fact that the IAT “scares people who say things like ‘Look, the water fountains are desegregated, what’s your problem.’” She also accused the test’s critics of having a “pathological focus” on black-white race relations and the black-white IAT for reasons that “will need to be dealt with by them in the presence of their psychotherapists or church leaders.”

This is the definition of a derailing tactic — shift the focus from critiques of the IAT itself, some of which in this case appeared in a flagship social-psych journal, to the ostensible moral and psychological failings of the critiquers.

A couple days ago, Quartz published its own article on the IAT, by Olivia Goldhill. The article covers similar ground and comes to similar conclusions as mine, and adds some new insights and analysis: The headline, “The world is relying on a flawed psychological test to fight racism,” captures things pithily. Goldhill’s piece clearly shows that Banaji and Greenwald are still trying to deflect and derail rather than fully engage with the process of evaluating their test:

It’s highly plausible that the scientists who created the IAT, and now ardently defend it, believe their work will change the world for the better. Banaji sent me an email from a former student that compared her to Ta-Nehisi Coates, Bryan Stevenson, and Michelle Alexander “in elucidating the corrosive and terrifying vestiges of white supremacy in America.” || Greenwald explicitly discouraged me from writing this article. “Debates about scientific interpretation belong in scientific journals, not popular press,” he wrote. Banaji, Greenwald, and Nosek all declined to talk on the phone about their work, but answered most of my questions by email.

The idea that journalists shouldn’t write about scientific controversies would have been highly questionable even before the replication crisis exploded onto the scene, but it’s hard to fathom why anyone would take this argument seriously in 2017. After all, the replication crisis was spurred in part by opaque research and peer-review processes, by people not sharing data, by social and professional structures that sometimes had the effect of short-circuiting real debate about the merits of ideas — particularly popular ones of the sort that often get glowing write-ups in, well, the “popular press” (Greenwald, of course, doesn’t appear to have any problems with positive coverage of the IAT). Journalism, when it’s done well, can serve as a useful check on all these tendencies. To be fair, Greenwald isn’t the only one who thinks that science should only be critiqued by those very close to a given controversy — this is an idea that seems to sometimes pop up among defenders of the old, deeply flawed social-psychological ways — but that isn’t how things should work.

Even more surprising, though, is an email Greenwald wrote to Goldhill which read, “The IAT can be used to select people who would be less likely than others to engage in discriminatory behavior.” This might come across as a fairly banal defense of his research project, but it isn’t: It’s the continuation of a very slippery pattern I identified in my article.

As I noted, in their 2013 best seller Blindspot, which helped the IAT carve out an even bigger place in the public imagination than it had already achieved, Banaji and Greenwald wrote that the test “predicts discriminatory behavior even among research participants who earnestly (and, we believe, honestly) espouse egalitarian beliefs,” and “has been shown, reliably and repeatedly” to do so. In fact, this is a “clearly … established” “empirical truth.” But then, just two years later, they argued in an academic paper unlikely to be read by the general public that due to the test’s methodological weaknesses, it is “problematic to use [it] to classify persons as likely to engage in discrimination,” and “attempts to diagnostically use such measures for individuals risk undesirably high rates of erroneous classifications.”

I referred to this as a “Schrödinger’s test” situation in which the test both does and doesn’t predict behavior at the same time. When the test’s creators are addressing lay audiences unfamiliar with its problems, it does predict behavior; when they’re addressing academic audiences familiar with what is now a years-long controversy, they acknowledge that it doesn’t. Greenwald’s quote to Goldhill just marks the latest example.

In other words:

Banaji and Greenwald in 2013, to the public: Our test has been shown, reliably and repeatedly, to predict behavior.

Banaji and Greenwald in 2015, to academics: Our test doesn’t predict behavior.

Greenwald in 2017, to the public: Our test predicts behavior.

So, once more: I disagree with Greenwald. Society desperately needs more open scrutiny of scientific claims, not less, whether in scientific journals, the media, or anywhere else. Especially when it comes to claims that seem to change every two years.

==

“tHe IaT iS bAsEd On ScIeNcE!!1!”

No, it’s based on ideology, and perpetuated by a multi-billion dollar church of DEI through faith by priests whose careers don’t exist without it.

https://www.thecut.com/2017/01/psychologys-racism-measuring-tool-isnt-up-to-the-job.html

https://qz.com/1144504/the-world-is-relying-on-a-flawed-psychological-test-to-fight-racism/

https://musaalgharbi.com/2020/09/16/diversity-important-related-training-terrible/

https://www.vox.com/identities/2017/3/7/14637626/implicit-association-test-racism

https://www.psychologytoday.com/us/blog/rabble-rouser/202203/12-reasons-be-skeptical-common-claims-about-implicit-bias

DOI: https://psycnet.apa.org/doi/10.1037/pspa0000314

Pre-Print: https://xlab.berkeley.edu/connect/Intersectional%20Implicit%20Bias%20Pre-Print.pdf

Intersectional Implicit Bias: Evidence for Asymmetrically Compounding Bias and the Predominance of Target Gender

Abstract

Little is known about implicit evaluations of complex, multiply categorizable social targets. Across five studies (N = 5,204), we investigated implicit evaluations of targets varying in race, gender, social class, and age. Overall, the largest and most consistent evaluative bias was pro-women/anti-men bias, followed br smaller but nonetheless consistent pro-upper-class/anti-lower-class biases. By contrast, we observed less consistent effects of targets' race, no effects of targets' age, and no consistent interactions between target-level categories. An integrative data analysis highlighted a number of moderating factors, but a stable pro-women/anti-men and pro-upper-class/anti-lower-class bias across demographic groups. Overall, these results suggest that implicit biases compound across multiple categories asymmetrically, with a dominant category (here, gender) largely driving evaluations, and ancillary categories (here, social class and race) exerting relatively smaller additional effects. We discuss potential implications of this work for under standing how implicit biases operate in real-world social settings.

[...]

General Discussion

Implicit bias is central to the study of social cognition. Given that people are multiply categorizable, understanding the influences of such intersectionality upon implicit bias is likely to be vital for understanding its effects in everyday social contexts. In the present research, we examined implicit evaluations of multiply categorizable social targets, testing two competing theories about intersectional intergroup bias. We also developed and tested the reliability of a novel method of measuring and modelling implicit bias at the level of individual targets.

In Study 1 we observed implicit evaluations of Black and White males to be driven solely by targets’ social class, with bias favoring upper-class over lower-class targets. In Study 2, we measured implicit evaluations of targets varying in race, gender, social class, and age, and found results to be primarily driven by a specific positive bias favoring upper-class female targets. In Study 3, we used similarly intersectional targets, and explored the impact of portraying targets in full-body versus upper body photographs on implicit evaluations. Here, we observed effects of targets’ race, with Asian and White targets evaluated more positively than Black targets, and of targets’ social class, with upper-class targets evaluated more positively than lower-class targets (though only when targets were displayed in full-body presentation). Most striking, however, was the dominant effect of target gender, with positive/negative evaluations of female/male targets accounting for the majority of variance in implicit bias.

In Study 4 we tested the generalizability of these results by recruiting representative samples of US adults, and measuring implicit evaluations not just via ST-IATs, but also via EPTs and AMPs. Across all measures, we observed target gender to be the largest driver of implicit evaluations, though its dominance was less pronounced in EPTs and AMPs than in ST- IATs. We also again observed effects of targets’ social class and race, though the effect of race was inconsistent across tasks, with participants displaying anti-Black bias in the ST-IAT, pro- Asian bias in the EPT, and anti-White bias in the AMP. Finally, in Study 5 we conducted an integrative data analysis to test a number of potential moderating factors. Results showed that while all groups of participants displayed pro-female implicit gender bias and pro-upper-class implicit social class bias, both biases were stronger among women than men. Results also showed the effect of race varied across racial groups, with Asians displaying a preference for Asian over White and Black targets, Black participants displaying a preference for Asian and Black targets over White targets, Latinos displaying a preference for Asian over Black targets, and Whites displaying no significant racial bias.

The present work makes theoretical, empirical, and methodological contributions to the study of implicit evaluative bias toward multiply categorizable targets. On a theoretical level, we believe our results are best accounted for by a synthesis of compounding bias and category dominance approaches. Consistent with the category dominance model (Macrae et al., 1995), we observed a single social category to exert a dominant influence on implicit evaluations of intersectional targets in each of our studies. In Study 1, social class was dominant. In Studies 3 and 4, target gender was dominant. And even in Study 2, despite its more complex results, target gender still uniquely accounted for substantially more variation in Target D Scores than any other target-level predictor. These results are consistent with the notion that when faced with complex social stimuli, social perceivers act as ‘cognitive misers’, and make implicit evaluations that are strongly influenced by specific social categories, and are relatively unaffected by others.

However, our results are also consistent with the notion that implicit biases compound–– at least to some extent––across multiple categories at once. In Studies 3 and 4, which used the most tightly controlled set of targets, we observed simultaneous effects of targets’ gender, race, and social class. And in Study 5b, we found little evidence that these results represented separate groups of participants attending solely to each factor. So, although we found little consistent evidence for the kind of multiplicative interaction effects suggested by the multiple jeopardy/advantage hypothesis (Ransford, 1980), we did find evidence of biases compounding additively––albeit highly asymmetrically––across multiple social categories, with the most negative implicit evaluations consistently being made of targets displaying multiple intersecting stigmatized social identities (in this case, lower SES males), and the most positive implicit evaluations being made of individuals displaying multiple intersecting positively-valued social identities (in this case, upper SES females).

The overall picture emerging from the present work is therefore one of theoretical compromise. Implicit evaluative biases toward complex multiply categorizable targets do appear to compound across categories, but do so asymmetrically, with a dominant category (here, target gender) playing a leading role, less dominant categories (here, target race and social class) exerting relatively small additional effects, and peripheral categories (here, target age) having little detectable influence.

==

“The Patriarchy™” being all mysterious ways again.

This study demonstrates the flawed and ideological nature of present workplace “unconscious bias” trainings. They’re shallow and univariate, to the point where the results are not replicable.

If you’re being put through an IAT that does not control for multiple categories, such as gender, age, and social class, what you’re being subjected to is essentially ideological dowsing.

By: Lee Jussim

Published: Mar 28, 2022

KEY POINTS

There is no consensually-agreed upon definition of implicit bias. This makes communicating about implicit bias quite difficult.

The Implicit Association Test (IAT) is the most common method for measuring implicit bias. Yet it has several flaws and limitations.

Those limitations, which are well-known among psychological scientists, are rarely acknowledged to the wider public, including students.

Implicit bias is in the air. Hillary Clinton famously declared that "implicit bias is a problem for everyone." When she was California Attorney General, now-Vice President Kamala Harris expanded implicit bias trainings for police and has attributed many things to implicit biases on her Twitter feed.

Given widespread distress over unconscious racism, many consulting firms now provide implicit bias trainings and assure you that they deliver. A simple Google search for "consultant, implicit bias training" yields pages and pages of hits. Is this kind of corporate response to the problem of implicit bias justified?

In a 2018 essay, West Virginia University sociologist Jason Manning argued that "unconscious racism" bears a resemblance to older superstitions about the evil eye and sympathetic magic—but rather than mysterious unseen supernatural forces, we have mysterious unseen unconscious forces. There are, of course, some differences between beliefs in the evil eye and unconscious racism. There is no evidence that people can harm you by looking at you, whereas there is a wealth of studies on implicit social cognition. But I argue that the comparison may still be justified because the evidence for unconscious racism is so weak.

A moral panic occurs when some substantial portion of a society creates a "folk devil"—members of the community considered deviant and dangerous—and exaggerates the dangers they pose. In one research report, for example, people massively overestimated the number of unarmed Black people killed by police in 2019, with more than half of all liberals surveyed estimating the number at 1,000 or more (when data indicated it was more likely to be around 27). And many people have been denounced or socially ostracized for opposing things like affirmative action and diversity, equity, and inclusion programs, even though the majority of Americans, including majorities of Black and Hispanic respondents, oppose basing college admissions or hiring on ethnicity or race.

Others have been denounced or publicly shamed for a maladroit compliment of a supermodel or for sharing a peer reviewed sociology article, if doing so is viewed as deviant behavior or contrary to widely-held views. I personally have witnessed a talk in which a Jewish professor was publicly called a "grotesque Nazi" by physics professors for having described Obama as "half-black" (he has a Kenyan father and a White American mother). Such responses can be seen as evidence of a moral panic and the creation of "folk devils" surrounding the topic of racism.

Returning, however, to the implicit bias consulting firms—it's possible to find powerful testaments to the effectiveness of their trainings. But what does the science actually say—not just about the trainings, but about implicit bias more generally?

The workhorse method for assessing implicit bias is the Implicit Association Test (IAT). You can take several of these here to see what they are like for yourself. But before you do, you should consider these reasons to be skeptical about any claims about implicit bias.

1. The peer-reviewed scientific literature has witnessed a great walking back of many of the most dramatic claims made on the basis of the IAT and about implicit social cognition more generally.

Several researchers in this domain have put together this repository of over 40 articles (and growing) identifying flaws, artifacts, limitations, and threats to the validity of the IAT. Here are the titles of just a few of those articles:

"More Error than Attitude in Implicit Association Tests"

"Unconscious Racism: A Concept in Pursuit of a Measure"

"Implicit? What Do You Mean? A Comprehensive Review of the Delusive Implicitness Construct in Attitude Research"

In fairness, there are still plenty of research articles that take implicit bias seriously, such as this one. The point is not that the notion has been completely debunked or the IAT shown to be completely worthless; I even do research using the IAT! Instead, the point is that most of the most dramatic claims about it have been debunked or, at least, shown to be dubious and scientifically controversial. These are not firm grounds on which to sell the public on the power and prevalence of unconscious racism or trainings to mitigate it.

2. There is no consensually-accepted scientific definition of implicit bias.

Across five articles, there might be three with different definitions and two that do not even provide a definition. No one knows what "implicit bias" even is—at least if "know" is taken to mean "a clear understanding shared by most scientists."

3. The IAT measures reaction times, not things that most people think of as bias.

Whether reaction times might be considered bias in some technical sense is beyond the scope of this essay. Reaction times are not what most people think of when they think about bias; the IAT does not directly measure racism, oppression, or unfairness. (Whether it does so indirectly with any reasonable success is controversial, as described next.) To claim reaction times constitute any sort of bias in the common understanding of the meaning of "bias" is to import a conclusion by fiat rather than evidence.

4. At best, the IAT measures the strength of association of concepts in memory.

When reaction times are faster for some pairs of concepts than others, it is possible that the two concepts for which reaction times are faster are more closely associated in memory. That's the only thing the IAT can directly show: the closeness of association of concepts in memory—not "bias," not "racism." But it goes downhill from there; a slew of statistical issues and methodological artifacts suggest that the IAT is not even a clean measure of strength of association.

5. The IAT may capture prejudice, stereotypes, or attitudes to some degree, but, if so, it is not a clean measure.

Critiques of the IAT have concluded that it contains more error than attitude or reflects actual knowledge about actual group differences and conditions; and that IAT scores reflect four separate phenomena, of which attitude is just one.

6. The IAT, as used and reported, has a potpourri of additional methodological and statistical oddities.

Although a deep dive into them is beyond the scope of this blog, the interested reader can find them laid out in gory and technical details here and in the online repository of articles.

7. Many of the studies that use IAT scores to predict behavior find little or no anti-Black discrimination specifically.

If "unconscious racism" was as powerful and pervasive as its advocates claim, one would expect some, and probably a great deal, of anti-Black discrimination in these studies. Its absence in many studies justifies significant skepticism about claims about the power or prevalence of implicit bias.

8. Whether IAT scores predict behavioral manifestations of bias beyond self-report prejudice scales is unclear, with some studies finding they do and others finding they do not.

Racial prejudice is real and is readily measured by self-report scales assessing attitudes towards various racial and ethnic groups. It is not clear that the IAT captures much beyond these self-report scales, though.

9. Procedures that change IAT scores have failed to produce changes in discriminatory behavior.

One might ask, then, the point of such procedures and why is there such enthusiasm for these trainings.

10. There is currently no clear evidence that implicit bias trainings accomplish anything other than teaching people about the research on implicit bias.

It's not that there is "no evidence." This is not a case of "absence of evidence is not evidence of absence." It is a case of "there is tons of evidence, but most of it shows little or no effect."

11. There is no evidence that IAT scores are "unconscious."

Research finds that people are quite good at predicting their IAT scores.

12. Critiques and discussions of its limitations or weaknesses are often not presented when the IAT is taught to introductory psychology students. Gambrill & Reiman, 2011, defined propaganda as “encouraging beliefs and actions with the least thought possible” (p. e19516). Does extolling research on the IAT without presenting its limitations and weaknesses constitute "encouraging beliefs and actions with the least thought possible"? I leave that to you to decide. Regardless, this omission is contrary to the aims of science.

What This All Means

Here is my advice to you: Take an IAT or two (which you can here) if you have not already, just to see what the buzz is about. But now you are armed with enough information to reject any simple-minded proclamations about unconscious racism or the supposed power of implicit biases.

==