Skip to content
 

Homosexuality and the number of older brothers and sisters, or, the difference between “significant” and “not significant” is not itself statistically significant

This paper, “Biological versus nonbiological older brothers and men’s sexual orientation,” by Anthony Bogaert, appeared recently in the Proceedings of the National Academy of Sciences and was picked up by several news organizations, including Scientific American, New Scientist, Science News, and the CBC. As the Science News article put it,

The number of biological older brothers correlated with the likelihood of a man being homosexual, regardless of the amount of time spent with those siblings during childhood, Bogaert says. No other sibling characteristic, such as number of older sisters, displayed a link to male sexual orientation.

I was curious about this–why older brothers and not older sisters? The article referred back to this earlier paper by Blanchard and Bogaert from 1996, which had this graph:

sibs1.png

and this table:

sibs3.png

Here’s the key quote from the paper:

Significant beta coefficients differ statistically from zero and, when positive, indicate a greater probability of homosexuality. Only the number of biological older brothers reared with the participant, and not any other sibling characteristic including the number of nonbiological brothers reared with the participant, was significantly related to sexual orientation.

The entire conclusions seem to be based on a comparison of significance with nonsignificance, even though the differences do not appear to be significant. (One can’t quite be sure–it’s a regression analysis and the different coef estimates are not independent, but based on the picture I strongly doubt the differences are significant.) In particular, the difference between the coefficients for brothers and sisters does not appear to be significant.

What can we say about this example?

As I have discussed elsewhere, the difference between “significant” and “not significant” is not itself statistically significant. But should I be such a hard-liner here? As Andrew Oswald pointed out, innovative research can have mistakes, but that doesn’t mean it should be discarded. And given my Bayesian inclinations, I should be the last person to discard a finding (in this case, the difference between the average number of older brothers and the average number of older sisters) just because it’s not statistically significant.

But . . . but . . . yes, the data are consistent with the hypothesis that only the number of older brothers matters. But the data are also consistent with the hypothesis that only the birth order (i.e., the total number of older siblings) matters. (At least, so I suspect from the graph and the table.) Given that the 95% confidence level is standard (and I’m pretty sure the paper wouldn’t have been published without it), I think the rule should be applied consistently.

To put it another way, the news articles (and also bloggers; see here, here, and here) just take this finding at face value.

Let me try this one more time: Bogaert’s conclusions might very well be correct. He did not make a big mistake (as was done, for example, in the article discussed here). But I think he should be a little less sure of his conclusions, since his data appear to be consistent with the simpler hypothesis that it’s birth order, not #brothers, that’s correlated with being gay. (The paper did refer to other studies replicating the findings, but when I tracked down the references I didn’t actually see any more data on the brothers vs. sisters issue.)

Warning: I don’t know what I’m talking about here!

This is a tricky issue because I know next to nothing about biology, so I’m speaking purely as a statistician here. Again, I’m not trying to slam Bogaert’s study, I’m just critical of the unquestioning acceptance of the results, which I think derives from an error about comparing statistical significance.

7 Comments

  1. Anonymous says:

    I'd be interested in how twins would factor in here. After all, the difference in their birth order is almost zero, so I'd be curious how sensitive their sexuality is to switching their birth order.

  2. Martyn says:

    Not being a subject matter expert either, I still have strong reservations over Blanchard and Bogaert's logistic regression model. For the effect of the number of older brothers, the counterfactual question it is addressing is this:

    What would have happened (in terms of sexual orientation) if the subject had had an extra sibling who was both older and a boy?

    But this imaginary experiment also changes the total family size and the birth order. If you want to address the theory that the prenatal environment is modified by previous male, but not female, foetuses then you should be asking this:

    What would have happened if one of the older siblings (if any) had been a different sex?

    This could be answered by a model with the following terms:

    Total number of siblings
    Number of older siblings
    Number of older brothers

    The last term would then give the odds ratio for number of older brothers, controlling for family size and birth order.

    The more recent paper by Bogaert in PNAS generated a lot of publicity because he was coming down on one side of the eternal nature/nurture argument by comparing the effects of biological vs non-biological older brothers. But he never made an explicit comparison. It is not convincing.

  3. another anon says:

    Bogaert's original study was strongly criticized. Many pointed out that siblings' influence on each other, particularly the older male sibling on the younger, is the most probable cause of any statistical showing that younger male siblings were more likely to be homosexual than the average in society.

    So he got grant money to do a follow-up. This time he would study male siblings where the other family siblings are adopted or from two-parent families. His statistics show that in these situations there is no trend towards homosexuality. Therefore, it must be because of the mother's hormones.

    Nonsense, again. For all these siblings know who is biologically connected. The emotive reaction to genetic older siblings (on the part of all family members) is bound to be different. Not only is there knowledge by family members. Those outside the family know it. And the intimacy of close contact from birth is bound to differ in such situations. In fact, if you take away his biological connection, it would tend to prove the opposite, i.e., that older male siblings' dominant relation and the inner-family dynamic of desire for approval from the biologic parents may have some effect on a homsexual orientation.

    This should lead social scientists to compare this situation with studies showing the relationship between first borns and older siblings with younger ones, rather than jumping to biologic conclusions.

    Note also, that all of this is just a wild assumption by someone trying to prove a biological connection (which no one has ever shown) by a sociologic statistic. His answer is: "It must be because …"

  4. It seems to me that the big problem with this study is that it doesn't address whether being an adoptee might have an effect upon being gay. I can well imagine that in a study where over half the respondents are from mixed families (and over 40% of these are adoptees) that not having many older nonbiological siblings would be associated with being an adoptee. Indeed, the adoptions website, http://encyclopedia.adoption.com/entry/only-child… says that "There are indications that as many as 50% or more of all adoptive parent couples (or singles) adopt one child only". And I can imagine many families who adopt adopt because they are childless. Thus, if adoptees are more likely to be gay, say because adoptees tend to have abusive biological fathers with gay tendencies, or because adoptive mothers aren't as careful about caring for adoptees, and neglect tends to cause gayness, this would give a much simpler explanation for his conclusion.

  5. Anonymous says:

    Stephen A Meigs has it backwards. It's people who come later in birth order who are more likely to be gay, not earlier. What he's suggesting would have created a bias in the data against the conclusion that was reached if the effect he predicts here is significant at all.

  6. No, I don't have it backwards. I was talking about nonbiological older siblings. The paper concludes not only that having older biological siblings is associated with gayness, but also that having nonbiological older siblings is not associated with being gay. Actually, the methodology is even worse than I suggested. Though the definition ("e.g., with adopted or step-siblings or as an adoptee") of blended family is vague in the paper, in the fourth, largest sample (55% of total), it would appear from the description of the sample that essentially every person who was raised without siblings was an adoptee. Therefore, those raised without any older nonbiological siblings would very much more tend to be adoptees, and since that is a most statistically significant group, the data may suggest mothing more than that being an adoptee makes one vulnerable to being gay and that being raised with older siblings (of any sort) makes one vulnerable to being gay, and the reason the data didn't suggest to the author the latter in the case of non-biological siblings is merely that the effect was canceled by the adoptee effect. For instance, if being gay is associated with having been forcibly sodomized, then since older male siblings have more power to forcibly sodomize younger male siblings than vice-versa, an association between gayness and being raised with older male siblings would make sense, and is not surprising to me. Perhaps I should have made it clearer what part of the paper I was arguing against.

  7. Andrew says:

    I think that various hypotheses are consistent with the data. The subject matter is so far from any of my expertise, I merely was pointing out that the difference between the "signif" and "non-signif" findings is not in itself statistically significance. Which actually opens the door to more hypotheses being consistent with the data shown. Further study is needed etc etc.