It may in fact make things worse.

>The Implicit Association Test has been used in online studies to assess implicit racial attitudes in over seven million participants. Although typically used as an assessment measure, results from four pre-registered experiments (N = 940) demonstrated that completing a Race IAT exacerbates the negative implicit attitudes that it seeks to assess. Increases in White participants’ negative automatic racial evaluations of Black people were observed across two different implicit measures (SC-IAT and AMP) but did not generalize to another measure of automatic racial bias (Shooter Bias task). Results highlight an important caveat for the Race IAT, but also for many other forms of psychological assessment: that by measuring, we often perturb the system that we wish to understand.

https://psyarxiv.com/vxsj7/

This might be an aspect of the same problem that DARE had. If you castigate some act but at the same time speak in a way that says that act is common then the effect of making it seem normal will tend to trump the disapproval.

>the effect of making it seem normal will tend to trump the disapproval.

I think it may be more due to the fact that people have a natural disinclination towards people pretending to be moral authorities, especially when their premise is revealed to be based on lies and pseudoscience

There were aspects of DARE that were based on Psuedoscience but lots of major and minor campaigns, from ending dueling to getting people not to steal things from the Petrified Forest National Park, have been derailed by this effect without involving lies or pseudoscience.

The problem is that the people suggesting the better ways are usually... promoting rather unpalatable ideas. Often either that there is no discrimination in the workplace ever and so we should stop talking about it, or that discrimination in the workplace is in fact normal and healthy.

I'd love to see a way of creating healthy, non-discriminatory workplaces, especially in industries that are usually pretty terrible.

>or that discrimination in the workplace is in fact normal and healthy

One thing that bothers me is that discrimination is seen as a four letter word, when it is something we do all the time and celebrate. Only specific forms are considered bad and we need to be more explicit on what and why.

For example, racial discrimination is bad. Discrimination based on what degree one holds is not. Or, given that not everyone has the same level of access to earn a degree, should we be so open in accepting such discrimination?

What about discriminating based on punctuality? It seems an easy one to justify, but there is a significant cultural impact on punctuality.

What about discrimination on factors that people can't control and that have no relation to work, but which are currently legally allowed. For example look at the correlation between height and pay. Should that be allowed and what efforts are we doing to stop it. And does it become more of a concern when we account for average height differences between genders and the gendered outcome that can result as a secondary result (paying short people less, given that women are on average shorter than men, results in paying women less).

These sort of conversations never seem to happen when it comes to discrimination in the work place, making me think the name of the issue doesn't accurately reflect what the acted upon concern is.

Enforcing culture is highly discriminatory. You have to talk, dress, and act a certain way within cultural boundaries, otherwise you're discriminated against.

Imagine you're in a job interview. You describe the position to a candidate, and the candidate says "Wow, that sounds cool."

You describe the position to a different candidate. That person says "Shxt, that's tight dog."

Does one sound more professional to you? Why? Are you going to make any inferences about that persons capability to be a good employee based on how they speak? Of course you are. Is that right? Probably not.

All norms reflect a certain kinds of power.

We seem to be in a period where norms are assumed to be malicious or harmful, where the majority has a moral obligation to minimize the advantages of following norms.

At bottom, though, that's just an attack on all norms.

A more honest framing is that the issue is disagreement over what discrimination means, not that it can't exist or is good. Probably the most contentious issue is representation itself. Can a group of people be overrepresented or underrepresented in an organization without the organization being discriminatory? James Damore made the case that it's probable. And what happened to him shows that it isn't about "unpalatable ideas", it's about avoiding even having the discussion.

Is it unpalatable to ask questions about how much discrimination exists and against whom it is directed? The only “evidence” I’m aware of are implicit association tests and Goldin’s study of blind auditions, neither of which demonstrate discrimination. And asking this question often just gets me downvoted. If we are so certain that there is discrimination and tons of it, why is the real evidence treated as a closely guarded secret while we proudly circulate the debunked evidence? Please note that this is the opposite of “we should not talk about it”.

I think you are getting downvoted because the argument is exasperating. It's argument by ignorance: "I don't know therefore it isn't true." It is very easy to find ample evidence of discrimination documented in published peer-reviewed studies. E.g https://www.nber.org/digest/sep03/w9873.html "The results indicate large racial differences in callback rates to a phone line with a voice mailbox attached and a message recorded by someone of the appropriate race and gender. Job applicants with white names needed to send about 10 resumes to get one callback; those with African-American names needed to send around 15 resumes to get one callback. This would suggest either employer prejudice or employer perception that race signals lower productivity."

The other exasperating thing is at presenting evidence almost never work. The step after someone has presented evidence of discrimination is for the person who doesn't "believe in discrimination" to challenge the result. Either the result doesn't actually prove discrimination or the study is flawed in some way so you can't draw any conclusions.

> I think you are getting downvoted because the argument is exasperating. It's argument by ignorance: "I don't know therefore it isn't true."

That would be unfortunate, because I'm very clearly not saying any such thing. Specifically, asking about the magnitude of discrimination is not asserting that it doesn't exist at all. For what it's worth (hopefully very little), I'm a True Believer in discrimination--it exists, I've experienced it. I'm skeptical about specific truth claims about discrimination (notably claims that it explains a significant part of any gap in the US). I'm also skeptical about social science's ability to accurately study discrimination due to its political homogeneity and track record (more on this below).

> It is very easy to find ample evidence of discrimination documented in published peer-reviewed studies.

Undoubtedly. It's easy to find peer-reviewed studies that support almost any popular conclusion; however, I'm not looking to confirm my own biases (nor anyone else's) but rather to understand what's really going on. Notably, neither the IAT study and Goldin's orchestra study should have passed peer review and yet they did and were held up as exemplary for decades. Between issues like these and the social science replication crisis, I don't put much stock in individual studies.

> The other exasperating thing is at presenting evidence almost never work. The step after someone has presented evidence of discrimination is for the person who doesn't "believe in discrimination" to challenge the result. Either the result doesn't actually prove discrimination or the study is flawed in some way so you can't draw any conclusions.

Discrimination is indeed difficult to prove; the reasonable response isn't to be frustrated that others remain skeptical after being shown problematic evidence, but rather to question why you are compelled by said problematic evidence. Skepticism is the reasonable position here. Note again that skepticism about various discrimination claims isn't the same thing as asserting the opposite of those claims. Note also that skepticism about various discrimination claims doesn't mean that we need to stop investigating discrimination; it merely means we don't have conclusive evidence (and consequently, we probably shouldn't be drafting policy that could hurt people or otherwise propagating myths about the conclusiveness of said evidence).

It's not exclusive to anti-bias "sensitivity training". It's pretty much all the industry of corporate training seminars. "Team-building", bunk "psychology tests"... You know the type. You probably had to endure them. They cost tens of thousands of dollars each, they're completely useless, but still, the managers with more budget than sense buy them, and it's a billion dollar industry.

I think you misinterpret why this is being done. It's not to train employees.

It's a hedge against accusations of racism.

So management can point at this and say: see? We did all we could!

Bingo. Not just a hedge against accusations but even a hedge against civil liability.

A similar story plays out in academic settings for Title IX training.

There is actually lots of good evidence that familiarity with unfamiliar ethnic groups and exposure to different cultural priorities makes people less prone to conflict and bad reactions.

I've worked at a lot of places, and I've yet to see an anti-bias training that was more than awareness-raising. I've never ever heard any suggest more radical ideas (such as restorative justice).

While we might argue that current D&I initiatives are not working at a pace we'd prefer, I do think the literature strongly suggest that the method of informed exposure and consideration for issues other than your own tends to work. What's more, these programs are extremely inexpensive compared to more radical restructuring of companies (or the often pie-in-the-sky suggestions of folks who will die on the hill of "it is a pipeline problem").

Some examples of not-too-old research on how simple conversations and exposure can change attitudes:

https://science.sciencemag.org/content/346/6215/1366

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2705986/

https://www.sciencedirect.com/science/article/pii/B978008088...

P.S., it's worth noting that if you haven't scanned the scientific literature carefully, you're committing the same error as the folks who fell for this goofy color theory scam by putting your own biases and intuition above the scientific method.

Well, it's a good way for management to show they did ... something.

And that's not necessarily as bad a thing as it sounds.

Racism, sexism, etc. are hard problems. You're talking about deeply embedded, pervasive messages. Of course they won't be solved by throwing a class at it. But shrugging your shoulders and saying, "Yeah, but we're not gonna do anything" becomes yet another one of those pervasive messages.

So it can't, and won't, fix the immediate problem. But it at least acknowledges the problem, rather than leaving it buried.

That's not to say it can't also be surrounded by cynicism, scamming, and bad faith. Those things are yet more indicators of the real underlying issues: people are willing to accept or ignore the racism, sexism, etc. that happen to other people because it's not important to them. Making the costs of it more visible doesn't reduce those costs, but it does provide a reason for people to start wanting to lower them.

This. I believe that's the main reason those tests exist. They are used by management as a legal/PR protection.

Discrimination training certainly has a function in reminding people of what they are prohibited from doing (legally or by policy), because then they can't claim ignorance of it later. It also has the benefit of reminding people to watch their own behaviour.

The training is largely for legal purposes. If someone needs to be fired for whatever it has to be demonstrated that they were 'trained' on the basis of that issue.

It depends on what you mean. About 25 years ago in Sweden the leading party the Social Democrats instituted a policy championed by feminists that meant that every other elected position should be occupied by a woman. Research has shown that that policy has led to more qualified candidates across the political spectrum.

There is also evidence from Sweden that attitudes towards rape and sexual assault has changed due to government policies. E.g questions such as "Is it ok to have sex with a girl who is too drunk to speak?" are answered differently today than 20 years ago.

Edit: The research is here: https://wwwsnsse.cdn.triggerfish.cloud/uploads/2017/05/sns-a... More qualified means higher level of education and more years of experience. Same metrics as used in most workplaces.

Also, I wonder why it only took someone five minutes to ask me for a source, but no one has yet to ask tenpies for a source... Some facts are easier to believe than others?

If affirmative action has such clear evidence that it works, why do not the Swedish government implement it as a general rule for all their employees? Sweden has a very high gender segregation rate, around 85% of people that has a full time employment do so in a gender segregated profession, and government work places tend to rank at the very top as the most gender segregated work places.

If it leads to more qualified employees, which I assume would correlates with their suitability, it would save money and reduce mistakes in key areas like healthcare which then would result in saved lives. A national policy would be a win-win for everyone.

A while back I looked if any municipality had a general hireing policy in order to combat gender segregation. So far I have yet to find one that even mention gender segregation as something to consider when hiring new people. There does not even seem to be one that something as vague as a goal to reduce gender segregation.

An internal policy for the list of candidates on their ballot, stating that every other candidate must be a woman.

This is not the same as the list of who eventually gets elected of course, as the probability of getting elected decreases the further down the list you go. Also the policy appears to apply per ballot, and since there are something like 30 regions with (possibly) independent ballots, there is still some room to stack one way or the other. In the event, last election they got a pretty even split: 48 vs 52.

For ballots for municipal elections, the candidate's order rarely changes due to person voting. Effectively, the order set by the party is the order people are elected in, barring exceptional circumstances (someone being very popular or unpopular).

Party list are one of the reasons that PR is not the silver bullet for democracy

It wasn't until 2018 that nonconsensual sex was classed as rape in Sweden (1). Which is somewhat surprising that it took that long given the political climate you outline.

(1) https://en.wikipedia.org/wiki/Rape_in_Sweden

>Research has shown...more qualified candidates...

First, can you link a source? Second, what does "more qualified" mean in the context of politicians? They are selected by popularity in the majority; capabilities are downstream from likeability. This, one could argue that "more qualified" could be construed as "excellent hair" or "large breasts." This is not to say that the job of political official does not have constraints or optimums, but the method of job placement is the issue in terms of my question.

The actual term used is "kompetent", and they talk about the general competency of the individual, and assume this correlates with their suitability as politicians. They then show that their measure of competency correlates positively with municipalities having better economy and better perceived service by the citizens. They also get more votes and things like that. The measure is basically if the person makes more than average within their peer group.

Report is here: https://www.sns.se/aktuellt/sns-analys-nr-41-konskvotering-k...

It is so clearly a con game and I consider it abuse to subject anyone to it.

Not only does it do nothing, but the purported problem—implicit association tests (IATs) “demonstrating” a high degree of racism in an organization or individual—is artificial: the IATs are themselves fraudulent; there is no evidence whatsoever that they measure racism at all. https://www.thecut.com/2017/01/psychologys-racism-measuring-...

EDIT: Downvoters, what's wrong with this comment? Are downvotes disagreement? If so, do you have anything to substantiate your position? If it's not disagreement, what could I do to improve the quality of my post to make it acceptable?

Exactly! I had 'unconcious bias' training at work, one of the learning outcomes of which was to understand the 'scientific basis' for unconcious bias. After reviewing the material and finding myself somewhat dubious of its claims, I - as a good scientist working in a technical role in a technical company - went off to research for myself. Turns out the supposed scientific basis is deeply flawed. And yet, this training is being rolled out in many companies across the UK: training based on nonsense. What's deeply insulting - and worrying - is that this should be pushed on educated, intelligent people whose jobs in no small part depend on sniffing out bullshit. Do they think we won't find out the truth? Are the management so clueless/compromised that they actually believe in it? IDK.

I think the right way to think about this is that psychology has the imprimatur of legitimate science (and some pscyhological research is indeed legitimate), but a lot of it is more about having an idea you like, conducting a few inadequate studies, and publishing them in a domain journal with little critical review and no reproduction.

Then, the folks who teach these courses see these papers (presumably in their undergraduate psych classes?) and end up believing them, becacuse they don't have the necessary scientific analytic/critical skills to evaluate them.

At my employer, the IAT was originally presented as scientific fact (with citations). Over time, the citations changed from "This is great!" to "This work appears to not be scientifically accurate, but we like the idea" to "we will not be using this, unless you want to try it voluntarily".

I think in general most people in society don't really have the ability to differentiate bullshit science from real science. I've actually left my scientific field permanently because I've seen far too enough examples of fraud and incompetence leading to tenure track positions.

> I think in general most people in society don't really have the ability to differentiate bullshit science from real science.

I would say in many ways they can't, at least not without significant time investment every time they need to make that assessment, which is often.

A domain expert is likely able to identify the smell of bullshit in their field extremely quickly. An expert in an adjacent domain will be able to apply some of their expertise, but not all, and sniff out some problems, but other might elude them. That is, a physicist might be able to note some likely problematic parts of statements to do with biology based somewhat on the science, somewhat on the shared methodology, and somewhat on the shared politics they might experience.

A layperson may have very little experience to draw upon for some scientific statement, and even when they have experience with the underlying principles (a general interest in the area, experience in college), that may not be enough to identify problem areas. That's not a problem we can expect the layperson to be responsible for correcting (at least not beyond a healthy skepticism), unless we expect them all to increase their knowledge in every domain that they might make an assessment about (this happens, but it happens at a societal level, and over the span of many years).

Psychology seems to share relatively little with most hard sciences, so it's particularly hard for both laypeople and scientists in other fields to assess without effort (as evidenced here though the multiple stories of people that put in that actual effort).

I'm not sure we'll see improvement on this issue until we have better norms about communicating results. There's a natural inflation that happens when different parties take a finding and explain it, and possibly over-emphasize portions of it. That this happens twice now, once with the PR department associated with the group doing the study and once by the media that goes by what the PR department say, means that the average person gets a very distorted view of what's actually going on.

"bias training doesn't eliminate bias" isn't the same thing as "bias doesn't exist".

There's no evidence, and certainly not a plausible model, that supports/explains how reaction times reflect attitudes, and how changes in attitude change reaction times.

It really has nothing to do with science at all. It's just another way to punish the majority for crimes of their ancestors, and to forcibly give power to a minority for simply existing and being a certain race or gender.

Nobody really wants equality. They want supremacy. If they did have equality, they would be treated like everyone else and lose all of their power/special treatment.

They are also tools of our politicians to be a permanent victim class, so the politicians can dole out justice and get votes.

If it were truly scientific we could test out and get certified, then we would only have to attend refresher courses.

I'm serious when I say this is one of the best things about being a contractor or consultant. I can just sign a paper saying that I'm responsible for my behavior.

I would immediately change your workplace if you are able to do so.

Understanding the "scientific basis" of something unscientific is a ritual. It's 2+2=5, a test of your willingness to say things that are obviously false when that's asked of you by those above you.

Let me ask you something: Is unconscious bias training the worst example of bullshit you have experienced in your professional life or the thing that has aggravated you the most? Does that tell you anything about your unconscious biases?

The point of the training is for companies to absolve themselves of culpability. If you act like a jerk after unconscious bias training you are on your own and the company hopes that it can't be blamed for it.

Please don't get personal about this. It only provokes others into doing worse, as happened downthread.

Your comment would be just fine with just the second paragraph.

https://news.ycombinator.com/newsguidelines.html

The second paragraph is the useful part of the comment. You seem to be reacting to the presence of the word "jerk", entirely ignoring the context.

Yes, we agree. The comment would be fine with just the second paragraph.

> Is unconscious bias training the worst example of bullshit you have experienced in your professional life or the thing that has aggravated you the most?

I can't see the parent making this claim at all; are you sure that your interpretation is in good faith?

They might be unconsciously biased.

I didn't act like a jerk before the training, never mind after :)

I had training called 'cultural awareness' or similar at another company which was basically very similar material, but without the bogus claims that it was rooted in science. It's the latter that I object to - if they are going to enforce behavioural standards just be honest about it, rather than gaslighting people with this arrant nonsense that we're all unknowing bigots through some kind of original sin.

The bullshit I've had to endure during my career includes "You're one of our best paid engineers, so don't share salary information with your colleagues.", "NDA means you can not take a job at a competing company for six months after your resignation." and "It is important that your back is towards me (the manager) so that I can see your monitor and ensure that you are working at all times."

My point is that even if unconscious bias training is 100% bullshit, it is among the most benign forms of business bullshit I've experienced. Thus, I cannot understand why it is so upsetting to so many.

> it is among the most benign forms of business bullshit I've experienced

I'd argue the opposite actually - it's insidious, because any objection is liable to get you labelled a racist etc., and it's dangerous, because it's effectively telling you that no matter how hard you try you cannot escape being a racist (etc.). The latter is pretty much a textbook definition of gaslighting. I don't want my employer gaslighting me.

What you describe is not bullshit, it's just straight up dishonesty and bullying.

Bullshit is distinguished by the speaker not even caring if it's true. Usually it is meant to distract you from what you are not supposed to notice.

Maybe it’s worth asking what other types of “bs” training they received at work.

Maybe it’s one among many or maybe it’s the only one. We don’t know.

That said, most training is about absolving the company of responsibility and putting it on you (security training for example).

On the surface it would seem a good idea to train against harassment at work, but I’d be keen to know if it has an effect on behavior or is it the HR consequences that affect behavior more than the training, for example.

If my job required me to go to a unconscious bias training I would treat it the same is if they forced me to go to church. I am sure companies would like it if everyone would treat each other as kin and share the religious experience, and the actually hours itself would unlikely be the worst experience compared to some of the worst days at the office, but I would still resign if they did it. Call it irrationality. I would simply no longer feel that there were any mutual trust and thus I would not feel safe in continuing working there.

I wonder how many companies can survive long term if employees don't trust their employer and the employer don't trust its employees.

I've seen studies[1][2] which indicate that unconscious racial bias training increases racist attitudes among those who are put through it. If that's true, unconscious bias training might be worse than bullshit, it might be actively harmful and counterproductive to its stated goals.

1 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6524213/

2 https://journals.sagepub.com/doi/pdf/10.1177/095001701771984...

Ah we're making a generation of Michael Scott Racists.

> Does that tell you anything about your unconscious biases?

There is so much to unpack about you here

> Yet we’re supposed to believe that we can do even better through different methods, and become even more successful than Apple and Amazon etc.

This is a prime example of why diversity training is needed. People are deeply biased in regards to seeing something successful and then thinking that is the formula to success. It's textbook survivorship bias. Furthermore, it completely misses the material basis for why these companies would have such hiring patterns: there were higher rates of asian and white college-educated men in CS, but, as hiring pools change, so do hiring patterns.

There isn't some innate ability to found software companies in white or asian men.

Yes, aspirants pattern-match from success and make tweaks they deem essential to their goals. That‘s the history of human endeavor. Already you see female software engineering hiring growing in priority in the companies that are struggling with hiring.

Doesn't the existence of women lead teams and initiatives throughout the history of the discipline of computer science suggest that you're deliberately ignoring lots of examples and distorting your filter based on ideological motivations?

It sure looks that way from here.

No, women are certainly capable of great software engineering the same way that men named Steve are capable of being great football players (Steve Young, Steve Largent). But since it’s clearly possible to build great software without many women, and win Super Bowls without Steve’s, If I’m a football coach I’m not going to start a Steve Initiative where I recruit players named Steve.

... wait. Didn't you just argue against broad pattern matching and then use logic that supported it?

Given your reasoning, why shouldn't someone note the plethora of successful Steve's and say, "Our team needs more Steves! Look at my pattern matching! Steves work!"

I have not been led to the same conclusion (we can balance out workforce racial population) by my unconscious bias training, I have been led to conclusions that I find quite reasonable (watch for these thought patterns that indicate racial bias, such as if the person is X then the person is also Y, also be sure to keep questions to things that are essential to the job role). We might be taking different diversity training.

Are you guys talking about things like this?

"here comes Karen, let me discriminate by noticing how I might discriminate so to make her more comfortable because I can't perceive my privilege but now I can alter my behavior just around her"

Isn't there some utility in keeping the workplace productive just by being sensitive, as opposed to quantifying if it "reduces discrimination"

No. Horoscopes are not real, while various biases are. Please stay out of the teaching profession!

Personal attacks will get you banned here. Please review https://news.ycombinator.com/newsguidelines.html and stick to the rules when commenting.

https://news.ycombinator.com/newsguidelines.html