Hacker News Reader

Comment by gboudrias | Hacker News Reader

The closer your field's demographics are to your society's, the lower the risk of (unconscious) bias is. Science as a whole should strive to be as unbiased as possible, as that's the closest we can get to objectivity.

The same logic applies to businesses, to reduce the risk of making mistakes that come with echo chambers.

View on HN

Using Swedish statistics from a few years ago, 12.5% of the population has a professional job which has at minimum 40% men and 40% women. 88.5% do not. Split per gender this was 88.4% for men and 88.6% for women.

If we state that bias is a risk then we must conclude that around 90% of the population work in echo chambers (more if we find that 60/40 is still quite bad segregation) and could use the benefit of reduced risk of mistakes. It would be interesting to hear what suggestions people have to address this generally so that the 88.4% of men and 88.6% women can reduce the risk.

Could you explain what point you're making? I can't tell if you're disagreeing with me or not.

Not an disagreement, but rather a remark. What ever the effect of gender segregation is, it would then be practically universal here because thats how the data look like.

In India there is a more equal split between women and men in the STEM fields, yet women are much less liberated then in Western countries. The reason being is that women there are forced to enter whatever career gives them the biggest paycheck, which happens to be STEM.

So, I'd say your assumption is incorrect.

I'm not saying it benefits the public necessarily, I'm saying it benefits those who opt for diversity. I don't expect STEM to have as big an impact on women's liberation as, say, business or politics. I don't think two cultures can be compared on this basis, my assumption is that, within one culture, diversified organizations will win out on average.

I think it depends in the task, and what the interest of the various groups are. If you have 10 engineers but only 2 of them are really passionate about it, and the other 8 would rather be doing something else, but due to incentives based on their minority status, they decided it was worth the extra money, you'd probably do a lot worse than a company which had 10 engineers that were all hired based on merit.

But diversity is only valuable to the extent that it helps create a better society. Like, ideally noone would even have to work! And there are plenty of examples of segregated (i.e. non-diverse) situations that we support becaues we deem them socially beneficial (e.g. sex-specific gyms and toilets, tall people competing in basketball, smart people are scientists, ...) (to be precise, personally I don't support these, but our society seems to, which would imply that the people on average do).

> The closer your field's demographics are to your society's, the lower the risk of (unconscious) bias is.

That's only true if all biases cancel each other out perfectly.

No, because this isn't a binary result. If group a and b both have bias 1, that bias won't be cancelled. But if they differ on bias 2, the combined group has less bias than a or b while not having 'cancelled all biases perfectly'. Perfection is a false goal while 'lower risk' is attainable and worthwhile.

Fair enough, let's forget the word "perfectly". Let's have a "practical" example. A society is made up of 99% "ethnic Germans" and 1% Jews. The year is 1933.

A faculty is made up of 90% "ethnic Germans" and 10% Jews. According to the hypothesis, getting that makeup closer to the societal demographics will result in a "lower risk of bias".

I hope it is now obvious to see that the hypothesis is a non-sequitur.

How is that a counter-example? They diverged from the overall population to an extreme level (one demographic over-represented by a 10x factor compared to the overall population) and also had an extreme level of bias (imprisoned and murdered a specific demographic).

The specific values are irrelevant, for the hypothesis to be true, it has to be true for all cases.

Let's take another example:

A planet somewhere in the milky way galaxy is populated equally by Red, Blue and Green Octopodes. They are biased differently in three dimensions:

  Red   (3,     2,    -2)
  Blue  (0,     3,     4)
  Green (23,  -56,   128)

The goal is to minimize bias. The hypothesis says bias in a group will be minimized by giving each color group an equal share in the group. However, at a first glance, we can immediately see that in order to minimize bias, we must reduce the number of Green Octopodes in the group. The hypothesis is wrong, because it doesn't generalize.

Note, I'm not making an argument against diversity. The average Green Octopode may hold extremist views, but they shouldn't be excluded from a group merely to reduce bias. Just like humans, Green Octopodes should be treated as the individuals that they are, not as a weight in a multivariate optimization problem.

I've seen no calls to increase the numbers of conservatives/Republicans in STEM, despite the fact that political ideology and affiliation is actually testing the way people think. For example, though the percentage of scientists who lean Rep. is 12% to the general public's 35%, the percentages are 81% and 52% for lean Dem.[1] And the gap in other political values is high as well.

Demographics are only slightly correlated with how people think, so this form of argument falls flat unless you think encouraging conservatives to get into science is far more important.

[1] http://www.people-press.org/2009/07/09/section-4-scientists-...

Maybe no one wants to spend money on this, but I think a greater diversity of viewpoints would lead to more trust from the public. It is a hard sell for either party though.

Is reducing bias the most important goal?

(Definitely not for me. Life today is much better than it was 50 years ago in communist Yugoslavia, despite higher inequality and bias, due to progress in science and business.)

Imagine how much more progress we could have had in science and business if we didn't de facto force out so many possible candidates who were interested in the field after they had their first kid.

Reply on news.ycombinator.com