But wouldn't that make it less likely? The first programmers had zero chance of being in the same field as their parents.
The numbers are supposed to based on the child generation. If the numbers were based on the parent generation ("likelihood of passing down job preference") then a growing field should indeed increase the likelihood.
> The numbers are supposed to based on the child generation. If the numbers were based on the parent generation ("likelihood of passing down job preference") then a growing field should indeed increase the likelihood.
If we denote the event "Child is a programmer" by C and the event "A parent is a programmer." by P, then the sentence "Computer programmers have parents who are computer programmers at a rate 6 times the rest of the population." can be expressed in terms of probability as P(P|C)/P(P) = 6. Bayes' rule tells us that P(P|C) = P(C|P)P(P)/P(C), so the sentence is equivalent to P(C|P)/P(C) or "Parents who are computer programmers have children who are computer programmers at a rate 6 times the rest of the population."
It doesn't matter whether you are looking at the child generation or the parent generation, the heritability relative to the general population will come out to the same number. A growing field may increase the P(C|P) part of the ratio, but it will also increase P(C) in general, so it is not clear whether the relative measurement will change at all, and in which direction.
Thought experiment: generation n has 1% fidgeteers, generation n+1 has 100% fidgeteers. All generation n fidgeteers will have passed on their profession (procreation assumed), but few generation n+1 fidgeteers will have followed in their parents' footsteps (assuming non-fidgeteers also procreate).
Anecdotal data point: 3rd generation computer programmer here, and I hope that one of my sons also takes up the mantle, although I haven't seen any evidence that it's likely yet. I took it up because I had computers to play with at home, and my dad had Logo (turtle graphics) on one of them.
I thought the same at first, but it could also be interpreted that if you think your job is dying out (or just not a nice job) then you'll steer your kids away from it, whereas if it's a good job with lots of prospects then you'll use your connections and skills to direct your kids towards the same career.
Kind of undercuts the "self-made" programmer meritocracy that seems to be a prevelant idea though.
The numbers are supposed to based on the child generation. If the numbers were based on the parent generation ("likelihood of passing down job preference") then a growing field should indeed increase the likelihood.