Teaching yourself about structural racism will improve your machine learning

I think beyond not feeding race in as a <i>feature</i> to any model, this stuff is mostly nonsense. If you include race as a feature, then I think it&#x27;s likely that the model will become racist, because race is so highly correlated to behaviors and patterns which are in large part the consequence of all sorts of things, including historical racism, that a model could easily mistake race as a causal factor. If you don&#x27;t feed race in as a feature, however, the outputs are hardly racist. My impression has been that by and large the argument <i>actually</i> being made is that &quot;we have been trying to correct for historical injustices by actively using race and gender as mechanisms for advantaging minorities and women, and an unbiased model is not properly accounting for these particular objectives.&quot;<p>Take something like a bank loan. If you had a model at a bank which took credit score, income, wealth, and collateral into account, black Americans would have loans rejected at a higher rate than white Americans. Is this model racist? No, this model doesn&#x27;t even know what race is, all it knows is credit scores, income, wealth, and collateral. Does the fact that black Americans used to be slaves in the US, or were kept out of certain housing markets, contribute towards the fact that black Americans, on average, have lower credit scores, income, wealth, and collateral? Of course. But is this model racist? Literally not at all. It is completely unbiased, and exactly what the model <i>should</i> be. If the case you&#x27;re making is that you think that there should be a national effort to correct for historical injustices that were done by the state by actively discriminating by race, that is a <i>completely</i> different discussion.<p>Having all of our decision-making apparatuses factor in the infinite pile of historical injustices that may have contributed to an individual&#x27;s particular circumstances is not the way to go. Keep models simple and limited to what is relevant for that particular criteria. Fix injustices further upstream, or you make the whole system a convoluted nightmare.

I did research briefly in a lab studying heart rate variability. HRV is a really interesting statistic in predicting heart health outcomes, and very interesting is the racial difference in HRV.<p>Basically, African Americans exhibit much higher heart rate variability, meaning their nervous system is much quicker to react to stimuli (quicker time to fight or flight response, for example) and this still isn’t well understood in the field.<p>A naive understanding is that racial physiology is just different. And plenty of people will stand by this. However, self reported stress scores offer some insight into the difference.<p>High stress African Americans with High HRV lived as long as Low stress, low HRV White&#x2F;Asian Americans. Most likely, the process by which the nervous system regulates itself is heavily influenced by life course events.<p>Medical science, in my experience, lacks in quantifying these social factors, and too often underplays their significance in determining physiological differences. Humans are incredibly dynamic systems, and the case can be made that we adapt to stimuli in order to survive. It’s certainly possible that the physiological difference we observe in different racial populations is due to survival based on this principal.<p>It’s only recently that I’ve seen research trying to get at these social&#x2F;physiological mechanisms, but as far as funding is concerned, hard biological sciences are more interesting. Everyone just wants to edit the genome and call it a day, but I think we could get much further if we understood how life events lead to physiological ailments later in life.

&gt; 4. Conclusion &quot;...grounding one’s work in an understanding of structural racism will improve model accuracy...&quot;<p>It is not an interesting result to say models not modeling reality are less accurate; the cogent discussion is to what degree systemic racism exists IN reality. This is textbook begging the question.<p>&gt; Acknowledgements: Conflict of Interest: None declared.<p>&gt; Funding: Whitney R. Robinson is supported by the National Institute of Minority Health<p>I am not familiar with standards of conflict declaration, but this looks like a pretty clear conflict of interest to me.

This paper fails to account for (or even mention) genetics. To expand on the breathing capacity example, consider the case of the sherpas.<p><a href="https:&#x2F;&#x2F;www.npr.org&#x2F;sections&#x2F;goatsandsoda&#x2F;2017&#x2F;05&#x2F;28&#x2F;530204187&#x2F;the-science-behind-the-super-abilities-of-sherpas" rel="nofollow">https:&#x2F;&#x2F;www.npr.org&#x2F;sections&#x2F;goatsandsoda&#x2F;2017&#x2F;05&#x2F;28&#x2F;5302041...</a>

The real problem is that there are mutually exclusive desiderata from your models.<p>1. If you have two people with identical relevant behavior and different races, you want the model to score them identically.<p>2. Each race should receive a comparable distribution of scores.<p>3. The scores should be as accurate of a predictor of ground facts as possible.<p>Relax the first desiderata and your model is now either explicitly or implicitly (via irrelevant proxy variables) using race to determine results, opening you up to racial discrimination lawsuits. Relax the second desiderata and your model is now creating disparate impact across racial groups, opening you up to racial discrimination lawsuit. Relax the third and you&#x27;re leaving accuracy, and thus money, on the table.

Disparate impact of results is NOT proof of racism. People are complicated - and some racial correlations are perfectly valid - at least until enough members of a racial group decide to change them...

&quot;Conflict of Interest: None declared.&quot;<p>In an article about structural racism, I would have expected more here.

This post was on the front page 10 minutes ago but now it can&#x27;t be found anywhere. It&#x27;s pretty disappointing that it has been removed solely on the basis of being about race or being controversial. Race and racism are a part of society and it presumably made it to the front page because it&#x27;s a topic that enough people found interesting to vote for...so why remove it?

People of every race find success in technology and in America more broadly. Many of these people&#x27;s ancestors came to America with nothing. Those groups which find more success than average generally have a cultural focus on education and other behaviors associated with responsible action. Groups that don&#x27;t succeed generally do not share these qualities.<p>It is these cultural differences which cause most of the group disparity in America, not &quot;structural racism&quot;, yet the &quot;critical race theorists&quot; (race hustlers and grievance mongers) and their followers ignore these major factors and replace them with straw men.<p>In order to fix a problem, it&#x27;s important to understand the actual causes. The sociologists and other assorted race hustlers will only divide us and lead us astray.

The definition for structural racism according to the article:<p>&gt; Structural racism refers to “the totality of ways in which societies foster [racial] discrimination, via mutually reinforcing [inequitable] systems...(e.g., in housing, education, employment, earnings, benefits, credit, media, health care, criminal justice, etc.) that in turn reinforce discriminatory beliefs, values, and distribution of resources,” reflected in history, culture, and interconnected institutions (Bailey and others, 2017).<p>I think I might be misunderstanding, but given this includes “culture”, is this so sufficiently broad such that hypothetical scenarios such as this (no idea if this is accurate) would be captured “white people are culturally more likely to use crystal meth than other racial groups, ergo they are victims of (a certain kind of) systemic racism”?<p>It seems like this is just a catchall for any kind of error associated with a racial group, and the article is merely cautioning against such errors. If so, it begs the questions “why not just say so?” and “why use such a loaded term like systemic racism?”.