4 ms·

Mostly, it's "'Of Men and Mice' isn't just a movie..." But, also: - Nature/Nurture is almost unsolvable for humans, at least within the confines of commonly a

Mostly, it's "'Of Men and Mice' isn't just a movie..." But, also: - Nature/Nurture is almost unsolvable for humans, at least within the confines of commonly accepted ethical standards - Even where you can show statistically significant between-group differences, they are almost certain to be swamped by variability between individuals. Example: Women are, on average, slightly older than men (because of higher life expectancy). The p-value here is as close to zero as you'll get in the real world. But, given two randomly selected people, one male and one female, knowing the above is close to useless! The probability of the woman being older differs from 0.5 by no more than a fraction of a percentage point. - The difference b/w mice and humans is that humans have the ability to overcome biology with reason.

PeterisP 6y ago
"Nature/Nurture is almost unsolvable" - I was under the impression that for most traits we have researched the evidence shows that the Nature/Nurture answer is simply "both" ? I.e. I agree that it's getting good experimental data on the exact mix of nature vs nurture is difficult or impossible with ethical methods, but for many traits we can (as far as I understand - it's not my field of research) demonstrate that it's definitely not 90/10 or 10/90 purely from observational studies. But perhaps it's more accurate to look at this from a "four quadrant" (simplified) perspective, with nature/nurture on one axis and hereditary/nonhereditary on the other axis. So you'd have hereditary nature (biological determinism) factors in one quarter, non-hereditary nature (biological variation, mutations, things like most cases of Down syndrome) in another; hereditary nurture (socioeconomic status, which affects a lot, cultural aspects and family values), and non-hereditary nurture, where (among other things) I'd be eager to place things which are under the individual willpower control even though they're not technically "nurture".
- IfOnlyYouKnew 6y ago
  Yes, indeed. I was thinking of gender roles, specifically, where twin studies are of little help (because "nature" is just sex, not "inherited from parents", and "nurture" is about the influences of society, not your immediate caretakers)
lliamander 6y ago
> Even where you can show statistically significant between-group differences, they are almost certain to be swamped by variability between individuals. Example: Women are, on average, slightly older than men (because of higher life expectancy). The p-value here is as close to zero as you'll get in the real world. > But, given two randomly selected people, one male and one female, knowing the above is close to useless! The probability of the woman being older differs from 0.5 by no more than a fraction of a percentage point. True, but sometimes it is the edges of the distribution that are really interesting, and there small differences in the mean can have a big impact. For instance, what is the male/female ratio in nursing homes, or among people older than 100? I don't know, but I'm willing to bet that it's a noticeable difference. Similarly true with the difference between male/female aggression as it pertains to the prison population.

PeterisP 6y ago

"Nature/Nurture is almost unsolvable" - I was under the impression that for most traits we have researched the evidence shows that the Nature/Nurture answer is simply "both" ? I.e. I agree that it's getting good experimental data on the exact mix of nature vs nurture is difficult or impossible with ethical methods, but for many traits we can (as far as I understand - it's not my field of research) demonstrate that it's definitely not 90/10 or 10/90 purely from observational studies. But perhaps it's more accurate to look at this from a "four quadrant" (simplified) perspective, with nature/nurture on one axis and hereditary/nonhereditary on the other axis. So you'd have hereditary nature (biological determinism) factors in one quarter, non-hereditary nature (biological variation, mutations, things like most cases of Down syndrome) in another; hereditary nurture (socioeconomic status, which affects a lot, cultural aspects and family values), and non-hereditary nurture, where (among other things) I'd be eager to place things which are under the individual willpower control even though they're not technically "nurture".

IfOnlyYouKnew 6y ago
Yes, indeed. I was thinking of gender roles, specifically, where twin studies are of little help (because "nature" is just sex, not "inherited from parents", and "nurture" is about the influences of society, not your immediate caretakers)

lliamander 6y ago

> Even where you can show statistically significant between-group differences, they are almost certain to be swamped by variability between individuals. Example: Women are, on average, slightly older than men (because of higher life expectancy). The p-value here is as close to zero as you'll get in the real world. > But, given two randomly selected people, one male and one female, knowing the above is close to useless! The probability of the woman being older differs from 0.5 by no more than a fraction of a percentage point. True, but sometimes it is the edges of the distribution that are really interesting, and there small differences in the mean can have a big impact. For instance, what is the male/female ratio in nursing homes, or among people older than 100? I don't know, but I'm willing to bet that it's a noticeable difference. Similarly true with the difference between male/female aggression as it pertains to the prison population.