Posted by shar on October 28, 2000, at 21:42:03
In reply to Re: LONG answer!!!, posted by pullmarine on October 27, 2000, at 2:04:38
> I have read that heritability is the proportion of variation of a trait that can be attributed to genetic factors as opposed to environmental ones. Can anyone give me a good idea of what that really means?
BTW, just kidding with the title of this post.
In statistics, people are always looking at variance (differences) and trying to explain what proportion may be due to factor a, factor b, or just random fluctuation in the data.
If there was no variance then everything would be the same for everyone, and the question would be answered. But, since there is always random fluctuation (caused by human error, the weather, if the respondent had a headache) one has to account (or try to) for proportions of variance due to the different sources.
So, the researchers said that the proportion of variance due to familial factors was modest in men and significant in women. (My simple interpretation.) They don't say there was a statistically significant proportion due to familial factors, which is pretty important in most social science research. Significant would mean that whatever percentages they found were not just due to chance, or random fluctuation in the data, but really due to the familial factor.
The best I could say would be it is probably significantly familial in women (but did the authors say that?). However, that is not the same as predicting the probability of a child having depression if one or more parent is depressed. All I would be willing to say from what I read in your post is that if you have a depressed female, you are significantly more likely to find that one or both of her parents experienced depression.
I am a cautious statistician. Pullmarine made some good comments on caveats as well. But since we didn't read the article (or at least I didn't) it's an educated guess.
Shar
>
> >In very general terms, the explanation for the first question is as follows (it might be easier to understand with a prinout):
>
> Genes may very well predispose certain characteristics (ie. genotype. is the gene there?for instance the genes for dark skin), but the characteristic may or may not come into existance unless certain environmental cues are present (phenotype. can you see the gene in action. is the skin dark? a person may be genotypically black, but phenotypically white).
>
> For many genes, environmental cues will trigger the genes into action (ie' taning, or some animals growing white fur instead of brown during winter due to changes in temperature).
>
> The observable in traits is called phenotype.
> Often, genes allow a limited range of possible oucomes that are expressed due to environmental factors.
>
> In other words, genes are the constraints within which different outcomes are possible depending on other factors (ie, environmental). a good analogy would be climatic conditions (don't try growing equatorial plants at the north pole, or don't expect snow at the equator)
>
> thus, heritability (the genetic constraints and range of possibilities that we get from our parents)is the proportion of difference in a trait (for instance, skin color) that can be attributed to genetic factors (how closely does the skin color of the offspring match that of the parents?)as opposed to environmental ones (how much of the variation is caused by sun exposure?). You can find that the genes responsible for skin color (there are eight of them if I remember correctly) are inherited. The proportion of variation in skin color is defined largely by these genes, but that sun exposure is a contributing environmental factor. A white person can tan but will not become black regardless of sun exposure, and vice a versa. Thus, skin color is very much an inherited trait Related to the family, not the environment).
>
> With depression, you will find that life circumstances (ie, environmental factors) will play a significant role in causing depression, and that genes may/or may not be partially responsible. It is very hard to assess whether depression is a natural response to life events, or possibly a learned behavior, or whether it's genetic.
>
> I would assume that the converse would then be true; in females genetic factors play a larger role.
>
> >No. because environmental factor do not affect men and women in the same ways. overall, women are taught to be caring, self- sacrificing, and face many hardships and responsibilities that most men don't face. Furthermore, women are usually more verbal about their feeling, thereby accounting for the higher numbers.
>
> > Does this also mean that the daughters of a depressed person are more likely to inherit a tendency towards depression than are the sons?
>
> No. It can be interpreted as an indication that women are more prone to depression than men. But it may very well be a confounding factor (statistical term), meaning that the rates are higher for women because they are better at reporting their feelings than men are, or that they are under more stress.
>
>
> >Does anyone know if there are statistics that say (for example), “If a person has one parent who has experienced depression, he/she has an X% chance of experiencing depression him/herself and if both parents have been depressed, then he/she has a Y% chance of experiencing depression”?
>
> >
> I've read that if one parent is manic-depressive, the chances are 25% for off-springs
> and that if both parents are, the chances increase to 75%, (indicating a dominant gene, see mendel or mendelism or mendelian genetics). But I truly doubt the accuracy of these numbers. source of percentages: touched by fire. k r jamison
>
>
> When you look at statistical data, these some of the questions to ask:
>
> Was the sample representative?
> was the study double blind?
> were the surveys misleading in anyway?
> Margin of Error.
> How were the surveys given, by whom and under what conditions.
> Were there any confounding factors?
> What are the mean and standard deviations?
> Results for the Null Hypothesis?
> Who is doing or funding the research and to what purpose?
>
>
> JOHN
>
>
> PS, check out the rand study on texas's alleged education progress under bush, i also recommend a short funny book called 'how to lie with statistics'
poster:shar
thread:1590
URL: http://www.dr-bob.org/babble/social/20001011/msgs/1786.html