> In science, we call out qualitative reasoning as being biased and unscientific.
Ha! I'm a scientist by training, and your claim makes me smile. Most of Science starts by qualitative reasoning and observation. It's because you notice phenomena that you emit hypotheses as to why they occur, and then you design experiments to generate data and verify your hypothesis (i.e. whether your qualitative understanding is correct or not).
Right, we use qualitative reasoning at the beginning and try to temper our biases separately, but how can you do unbiased evaluation without numbers? Even the social sciences has to rely at numbers and statistics eventually.
> but how can you do unbiased evaluation without numbers?
First, collecting data must be made to answer a question. The current way of asking ethnicity based on unvalidated criteria (basically what you identify yourself as) does not mean anything. It's rubbish as data, because there are almost no "pure" individuals in the US anymore, people have been mixed for generations.
The way the current data is used is to reach a political agenda to say that we are in a state of inequality between races and sexes and that the government has to step in to fix things, hence you need the government to spend money and resources on this, etc... It's NOT a scientific study at work, it's data used for political purposes.
Plus, it's not unbiased either because it's not in an observational state. Individuals and companies are aware of these ratios in these companies and know that they are expected to do something about it. That's not science at work, it's social pressure at work.
> It's rubbish as data, because there are almost no "pure" individuals in the US anymore, people have been mixed for generations.
It's not rubbish. You can't simultaneously discuss statistics about black incarceration or female underrepresentation in tech while also denying that such classifications even exist in the first place. The lines blur sometimes, but pretending there are no lines denies reality.
> that the government has to step in to fix things
You're putting the cart before the horse. This is a private company's data, not any specific recommendation for government action.
>Plus, it's not unbiased either because it's not in an observational state. Individuals and companies are aware of these ratios in these companies and know that they are expected to do something about it. That's not science at work, it's social pressure at work.
First, some companies just plain don't care and don't feel any social pressure because their insulated from any real feedback or criticism. Second, any social science work includes some degree of bias because we're not all robots. Saying no possible conclusions can be drawn from demographic data is unscientific and akin to global warming denial.
> You're putting the cart before the horse. This is a private company's data, not any specific recommendation for government action.
The federal government (and potentially individual States too) has been active for years to enforce quota in various domains to reduce "discrimination". Of course companies feel the pressure to do something about it, because if they don't, they may be targeted next in terms of Employment Laws.
> Second, any social science work includes some degree of bias because we're not all robots. Saying no possible conclusions can be drawn from demographic data is unscientific and akin to global warming denial.
A proper social study should always lead to further studies unless you are crystal clear on how to read the data out. Because "we are not robots", the explanations are not always simple and it's not JUST because there's racism or discrimination that there are differences in who gets what job. It's just like if you were saying that there's racism against white people among construction workers, because most of them are not white. This would be missing the entire point because you'd be focused on the numbers instead of trying to understand why it is so and what are the incentives in place.
So yeah, most "demography data based conclusions" are rubbish because they do not focus on explaining the individuals behaviors and what's in it for each of them. And yeah, global warming is actually very similar: data is sparse, grossly extrapolated, and used as a political agenda and a source of new taxes. Global Warming may be happening, but certainly not as fast as Al Gore wanted us to believe, and whether CO2 is the real culprit is another matter for discussion. But that's a whole other topic not relevant to the point discussed here.
Who said that? Of course you can identify problems without having any number. That's called qualitative understanding.