cross-posted from: https://lemmygrad.ml/post/3768443
Researchers estimate that each additional centimeter of height is associated with a 1.30% increase in annual income.
https://www.forbes.com/sites/traversmark/2020/04/16/your-height-has-a-big-impact-on-your-salary-new-research-seeks-to-understand-why/
You want to study the effect of a certain trait on a group. So you take two groups and try to make them as even as possible, aside from the trait you are testing. Ideally you'd match every guy in your sample with someone who is mostly identical in upbringing, location, etc, but taller. That's how you make sure it's actually height causing the effect and not something else. If you don't account for this, you might end up with a bunch of bourgeois short kings getting compared to 6 foot tall poor people, and might even come the conclusion that height has a negative correlation with wealth.
It's not saying "height matters, not those other things." But rather "height is one of those things that effect socioeconomic status, statistically."
I don't get how making sexual difference equal does anything other than make sexual difference disappear. When half the population is shorter because of sexual difference, and also paid less due to structural discrimination on the basis of gender/sex, making those differences equal makes the actual causes disappear. I just don't get it. I have no background in labs or experimentation so sorry if I'm being dense.
Marx flattens some differences in order to illustrate class distinctions and disappear certain confusing elements before working them in later, but he explains why and how, and what effect this will have.
But how would you do this with sex, when the thing you are measuring, height and pay, are both directly related, either physiologically or socially? I couldn't find the link to the study or more info, and the fact this is in Forbes makes my
register a beep
The purpose is to measure the effect of a single variable, so you make sure to correct for all other variables. For example, to measure the effect of height you might compare white men only against white men, black women only against black women, etc.
In a study measuring the gender pay gap, they would be correcting for variables other than gender.
Is that what "this estimation assumes other factors associated with earning potential — for instance, gender, age, years of schooling, and location — are held equal," means? Cuz that's not what it sounds like. Assuming things are equal isn't this statistical matching thing you are talking about
Edit: I found the study so I'm trying to figure out how these other factors are controlled for.
Yes. And usually these factors are controlled for when picking your study's sample.
https://www.statisticshowto.com/matched-samples/
"All factors held equal" doesn't mean the different factors are equal in effect, just that the sample is chosen such that the other factors are identical across the two groups.
Yes, that's what it means - if all other factors they list are equal, meaning if the only difference is height.