Identical twins are still definitely different in some ways (whether raised in the same family or not), and I think the source of those differences is not well understood. See “§1.4: What is E, really?” here.
Random example on my mind: My spouse is noticeably more happy and well-adjusted this year than last year, in a way that I figure would probably have a small but measurable effect on a personality survey, because she left a job she hated for a job she liked. But she obviously has the same genes now as she did last year.
But she obviously has the same genes now as she did last year.
Leaving the job is not an example because it could have increased the heritability of her personality survey. Pretty much all traits change in heritability over a lifetime (eg. Wilson effect for IQ or think about, say, psychiatric disorders) and other variance components as well, so you can’t point to a single change and say it is an example of E because ‘A didn’t change’ - well, neither did C (if an adult) or D, they could all be increasing due to that single change (you can view it as regression to a mean as she finally leaves an outlier job, as most people do not ‘hate’ their job), so in fact, it may be any of ACD!
I think you’re misunderstanding what I was trying to say, sorry I left out the details. Here’s what I had in mind:
In the methodology of a classic twin study, . So the question “why is E more than zero?” is the same as asking “why do identical twins reared together not get exactly the same scores on personality questionnaires and health questionnaires and vocabulary tests etc.?”
Now, my spouse (call her S) had a job she hated at age 39 and a job she liked at age 40. If she had an identical twin (call her S’), and the two of them happened to participate in a heritability study when they were both age 40, would S’ have also had a job she liked? Or if they had participated in a heritability study when they were both age 39, would S’ have also had a job she hated? Probably at least one of those questions has an answer “no”, right? Unless there’s an extraordinary coincidence where S & S’ find their ways out of bad job situations in the same year, well into adulthood.
And if one of those questions has an answer “no”, then that’s a kind of thing that would have pulled r down from 1.0, a.k.a. contributed to E, at least for a certain fraction of adulthood.
I’m not saying this one thing is a BIG effect, but I think there are a lot of things like that, and if you add them up, I do think it explains some noticeable fraction of E. I’d put it in the “luck of the draw” category in my breakdown of §1.4 here.
(That’s assuming I’m right that the bad job was dragging her down enough to show up on a personality survey, but I really think it was, not hugely but nonzero.)
My point is that in your scenario, it is entirely possible that S’ had managed to obtain a job she liked because she was genetically predisposed (eg. having a nice personality), but that S had not for exogenous random bad luck, and that if they were measured at age 39, they would seem unusually discordant twins—except it was just bad luck, and S would regress to her mean at age 40 and now they would be concordant. You can’t just handwave it and say, ‘S just got a new better job! Heritability must be going down!’ The new job could well be heritability going up. Your ‘random example’ is just poorly chosen and irrelevant to the discussion since it’s not at all obvious that it is an example of E going up and ACD going done just because “she obviously has the same genes now as she did last year.” There could be a time reversal: the bad job could have been the E, not the good one!
I agree with what you wrote. I never said anything about E going up or heritability going down. I’m not sure how I gave you that impression.
It seems that we’re on the same page that bad luck exists (and good luck also exists), and that bad and good luck can affect one’s life circumstances in a way that may impact questionnaire results, and can thus contribute to identical twins being discordant, at any given time. That’s all I was saying. It’s meant to be a simple and obvious point.
Identical twins are still definitely different in some ways (whether raised in the same family or not), and I think the source of those differences is not well understood. See “§1.4: What is E, really?” here.
Random example on my mind: My spouse is noticeably more happy and well-adjusted this year than last year, in a way that I figure would probably have a small but measurable effect on a personality survey, because she left a job she hated for a job she liked. But she obviously has the same genes now as she did last year.
Leaving the job is not an example because it could have increased the heritability of her personality survey. Pretty much all traits change in heritability over a lifetime (eg. Wilson effect for IQ or think about, say, psychiatric disorders) and other variance components as well, so you can’t point to a single change and say it is an example of E because ‘A didn’t change’ - well, neither did C (if an adult) or D, they could all be increasing due to that single change (you can view it as regression to a mean as she finally leaves an outlier job, as most people do not ‘hate’ their job), so in fact, it may be any of ACD!
I think you’re misunderstanding what I was trying to say, sorry I left out the details. Here’s what I had in mind:
In the methodology of a classic twin study,
Now, my spouse (call her S) had a job she hated at age 39 and a job she liked at age 40. If she had an identical twin (call her S’), and the two of them happened to participate in a heritability study when they were both age 40, would S’ have also had a job she liked? Or if they had participated in a heritability study when they were both age 39, would S’ have also had a job she hated? Probably at least one of those questions has an answer “no”, right? Unless there’s an extraordinary coincidence where S & S’ find their ways out of bad job situations in the same year, well into adulthood.
And if one of those questions has an answer “no”, then that’s a kind of thing that would have pulled r down from 1.0, a.k.a. contributed to E, at least for a certain fraction of adulthood.
I’m not saying this one thing is a BIG effect, but I think there are a lot of things like that, and if you add them up, I do think it explains some noticeable fraction of E. I’d put it in the “luck of the draw” category in my breakdown of §1.4 here.
(That’s assuming I’m right that the bad job was dragging her down enough to show up on a personality survey, but I really think it was, not hugely but nonzero.)
My point is that in your scenario, it is entirely possible that S’ had managed to obtain a job she liked because she was genetically predisposed (eg. having a nice personality), but that S had not for exogenous random bad luck, and that if they were measured at age 39, they would seem unusually discordant twins—except it was just bad luck, and S would regress to her mean at age 40 and now they would be concordant. You can’t just handwave it and say, ‘S just got a new better job! Heritability must be going down!’ The new job could well be heritability going up. Your ‘random example’ is just poorly chosen and irrelevant to the discussion since it’s not at all obvious that it is an example of E going up and ACD going done just because “she obviously has the same genes now as she did last year.” There could be a time reversal: the bad job could have been the E, not the good one!
I agree with what you wrote. I never said anything about E going up or heritability going down. I’m not sure how I gave you that impression.
It seems that we’re on the same page that bad luck exists (and good luck also exists), and that bad and good luck can affect one’s life circumstances in a way that may impact questionnaire results, and can thus contribute to identical twins being discordant, at any given time. That’s all I was saying. It’s meant to be a simple and obvious point.