6 ms·

An essential concept for this discussion: Another problem that Plomin encounters with explaining his findings is that people often confuse group and individual

An essential concept for this discussion: Another problem that Plomin encounters with explaining his findings is that people often confuse group and individual differences – or, to put it another way, the distinction between means and variances. Thus, the average height of northern European males has increased by more than 15cm in the past two centuries. That is obviously due to changes in environment. However, the variation in height between northern European males is down to genetics. The same applies to psychological traits. “The causes of average differences,” he says, “aren’t necessarily related to causes of individual differences. So that’s why you can say heritability can be very high for a trait, but the average differences between groups – ethnic groups, gender – could be entirely environmental; for example, as a result of discrimination. The confusion between means and variances is a fundamental misunderstanding.”

noobermin 8y ago
How does this work? How can one thing that causes the motion of the average not also affect the variation? How wouldn't the one which inherited the worst genes not benefit from environment?
- raincom 8y ago
  It is like how rising tides lift all boats. A good environment helps people without the right genetics; but this does not mean that this group can beat another set with right genetics in the same good environment.
  - noobermin 8y ago
    What you just said sounds like a misunderstanding of the words "average" and "variance". Usually when things like intelligence is modeled, it is modeled as a normal distribution. Moving the average for a normal distribution moves the majority of people in the distribution. Variance just expresses how wide the distribution is, or how far you can expect people to deviate from the average. What you're saying seems to suggest the distribution is multimodal. I'm not sure that's what they said.
  - forapurpose 8y ago
    > A good environment helps people without the right genetics; but this does not mean that this group can beat another set with right genetics in the same good environment. I think that's backwards from what Plomin is saying in that quote. He's saying that, regardless of genetics, a better environment will cause the average performance of the group experiencing it to outperform the average performance of a group experiencing a worse environment. And he's saying that DNA influences the amount of each group's variation around that average performance. For example, if we were looking at height, * There are two groups, 1 and 2; people in Group 1 have one kind of tallness DNA (tDNA, something I just made up), and people in Group 2 have another kind of tDNA; also, Group 1 grows up in an environment with better nutrition and exercise than Group 2. Both groups are one gender, let's say women, to simplify height statistics. * Group 1 averages a height of 5'7" due to their relatively better environment, and due to their tDNA their variance is that 50% are between 5'5" and 5'8", a 3-inch range, and 95% are between 5'3" and 5'10", a 7-inch range. * Group 2 averages a height of 5'5" due to their relatively worse environment, and due to their different tDNA their variance is larger: 50% score between 5'3" and 5'9", a 6-inch range, and 95% are between 5'0" and 5'11", an 11-inch range.
golemiprague 8y ago
Actually I couldn't understand this concept, seems to me he is just afraid of the implications of his theory. If anything I would think that the multiplication of the genetic effect creates much larger differences between groups than between individuals. It is similiar to sport, people like to watch one on one female sport like tennis or various olympic sports because the effect of them being slightly worst than men doesn't manifest that much. But once you look at team sports like football or basketball it just multiplies and it looks so much worst than the men equivalent.
leoc 8y ago
It seems the hypothetical may be doing a lot of work there for Plomin: see pp. 9-10 of https://www.tandfonline.com/doi/abs/10.1080/02680939.2016.1139189 https://www.tandfonline.com/doi/abs/10.1080/02680939.2016.11...

noobermin 8y ago

How does this work? How can one thing that causes the motion of the average not also affect the variation? How wouldn't the one which inherited the worst genes not benefit from environment?

raincom 8y ago
It is like how rising tides lift all boats. A good environment helps people without the right genetics; but this does not mean that this group can beat another set with right genetics in the same good environment.
- noobermin 8y ago
  What you just said sounds like a misunderstanding of the words "average" and "variance". Usually when things like intelligence is modeled, it is modeled as a normal distribution. Moving the average for a normal distribution moves the majority of people in the distribution. Variance just expresses how wide the distribution is, or how far you can expect people to deviate from the average. What you're saying seems to suggest the distribution is multimodal. I'm not sure that's what they said.
- forapurpose 8y ago
  > A good environment helps people without the right genetics; but this does not mean that this group can beat another set with right genetics in the same good environment. I think that's backwards from what Plomin is saying in that quote. He's saying that, regardless of genetics, a better environment will cause the average performance of the group experiencing it to outperform the average performance of a group experiencing a worse environment. And he's saying that DNA influences the amount of each group's variation around that average performance. For example, if we were looking at height, * There are two groups, 1 and 2; people in Group 1 have one kind of tallness DNA (tDNA, something I just made up), and people in Group 2 have another kind of tDNA; also, Group 1 grows up in an environment with better nutrition and exercise than Group 2. Both groups are one gender, let's say women, to simplify height statistics. * Group 1 averages a height of 5'7" due to their relatively better environment, and due to their tDNA their variance is that 50% are between 5'5" and 5'8", a 3-inch range, and 95% are between 5'3" and 5'10", a 7-inch range. * Group 2 averages a height of 5'5" due to their relatively worse environment, and due to their different tDNA their variance is larger: 50% score between 5'3" and 5'9", a 6-inch range, and 95% are between 5'0" and 5'11", an 11-inch range.

golemiprague 8y ago

Actually I couldn't understand this concept, seems to me he is just afraid of the implications of his theory. If anything I would think that the multiplication of the genetic effect creates much larger differences between groups than between individuals. It is similiar to sport, people like to watch one on one female sport like tennis or various olympic sports because the effect of them being slightly worst than men doesn't manifest that much. But once you look at team sports like football or basketball it just multiplies and it looks so much worst than the men equivalent.

leoc 8y ago

It seems the hypothetical may be doing a lot of work there for Plomin: see pp. 9-10 of https://www.tandfonline.com/doi/abs/10.1080/02680939.2016.1139189 https://www.tandfonline.com/doi/abs/10.1080/02680939.2016.11...