[mikeash]

What is the difference between “women are biologically unsuited for tech jobs” and “the distribution of personality traits is different in men and women, which accounts for the skewed gender distribution in the tech industry”?

[belorn]

The difference between preference and ability is one indicate what a person want to do, and the second dictate what a person can do.

For example, people of red hair have generally higher sensitivity to ultraviolet light["Variants of the melanocyte-stimulating hormone receptor gene are associated with red hair and fair skin in humans". Nature Genetics. 1995]. As a result a person with red hair will easier get sun burn and could very likely have a distribution that lean slightly away from professions which exposes the body to a lot of ultraviolet light. This do not mean that red hair color make someone unsuited to do outdoor work. A increased sensitivity to ultraviolet light is not a factor in a persona ability to perform work outdoor, only how comfortable such work would be.

[mikeash]

Seems to me that redheads are unsuited to outdoor work in more sunny climates without modern technology to help. Sunburn isn't merely uncomfortable, it can be debilitating and even deadly. So I don't think that's a very good example.

Back to the actual case at hand, I don't see why "personality traits" would automatically translate to "preference" rather than "ability." Even if it does, don't people do a better job when it's something they enjoy?

[belorn]

You are thinking of the extreme end when talking about debilitating and deadly sunburns, and is not within normal variance. Death by sunburn is a extreme rare event, generally caused by other contributing factors such as dehydration.

If we take the extreme end of everything, practically all in life is deadly. Salt is extremely deadly at high doses, but it is also suitable as food. Pure oxygen is deadly, but also suitable in medicine. Looking at the extreme part of variance is unhelpful to determine suitability when suitability is in the 99.99th percentile and unsuitability is outside it.

> don't people do a better job when it's something they enjoy?

If one assume that people pick jobs which they enjoy, then that issue is a bit moot. If 10% of the population like a particular activity, then it seems naturally that those 10% will do a good job. There is no logical argument that say that the 10% are not suited for the job just because 90% don't want the job and thus don't seek it. Preference of the 90% don't dictate the suitability of the remaining 10%.