IQ heritability -a question to knowledgeable users

[LiLim sAtaN]

So recently I began thinking that if I'll have kids in the future, then I'll do everything so that they are smarter than me and don't have my bad traits. And there is my question - if I choose a mate who is smarter than me, with an IQ of e.g 135-140 (a science whiz), how will my changes of having gifted children increase? If I choose someone coming from a family with a prior history of academic excellence, will they be increased more than as if I choose a gifted person who doesn't have gifted parents? What about inheriting psychological traits?

 

I can relate. I do not want to reproduce without knowing that thus offspring(s) will be of an equal or superior intelligence. I also have ruled out certain "Colors" of the world that I will not even have sex with to even begin any kind of reproduction. In addition, I'm only proud of one of my inherited nationalities and If reproduction were to occur thus offspring(s) must have a larger amount of that nationality than I do (not sure if nationality is correct term).

 

Your desire in wanting to produce an highly intelligent offspring, and might I add naturally, should be more of a priority in the world. Instead I am under the impression that more research is done to further the ability to repair defects, mutations. Mutations that might not exist, or rather could become extinct naturally, by aiding in the achievement of a variation within your suggestion. However, that may only apply to certain "Color" groups and might not be possible for others. In addition, where Is the justification In researching the ability to change physical attributes, affecting appearance, in a fetus (or a baby)?

[hypervalent_iodine]

 

I can relate. I do not want to reproduce without knowing that thus offspring(s) will be of an equal or superior intelligence. I also have ruled out certain "Colors" of the world that I will not even have sex with to even begin any kind of reproduction. In addition, I'm only proud of one of my inherited nationalities and If reproduction were to occur thus offspring(s) must have a larger amount of that nationality than I do (not sure if nationality is correct term).

 

Your desire in wanting to produce an highly intelligent offspring, and might I add naturally, should be more of a priority in the world. Instead I am under the impression that more research is done to further the ability to repair defects, mutations. Mutations that might not exist, or rather could become extinct naturally, by aiding in the achievement of a variation within your suggestion. However, that may only apply to certain "Color" groups and might not be possible for others. In addition, where Is the justification In researching the ability to change physical attributes, affecting appearance, in a fetus (or a baby)?

 

 

!

Moderator Note

Hi LiLim sAtaN,

 

Welcome to SFN. You may want to have a quick look at the rules you agreed to upon signing up to this site, specifically the following rule:

 

 

 

Be civil.

• No flaming. Refrain from insulting or attacking users in a discussion.
• Avoid the use of vulgar language.
• Slurs or prejudice against any group of people (or person) are prohibited.

 

 

!

Moderator Note

We will not tolerate racism here. Please be more mindful of this in the future.

[LiLim sAtaN]

I'm not sure what was racist within my reply. Without knowing the proper terminology I was trying to convey my desire to keep Native American DNA within my offspring by not interbreeding. I'm not even sure if that is the proper way to phrase the statement, but after trying to find the right terminology for over two hours I'm going with that simple, and hopefully not racist statement.

[ahyaa]

There is a serious degree of misunderstanding of how genetics works. DNA isn't rigidly expressed. How your genetic code is interpretted and what results it produces when building you is highly dependent on environment and this has been found to be especially true of intelligence. Brain development is highly dependent upon proper nutrition and an engaging environment.

 

I'm pretty late on this topic... but in response to the OP I completely agree with this. Even if you were to choose a mate with all the "smart" genes there's a good chance they may not be turned on because as Delta1212 said, genes are highly regulated by environment and upbringing of an organism. So even if you have the capability of producing these "smart" proteins by having the genes, they may not be expressed and so you may not ever end up having a "smart" kid. Human heredity is so complicated, I'd think if people truly understood it we'd have taken measures by now to leave the best working genotypes for our offspring.

[Lightmeow]

I'm not sure what was racist within my reply. Without knowing the proper terminology I was trying to convey my desire to keep Native American DNA within my offspring by not interbreeding. I'm not even sure if that is the proper way to phrase the statement, but after trying to find the right terminology for over two hours I'm going with that simple, and hopefully not racist statement.

Well worded, I suggest you edit your last post and change the racist part to that to prevent offending other people in the future.

Edited December 24, 2013 by Lightmeow

[iNow]

I do not want to reproduce without knowing that thus offspring(s) will be of an equal or superior intelligence. I also have ruled out certain "Colors" of the world that I will not even have sex with to even begin any kind of reproduction.

Nope. Not racist at all. I'm really glad you spent 2 hours formulating this sentence to ensure your meaning was clear.

[CharonY]

I'm not sure what was racist within my reply. Without knowing the proper terminology I was trying to convey my desire to keep Native American DNA within my offspring by not interbreeding. I'm not even sure if that is the proper way to phrase the statement, but after trying to find the right terminology for over two hours I'm going with that simple, and hopefully not racist statement.

 

Here is the thing, there is no Native American DNA. You will have mix of alleles (gene variants) that are found in almost all human groups (with very few exceptions in very small, highly inbred populations, maybe). In your population (depending on how isolated it is) certain combinations may be more common than in other populations. Also note that the composition of alleles in each population are not static and have changed massively over time. This is not only due to interbreeding but also due to selection and changes thereof. And example could be the accessibility of new food sources.

[Discere et Docere]

I have a harsh, unpleasant personality inherited from my paternal grandfather and I'm prone to anxiety and hypochondria, inherited from my paternal grandmother. I'm also extremely lazy and envious of others, especially people who are smarter and more hard-working than me."

Do something about that.

 

Personality cannot be inherited. Genetic makeup is really what is inherited. Genetic makeup also pertains to the shape, size, and growth patterns of the brain, which do influence personality and intelligence through neuronal behavior and amount. However, they do not control it.

 

Personality is formed by different aspects and experiences throughout life, and it can change easily.

 

Intelligence, while also dependent upon the aspects of our brain, is also affected by our nutrition, health habits, education, and mental stimuli. I recommend reading about brain plasticity. A good book to start off with is, "The Brain That Changes Itself" by Norman Doige, M.D. I also recommend watching this TEDTalk: http://www.ted.com/talks/james_flynn_why_our_iq_levels_are_higher_than_our_grandparents.html.

[Strange]

Personality cannot be inherited.

 

I find that hard to believe.

 

For example: http://www.sciencedaily.com/releases/2012/05/120516115903.htm

[chadn737]

There is a serious degree of misunderstanding of how genetics works. DNA isn't rigidly expressed. How your genetic code is interpretted and what results it produces when building you is highly dependent on environment and this has been found to be especially true of intelligence. Brain development is highly dependent upon proper nutrition and an engaging environment.

 

It is a complex organ and sub-par building materials will result in a brain that reaches less than full potential. A hazardous, stressful environment which discourages mental development will also hinder mental development in comparison with a secure environment that fosters learning.

 

It's enough of an effect that people with "average" genetic intelligence can have well above average intelligence while those with "exceptional" genetic intelligence can be comparatively stunted. And choosing a partner based on her own intelligence won't tell you whether she's someone who has exceeded her potential despite average genetics because of environmental factors, or whether she's someone who has exceeded her environment because of even more exceptional genetics. You could get lucky or wind up with a lemon either way.

 

The single best thing you can do to improve the intelligence of your children is to properly care for them, not choosing a mate for her presumed genetics. Adequate nutrition and encouragement to enjoy learning won't turn an average person into Einstein, but they will allow your child to reach his or her full potential which will probably be above what most people achieve regardless of how "natively" smart they are.

 

I disagree with the assertion that genetics is so dependent upon environment. It really depends on the specific trait. Of course if you are developing under conditions of starvation you will not reach your full potential....but unless you live in the 3rd world, thats really not a major factor. There is a reason the heritability of IQ is so much higher in developed nations like the US....such environmental factors are a lot less variable and extreme and the majority of people probably have the ability for their genetic potential to be realized.

I think people in this thread are way underestimating the role of genetics and overestimating the role of environment.

[overtone]

 

 

There is a reason the heritability of IQ is so much higher in developed nations like the US....such environmental factors are a lot less variable and extreme and the majority of people probably have the ability for their genetic potential to be realized.

The average poor kid from the inner city, when they start school, has heard spoke to them by an adult less than half as many words as the average suburban wealthy kid. And the quality of that speech - the complexity of it, range of topics, etc, has been also much reduced in comparison. http://centerforeducation.rice.edu/slc/LS/30MillionWordGap.html

 

Being subjected to violent or abusive discipline as a child correlates with a lower adult IQ, and also correlates with violent or abusive discipline of children in turn.

 

Having a higher levels of lead in one's blood, suffering more frequenct childhood infections, diseases, and injuries, eating lower quality food, having no books in the house, spending too much time in a condition of stress (noise, interrupted sleep, threat) etc etc etc - many factors correlate with lower/higher IQs in the US, that also correlate with low income and/or despised sociological "race", and tend to run in generations without being genetically heritable.

[petrushka.googol]

Consider this kim ung yong who is said to have the highest iq had his parents born at the same time. How do you explain this?

[SlavicWolf]

Epic act of forum necromancy. Resurrecting a month-old thread...

[chadn737]

The average poor kid from the inner city, when they start school, has heard spoke to them by an adult less than half as many words as the average suburban wealthy kid. And the quality of that speech - the complexity of it, range of topics, etc, has been also much reduced in comparison. http://centerforeducation.rice.edu/slc/LS/30MillionWordGap.html

 

Being subjected to violent or abusive discipline as a child correlates with a lower adult IQ, and also correlates with violent or abusive discipline of children in turn.

 

Having a higher levels of lead in one's blood, suffering more frequenct childhood infections, diseases, and injuries, eating lower quality food, having no books in the house, spending too much time in a condition of stress (noise, interrupted sleep, threat) etc etc etc - many factors correlate with lower/higher IQs in the US, that also correlate with low income and/or despised sociological "race", and tend to run in generations without being genetically heritable.

 

 

I never said there weren't exceptions or that the environment did not play a role. However, as a whole, the environment is a lot less variable, even in the US as a whole, than it is outside of the US.

 

An important question that we need to ask is at what age are we interested in? Are we interested in the IQ of 2 year olds, 8 year olds, adults? I would maintain that ultimately we are interested in the adult IQ, as smart 2 year olds are sill pretty inept at most things. If we consider then adult IQ, then it appears that genetics is the major source of variability, regardless of environmental factors. Certainly environmental factors play a role, and in fact they appear to be a more important role at younger ages. Socioeconomic status will drive the average IQ of a two year old, but in adulthood, it becomes less of a factor, less a source of variance.

 

Here is a figure from a metastudy of genetic and environmental influences on cognition:

 

 

Elliot M. Tucker-Drob, Daniel A. Briley, and K. Paige Harden

Genetic and Environmental Influences on Cognition Across Development and Context Current Directions in Psychological Science October 2013 22: 349-355,doi:10.1177/0963721413485087

 

 

Somewhere around the age of 8, and then increasing with age, the variance explained by genetics predominates over that of unique and shared environments. If we were to focus only on early ages, certainly socioeconomics is a significant factor, but this influence goes away with time.

 

And haven't you ever heard the old maxim "Correlation does not equal causation". One could say that low IQ correlates with violent or abusive discipline, implying that IQ is the driving factor, not disciplinary action.

Edited January 21, 2014 by chadn737

[SlavicWolf]

I wish that advanced genetic engineering was already available. I one-two generations all people would be smart, healthy and good looking. I would not hesitate even for a while before genetically engineering my kid.

[overtone]

I never said there weren't exceptions or that the environment did not play a role. However, as a whole, the environment is a lot less variable, even in the US as a whole, than it is outside of the US.

And so are the IQ scores.

 

 

 

Certainly environmental factors play a role, and in fact they appear to be a more important role at younger ages. Socioeconomic status will drive the average IQ of a two year old, but in adulthood, it becomes less of a factor, less a source of variance.

Some of the specific environmental factors known to influence IQ are cumulative - lead poisoning, violent discipline and physical trauma, the cognitive effects of factors such as deafness or lazy eye, for example, often build up over development, affecting the older child more than the infant. If these factors are also correlated with genetics - as child abuse, physical trauma, and lead poisoning are - the result will be a false correlation of adult IQ with genetic inheritance.

 

Or as your linked study noted:

 

 

Transactional models predict that people in high-opportunity contexts actively evoke and select positive learning experiences on the basis of their genetic predispositions; these learning experiences, in turn, reciprocally influence cognition. The net result of this transactional process is increasing genetic influence with increasing age and increasing environmental opportunity.

So the question here would be: what sort of "genetic predispositions" are we talking about, and what exactly constitutes a "high opportunity context" for them? That is not at all the same as an inherited IQ score, and argues as much for a stronger as a weaker environmental role.

 

We might even have to reinterpret the basic meaning of the variation on given IQ test given in different "opportunity contexts" from those it was designed to measure.

 

We might, in particular, keep in mind the discoveries of Claude Steele and J Aronson (overview: http://www.nytimes.com/2012/10/07/opinion/sunday/intelligence-and-the-stereotype-threat.html?_r=0 ), researchers who have created false correlations of IQ and genetics in US adults by arranging such simple and universal circumstances as labeling the test an "IQ Test" and having the test subject check a box identifying their sex, race, ethnicity, or country of origin.

Edited January 21, 2014 by overtone

[chadn737]

And so are the IQ scores.

 

 

Do you have support that IQ scores are more variable in other nations? The key here being "variability", not absolute level. A lot of people confuse the fact that in discussing heritability, we are talking about variances within a population, not absolute levels. So what I would be interested in here is in the standard deviations or some other measure of variance and not the actual scores themselves.

 

And again, I am not denying the role of the environment, I am arguing against those that minimize the genetics.

 

Some of the specific environmental factors known to influence IQ are cumulative - lead poisoning, violent discipline and physical trauma, the cognitive effects of factors such as deafness or lazy eye, for example, often build up over development, affecting the older child more than the infant. If these factors are also correlated with genetics - as child abuse, physical trauma, and lead poisoning are - the result will be a false correlation of adult IQ with genetic inheritance.

 

 

Whether or not environmental factors are cumulative will not really be of significant consequence. Cumulative effects would fall under shared or unique environment. These factors are not determined by simple correlation, nor is the increasing effects of genetics with age.

 

Or as your linked study noted:

Quote

Transactional models predict that people in high-opportunity contexts actively evoke and select positive learning experiences on the basis of their genetic predispositions; these learning experiences, in turn, reciprocally influence cognition. The net result of this transactional process is increasing genetic influence with increasing age and increasing environmental opportunity.

So the question here would be: what sort of "genetic predispositions" are we talking about, and what exactly constitutes a "high opportunity context" for them? That is not at all the same as an inherited IQ score, and argues as much for a stronger as a weaker environmental role.

We might even have to reinterpret the basic meaning of the variation on given IQ test given in different "opportunity contexts" from those it was designed to measure.

 

We aren't talking about an inherited IQ score. There seems to be this confusion that its either all genetic or all environment, rather than recognizing that we are discussing different degrees of contribution to the variance in scores.

 

Let's discuss what this one explanation means. The paper in fact gives two potential explanations for this observation, the second you failed to mention, which amounts to quote mining. "High opportunity context" means that with increasing age, the autonomy of the individual increases, this allows their predispositions to come to the forefront. Consider the aforementioned exposure to vocabulary. At 2 years of age, a child will not have the ability to read and little freedom to meet individuals beyond their family and those the family wishes them exposed to. With increasing age, the child will be exposed to more, at school they will have more freedom to interact with different children. They will have the opportunity to increasingly pursue their own interests. The argument here in this explanation is that individuals with genetic predispositions to higher IQs will have the opportunity to pursue those activities and interactions that feedback into their IQ score.

 

So essentially, the explanation is that at younger ages, lack of autonomy holds everyone to a common level. We can make the analogy to say the more extreme forms of communism that prevented anyone from rising above the others regardless of ability. With increasing autonomy, you the chains are taken off and ones natural abilities begin to come to the fore.

 

In fact, if you are going to argue against me from the very paper I quote, you would be better off referring to the latter analysis of Socioeconomic Status (SES)

 

 

 

The caveat of these two graphs is that they are of a single age. The first graph is from age 2, the second from age 17. One thing that stands out is that while at age 2, the genetic component of IQ drops to 0 after -1 standard deviation, at age 17, there is still a genetic influence even at -2 standard deviations. This again goes back to the fact that with increasing age, the influence of genetics increases. This is true even at the lowest levels of SES.

 

Now if we recall what 1 and 2 standard deviations are, then we can begin to get a more accurate picture of how much influence genetics has. The key here is that effects of genetics in explaining variance increases with income/SES.

 

At -1 standard deviation, ~16% of the population is going to be below that line. The remaining ~84% is above it. You don't reach 0% genetics until you drop to almost -2 standard deviations, which is only ~2-3% of the population, but this effect is observed only in very young children. If at -2 standard deviations, the effects of genetics is 0.2, that means that ~20% of the variation is explainable by genetics in those circumstances. That is not insignificant.

 

We might, in particular, keep in mind the discoveries of Claude Steele and J Aronson (overview: http://www.nytimes.c...hreat.html?_r=0 ), researchers who have created false correlations of IQ and genetics in US adults by arranging such simple and universal circumstances as labeling the test an "IQ Test" and having the test subject check a box identifying their sex, race, ethnicity, or country of origin.

 

 

I don't find your reference to Steele and Aronson of particular relevance for several reasons.

 

1) These studies do not measure genetics or even try to correlate genetics. At no part do they attempt to measure heritability or make such a correlation.

2) It specifically deals with between group differences, but I am not really talking about between-group differences.

3) In typical twin studies, such simplistic explanations don't really hold when you are looking at the variance between say two identical twins.

Edited January 21, 2014 by chadn737

[overtone]

The caveat of these two graphs is that they are of a single age. The first graph is from age 2, the second from age 17. One thing that stands out is that while at age 2, the genetic component of IQ drops to 0 after -1 standard deviation, at age 17, there is still a genetic influence even at -2 standard deviations. This again goes back to the fact that with increasing age, the influence of genetics increases

That is a presumption, not a fact. The study does not deal with cumulative environmental influences with sufficient rigor to allow that presumption. Until the environmental influences have been dealt with, the degree to which the convergence measured is due to genetic factors remains uncertain.

 

 

1) These studies do not measure genetics or even try to correlate genetics. At no part do they attempt to measure heritability or make such a correlation.

Are you arguing that there is no genetic component underlying the between group socioeconomic divisions in the US, and everywhere else? I remind you that there is such a thing as inherited wealth, inherited land, inherited dietary and other influences, inherited "opportunity", and inherited social status in most cultures making use of IQ tests - and these factors are often cumulative, with increasing influence on the aging test subject.

 

 

 

2) It specifically deals with between group differences, but I am not really talking about between-group differences.

That is a conclusion of those studies, not an assumption. The test subjects are not grouped in advance by the researchers. Individual IQ scores are correlated with self-identified group membership, not a priori groupings. Self-identified and presumptively genetically correlated group membership appears to be an environmental influence on IQ scores of individuals - and such influences, like many others, increase with age from infancy.

 

They have to be removed, explicitly, before the degree even (let alone the kind) of genetic influence is revealed.

 

 

 

One thing that stands out is that while at age 2, the genetic component of IQ drops to 0 after -1 standard deviation, at age 17, there is still a genetic influence even at -2 standard deviations.

That has not been demonstrated. The complexity of the environment involved has not been recognized and dealt with.

 

 

 

"High opportunity context" means that with increasing age, the autonomy of the individual increases, this allows their predispositions to come to the forefront.

The "predispositions" that are allowed to come to the fore vary considerably according to aspects of the environment that are strongly correlated with genetics. An increasingly lead poisoned child is going to have different "predispositions" than one raised in a cleaner environment, for example. That may easily produce a false correlation between genetic inheritance and IQ scores, and would also account for a significant amount of convergence in the scores - exactly how much can only be determined by dealing explicitly with, say, lead exposure in the test subjects.

 

Look, the presumption that genetic inheritance influences IQ scores seems quite safe, to me. The degree and more importantly the nature of that influence is a very complicated matter - beginning with a clear description of what exactly an IQ score measures in different people, by the way.

 

Nothing in those studies allows a specific number, or anything like precision, to be attached to that presumption. It's interesting data, and focuses ones attention, but the problems are obvious. When the environment is correlated with the genetics, and is as complex as it is, extracting a clear genetic component is going to be very difficult, and its meaning essentially impossible to forecast.

Edited January 25, 2014 by overtone

[chadn737]

That is a presumption, not a fact. The study does not deal with cumulative environmental influences with sufficient rigor to allow that presumption. Until the environmental influences have been dealt with, the degree to which the convergence measured is due to genetic factors remains uncertain.

 

 

This paper is a review and meta-analysis of the research on the subject. Actual estimates of heritability and control for environmental factors were done in the specific studies cited.

 

Specifically what do you mean by the fact that they fail to account for cumulative environmental influences? Your emphasis on these effects being cumulative doesn't really make sense because in studies such effects will be distinguished as either shared or unique environment, not as genetic. In fact, the argument of cumulative environmental influences makes even less sense given the increasing influence of genetics later in life. For instance, in a twin study, as a set of twins age, the environment will not be equal for both individuals. This should result in greater variance between the two twins, which in a twin study would indicate increasing environmental influences, not genetic.

 

So your objection really doesn't make sense to me. Could you specifically state then what sort of "controls" are needed? Can you provide any papers that indicate that cumulative environmental influences masquerades as genetic influences in such studies?

 

Are you arguing that there is no genetic component underlying the between group socioeconomic divisions in the US, and everywhere else? I remind you that there is such a thing as inherited wealth, inherited land, inherited dietary and other influences, inherited "opportunity", and inherited social status in most cultures making use of IQ tests - and these factors are often cumulative, with increasing influence on the aging test subject.

 

 

This objection makes no sense, because it is in reply to an argument I was making against the work of Steele and Aronson. I was stating that their research does not measure heritability in any way or deal with genetics. How you make the leap from that to the idea that I am arguing that there is no genetic component to between group socioeconomic divisions is beyond me. Maybe you could help me understand the flow of your thinking here, because it seems to be very out of context.

 

As I have stated several times now, cumulative environmental effects will fall out as either unique or shared environmental effects, not genetics. Can you explain to me how these effects would masquerade as genetic and not environmental? Our measures of heritability come from some very sophisticated studies that account for such influences. For example, studies of adopted children allows us to break the effects of "inherited wealth, etc". For example, the Colorado Adoption Project follows the cognitive development of adopted siblings. Since these children are split and raised in different environments, we can break the effects of such "inherited" environments:

 

"The current study used data from the Colorado Adoption Project (CAP) to examine cognitive development from infancy through adolescence. Not only is the adoption sibling design a powerful test of genetic and environmental influences on family resemblance, the CAP is the only longitudinal adoption study in existence that measures cognitive skills from infancy into late adolescence (see Petrill et al., 2003; Plomin et al., 1997). The current analysis included measures from adoptive and matched biological control sibling pairs, collected over the entire period of the CAP, to examine the development of cognitive ability throughout the first 16 years of life."

 

Genetic and Environmental Contributions to General Cognitive Ability Through the First 16 Years of Life

 

At no age in this study does the heritability drop below ~0.3 and often it is over ~0.5.

 

That is a conclusion of those studies, not an assumption. The test subjects are not grouped in advance by the researchers. Individual IQ scores are correlated with self-identified group membership, not a priori groupings. Self-identified and presumptively genetically correlated group membership appears to be an environmental influence on IQ scores of individuals - and such influences, like many others, increase with age from infancy.

They have to be removed, explicitly, before the degree even (let alone the kind) of genetic influence is revealed.

 

 

1) What is a "conclusion" and not an "assumption"? Recall that I merely pointed out that Steele and Aronson were discussing the effects of group identity. None of the arguments I have made have anything to do with between-group differences. If you understand genetics, then you know it is possible for genetic influences to contribute significantly to variance within populations even if it does not contribute to variance between populations. I am only arguing for the importance of genetic influences in a general sense, which applies to within population comparisons, I am stating nothing regarding between group differences.

 

2) I also take issue with the idea that "self-identification" is necessarily an accurate portrayal of genetic membership. I am classified as "white". My ancestry is largely German, but the person of largely Greek decent is also classified as "white", even though genetically we are not of the same population. The typical American black individual has a significant proportion of admixture from European populations. We say that Obama is "black", but half his ancestry is European. So self-identification is often a poor measure of actual genetic membership and similarity.

 

3) Self-identified group membership really will not influence the estimation of heritability in twin studies, sibling comparisons. In a twin studies, the shared and unique environmental effects are determined using the variance between twin pairs. So if you measure the variance between two identical black twins, both will presumably self-identify as "black". The effect will be accounted for then.

 

4) Many of the most extensive twin studies have been conducted in relatively homogenous populations, such as the Netherlands Twin Registry.

 

That has not been demonstrated. The complexity of the environment involved has not been recognized and dealt with.

 

Sure it has. These are not simplistic studies, they take into account the complexity of the environment.

The "predispositions" that are allowed to come to the fore vary considerably according to aspects of the environment that are strongly correlated with genetics. An increasingly lead poisoned child is going to have different "predispositions" than one raised in a cleaner environment, for example. That may easily produce a false correlation between genetic inheritance and IQ scores, and would also account for a significant amount of convergence in the scores - exactly how much can only be determined by dealing explicitly with, say, lead exposure in the test subjects.

 

 

 

The effects of lead poisoning will show up as either a shared or unique environmental effect. The power of methods of measuring heritability is that one doesn't need to know specific environmental effects, they will show up as the source of variance between individuals. There are assumption free approaches of estimating heritability from association studies, these show significant degrees of heritability: Genome-wide association studies establish that human intelligence is highly heritable and polygenic.

 

Look, the presumption that genetic inheritance influences IQ scores seems quite safe, to me. The degree and more importantly the nature of that influence is a very complicated matter - beginning with a clear description of what exactly an IQ score measures in different people, by the way.

Nothing in those studies allows a specific number, or anything like precision, to be attached to that presumption. It's interesting data, and focuses ones attention, but the problems are obvious. When the environment is correlated with the genetics, and is as complex as it is, extracting a clear genetic component is going to be very difficult, and its meaning essentially impossible to forecast.

 

 

What an IQ score measures exactly is debatable and obviously imperfect. However, IQ does correlate strongly with many aspects we consider to be "intelligent", so as a proxy, it works.

 

Of course you can't achieve a specific number, because we are talking about the variance of a trait and that will always be context specific. The argument that the data is "complex" is kind of a copout. The complexity of the situation is recognized by modern researchers and they take these factors into account in their models. The meaning is pretty clear, there are segregating variants in the population that account for a significant amount of the variation in human cognitive ability. We don't find this to be particularly controversial for many other traits, such as height, which has a height of ~0.8. Only when it comes to traits like intelligence, does there seem to be an attempt to explain and challenge the same methods used for non-controversial traits.

Edited January 26, 2014 by chadn737

[overtone]

Sure it has. These are not simplistic studies, they take into account the complexity of the environment.

The single example of lead poisoning during development is enough to deny that assertion. Dozens of other factors are also neglected (stereotype handicap, as Steele and Aronson demonstrate, merely one of the more interesting).

 

And that presumption - that the environmental influences have been recognized and controlled for somehow - is key to the conclusions drawn.

 

 

 

The complexity of the situation is recognized by modern researchers and they take these factors into account in their models.

They don't. They assume they have recognized them and taken them into account. I doubt that very much, for the reasons indicated, and so far no evidence reassuring me has appeared.

 

We have a situation with key resemblences to the famous situation in learning theory based on rat studies, when several decades of research product had to be re-evaluated and in many cases discarded upon the discovery and demonstration that rats were far more sensitive to sound, touch, and smell than had been assumed - the extraordinary measures that it turned out were necessary to set up even a simple T maze without providing clues to the rat other than the patterns of reward supposedly being tested revealed that most of the assumptions were delusions founded in researcher obliviousness.

 

It takes even more extraordinary measures to isolate and remove environmental influences on IQ test scores.

 

 

 

We don't find this to be particularly controversial for many other traits, such as height, which has a height of ~0.8.

We should, with height, as John Komlos and several subsequent researchers have thoroughly demonstrated - and created some controversy, btw. The finding that height in individuals is strongly correlated with income inequality in their society, for example, is not without controversial implications. Edited January 27, 2014 by overtone

[chadn737]

The single example of lead poisoning during development is enough to deny that assertion. Dozens of other factors are also neglected (stereotype handicap, as Steele and Aronson demonstrate, merely one of the more interesting).

And that presumption - that the environmental influences have been recognized and controlled for somehow - is key to the conclusions drawn.

 

 

Its interesting the arguments you skip over and ignore. For instance I asked you to: "Could you specifically state then what sort of "controls" are needed? Can you provide any papers that indicate that cumulative environmental influences masquerades as genetic influences in such studies?"

 

The example lead poisoning does not deny that assertion. The design of studies that measure heritability, such as twin studies and sibling comparisons can account for unknown environmental effects, even cumulative ones like lead poisoning. These effects will fall out as a either the shared or unique environmental effect, not the genetic one. These methods do not require specific environmental factors to be identified as an environmental factor. I don't know if you are confused over this fact, but just as the heritability, h², of a trait does not need to specify specific genes, the environmental factors do not need to specify the specific environmental factor.

 

For example, in modern twin study designs, researchers will generate multiple models assigning different degrees of influence to each factor. These models can then be compared to each other to identify which model fits the data the best. In this way we can even see the effect of dropping an entire factor.

 

They don't. They assume they have recognized them and taken them into account. I doubt that very much, for the reasons indicated, and so far no evidence reassuring me has appeared.

We have a situation with key resemblences to the famous situation in learning theory based on rat studies, when several decades of research product had to be re-evaluated and in many cases discarded upon the discovery and demonstration that rats were far more sensitive to sound, touch, and smell than had been assumed - the extraordinary measures that it turned out were necessary to set up even a simple T maze without providing clues to the rat other than the patterns of reward supposedly being tested revealed that most of the assumptions were delusions founded in researcher obliviousness.

It takes even more extraordinary measures to isolate and remove environmental influences on IQ test scores.

 

 

 

 

You seem to be confused over the need to recognize specific factors. Measuring heritability does not require one to know that lead poisoning is a factor in order for the variance caused by lead poisoning to show up as either a shared or unique environmental component. These methods are generalized enough that they can separate out the source of variance for each factor without specifying them exactly.

 

We should, with height, as John Komlos and several subsequent researchers have thoroughly demonstrated - and created some controversy, btw. The finding that height in individuals is strongly correlated with income inequality in their society, for example, is not without controversial implications.

 

 

Please provide specific papers or sources.

 

By the way, John Komlos would agree with me. Here is a talk he gave...let me quote directly "And of course genetics plays a big role. Lets not forget that. But there is a remaining 20% that depends on external circumstances..."

 

From John Komlos's own mouth, environmental factors account for 20% of the variance. That means that 80% of the variance comes from genetics. That is exactly the finding of the research I previously mentioned, where the heritability of height is ~0.8.

 

You seem to be implying that just because a strong correlation exists for some environmental factor, that this negates the strong role of genetics, but that is simply false. You can have a strong correlation between income inequality and height and still have genetics as the primary source of variance.

Edited January 27, 2014 by chadn737

[overtone]

 

 

The design of studies that measure heritability, such as twin studies and sibling comparisons can account for unknown environmental effects, even cumulative ones like lead poisoning. These effects will fall out as a either the shared or unique environmental effect, not the genetic one

Unless you have measured the lead exposure patterns (and the several other such hidden influences) of your subjects, you don't know whether you are dealing with a shared or unique environment. These studies you refer to rest entirely, in their conclusions, on the ability of the researchers to distinguish shared from unique environments, in the circumstances relevant to IQ test scores.

 

If the relevant environment varies in ways correlated with the genetics, it will produce a false correlation in the IQ scores. If the researchers are unaware of this variation, they are likely to read the correlation as real. Since none of these studies, to my knowledge, have controlled for various known influences on IQ scores that do correlate with genetic heritage (lead poisoning, stereotype effect, etc), I have little confidence in their classifications of environments as shared or unique. Is that unreasonable of me?

 

The discovery of such factors as stereotype bias (Steele, Aronson, et al) showed that what had been assumed to be a shared environment (similar socioeconomic circumstances, similar test regimes, etc, ) in many studies of IQ scores was not, that the environments of the test subjects varied in ways that had influenced the scores of the tests in consistent and apparently misleading ways - that happen to be genetically correlated, we note.

 

Another example: studies have shown that a given family in a given house is not necessarily a shared environment for the various children - the different ages, the different sexes, the different birth orders, the different physiologies and illnesses, all have their influences - they have to be controlled, if they are relevant to the aspect being measured and you want to reveal the genetic component in isolation.

 

 

 

From John Komlos's own mouth, environmental factors account for 20% of the variance. That means that 80% of the variance comes from genetics. That is exactly the finding of the research I previously mentioned, where the heritability of height is ~0.8.

That's out of context, and probably (from his writings, I don't bother with youtube videos) doesn't mean quite what you seem to think -

 

one of his key findings was that the intergroup variation in human height seems to be almost completely environmental, and only within a given shared environment does a large contribution of genetics to variance show up - he now recognizes a shared environment by the group's median height and variance, in other words -

 

You seem to be arguing in the opposite direction - assuming, as people talking about human height did for centuries (and concluded that Chinese were shorter than Englishmen by inheritance, for example) that you can recognize a unique environoment, accurately identify a shared environment.

 

Another issue in comparing IQ and height is that the variance in IQ is the only visible aspect - nobody knows what the standard human minimum allotment of IQ looks like, or what the maximum would be, or even if a linear and one dimensional scale is appropriate.

 

Still another issue is the instability of IQ scores - human height does not vary from day to day, or measurement device to measureement device, by any large fraction of the population variance or the available range for a given individual.

 

These issues become critical, to the point of dominance, when researcher begin to postulate feedback mechanisms - when the "intrinsic" IQ is supposed to be governing the environmental influences, selecting opportunities for its own enhancement.

[chadn737]

Overtone, I would ask that you please provide sources for your claims in order to further the discussion. Being able to provide the original data, (i.e. not the New York Times) is important to scientific discussion is it not?

 

Specifically I would like sources on:

 

1) Cumulative environmental effects masquerading as genetic effects (I've asked you for this twice now I think)

 

2) And the specific papers of John Komlos that you cite.

 

Links will be enough, or even titles. I can more than easily access any papers.

 

Thanks, I'll reply later when I have more time.

[jmcnell]

Intelligence is defintely a heritable trait but only to an extent. If you look at some brilliant people they usually had very supportive/intelligent parents. This is not always the case and you can see millions of examples otherwise. I don't believe this limits anyone. We all come with traits that we don't want. It is our duty to try to minimize our bad traits and exploit our good ones.

 

The best quote I tell people that bring this question up is the "12 infant quote" from American psychologist John Broadus Watson.

 

Give me a dozen healthy infants, well-formed, and my own specified world to bring them up in and I'll guarantee to take any one at random and train him to become any type of specialist I might select – doctor, lawyer, artist, merchant-chief and, yes, even beggar-man and thief, regardless of his talents, penchants, tendencies, abilities, vocations, and race of his ancestors. I am going beyond my facts and I admit it, but so have the advocates of the contrary and they have been doing it for many thousands of years.

 

when I took an art class my teacher explained that Frank Lloyd Wright was surrounded by pictures of buildings and architecture by his mother while he was a baby. She essentially molded him to be an architect. Although I can't find anything online but I'll take her word for it.

[chadn737]

Unless you have measured the lead exposure patterns (and the several other such hidden influences) of your subjects, you don't know whether you are dealing with a shared or unique environment. These studies you refer to rest entirely, in their conclusions, on the ability of the researchers to distinguish shared from unique environments, in the circumstances relevant to IQ test scores.

 

 

It really doesn't matter. If with one set of individuals lead exposure falls out as the unique environmental factor and in another set of individuals as the shared environmental factor, then that source of variance will still be either unique or shared and not genetic. We are interested here in the genetic aspects of IQ, so the results are still valid.

 

If the relevant environment varies in ways correlated with the genetics, it will produce a false correlation in the IQ scores. If the researchers are unaware of this variation, they are likely to read the correlation as real. Since none of these studies, to my knowledge, have controlled for various known influences on IQ scores that do correlate with genetic heritage (lead poisoning, stereotype effect, etc), I have little confidence in their classifications of environments as shared or unique. Is that unreasonable of me?

 

 

And how will the environment correlate with genetics? I have asked you to support this claim several times now and you simply keep reasserting it. Thats the fallacy known as argument from repetition. In order for it to do so, every individual in the study would have to be exposed to the same environmental factors, to the exact same degree, with the same effect, otherwise its going to cause variance in a way that falls out as either shared or unique environment. I know of no example where such environmental factors are so closely correlated with genetics that they masquerade as genetics and given how unlikely it is for such diverse sets of individuals to be exposed to such identical environments, the burden is really on you to show that it is. You talk about lead exposure a lot, as if it is some trump card, but those individuals most likely to be exposed to lead in the environment are typically those where measures of heritability show greater amounts of environmental variance, just as one would expect if lead exposure or any other environmental factor were a significant factor in development. There is a reason why in the studies I mentioned previously, heritability drops at the most extreme deviations.

 

So rather than overestimating heritability, I think it is far more common to underestimate it. Its a lot easier for genetics factors to masquerade as environmental factors due to noise, than it is for environmental factors to masquerade as genetic.

 

The discovery of such factors as stereotype bias (Steele, Aronson, et al) showed that what had been assumed to be a shared environment (similar socioeconomic circumstances, similar test regimes, etc, ) in many studies of IQ scores was not, that the environments of the test subjects varied in ways that had influenced the scores of the tests in consistent and apparently misleading ways - that happen to be genetically correlated, we note.

 

 

That's not how shared environment is measured in most IQ studies to my knowledge. There are adjustments for socioeconomic status, but if we take the typical twin study, the shared environment is estimated from monozygotic twins by subtracting out the effects of genetic variance, since monozygotic twins share to a large degree the same environment. This method is very robust to such effects and then there are the assumption free methods in genome-wide association studies that estimate the heritability directly from the genome. I linked to such a study previously.

 

Furthermore, as I pointed out to you twice now, the factors studied by Steele and Aronson have nothing to do with genetics. There results are not correlated with genetics either. In measuring heritability, race and sex are not used as proxies of genetics. I pointed out to you earlier how something like twin studies, where the individuals in the study are of identical race, are an inherent control for any such variance. In fact, if two twins scored differently because of their self-perception, this would fall out in the study as a random variable...i.e. the unique environment.

 

Another example: studies have shown that a given family in a given house is not necessarily a shared environment for the various children - the different ages, the different sexes, the different birth orders, the different physiologies and illnesses, all have their influences - they have to be controlled, if they are relevant to the aspect being measured and you want to reveal the genetic component in isolation.

 

 

Indeed, such things are controlled for in sibling studies. This is actually an inherent strength of twin studies because many of these factors are controlled for automatically. Of course, there are the aforementioned adoption studies, where the basic study design automatically leads us to the assumption that the environments are not identical, allowing for even more robust and reliable measures of heritability. And again, we are left with the results of genome-wide association studies that allow us to measure heritability directly from the genome.

 

That's out of context, and probably (from his writings, I don't bother with youtube videos) doesn't mean quite what you seem to think -

 

No its not out of context. There is no other way to interpret the claim that "But there is a remaining 20% that depends on external circumstances". What do you think he means by external circumstances if not the environment? And if ~20% of the variance depends on the environment, where does the remaining ~80% come from...magic? Is it mere coincidence that this figure matches perfectly the estimates of heritability in height being ~0.8?

I normally don't bother with youtube videos either, but then I asked you to give me specific references, I frankly don't have time to read through all of his papers on the matter, I have a stack of other papers with higher priority. If you don't like youtube videos, then you surely can take the time to give me specific papers to back your claim.

one of his key findings was that the intergroup variation in human height seems to be almost completely environmental, and only within a given shared environment does a large contribution of genetics to variance show up - he now recognizes a shared environment by the group's median height and variance, in other words -

 

Define for me what you mean by intergroup? I have stated several times that I am not making any argument for the variance between populations, but within populations. Every geneticist worth his salt can tell you cant make assumptions about the differences between populations based on the differences within populations. But such differences does not negate the heritability of a trait within the population. You seem to be arguing against claims that I have never made.

You seem to be arguing in the opposite direction - assuming, as people talking about human height did for centuries (and concluded that Chinese were shorter than Englishmen by inheritance, for example) that you can recognize a unique environoment, accurately identify a shared environment.

 

 

That is not at all what I am claiming. Again you are arguing against genetic effects between populations. You also seem to be confused on what I mean by shared environment. Shared environment does not have to be at the population level. If we consider a typical twin study design, the shared environment is intrinsic to each pair of twins...namely each pair of monozygotic twins. You could have a 100 different twin pairs with a 100 different shared environments and still estimate the effect of the shared environment. This is because of how the comparison is done, with the measure of shared environment being from the variance of the twin pairs themselves, not across the entire population.

 

Another issue in comparing IQ and height is that the variance in IQ is the only visible aspect - nobody knows what the standard human minimum allotment of IQ looks like, or what the maximum would be, or even if a linear and one dimensional scale is appropriate.

 

 

Irrelevant. What these studies are trying to explain is the variance itself, not the minimal value. If the minimal human IQ was 80, we would not necessarily be interested in what causes that first 80, but why one person has 85 and another has 95.....we are interested in the variance between that 85 and 95, not the first 80 points.

 

Still another issue is the instability of IQ scores - human height does not vary from day to day, or measurement device to measureement device, by any large fraction of the population variance or the available range for a given individual.

 

 

Sure it does. Human height varies with age, and the measurer/device (its called technical error). Thats why we do not rely on measurements from single individuals, but use larger sample sizes, whether for height or IQ.

 

These issues become critical, to the point of dominance, when researcher begin to postulate feedback mechanisms - when the "intrinsic" IQ is supposed to be governing the environmental influences, selecting opportunities for its own enhancement.

 

 

 

Well in that case, ~150 years of genetic research is complete bunk....right? Of course not. The point is that countless studies have demonstrated that IQ is very heritable. The irony being, that postulated feedback measurements means that overtime, genetics plays an increasingly important role over environment.

 

I look forward to you supplying me with some actual references in your next post.

Edited January 28, 2014 by chadn737