Playing fast and loose with science

Following the comments by Prime Minister Goh in Time magazine about having gay persons in the civil service (see the article Gay civil servants, and what next?) the gay issue has blown up in Singapore like never before.

The fundamentalist Christians have been vociferous in their objections. They particularly disliked the Prime Minister saying that gay persons are "born that way".

The Church of Our Saviour (COOS) sent a letter to the various newspapers, entitled "Born gay? There is no scientific proof", but it wasn't published by any of them. We heard that they were extremely upset. So, in dogged pursuit of their Fundamentalist Agenda, they put it up on their website, together with another letter, "The truth about being gay". Both these letters are put here for easy reference, in the 2nd and 3rd of the right-side boxes. Word came to us that COOS planned to circulate these letters to Members of Parliament and other influential people.

They also encouraged individuals to write to the press, etc, using the same template. You can see an example of that in Thio Su Mien's letter to Streats, which I have archived among the various letters in Letters in Streats, 15 - 24 July 2003.

My own views about the "Born gay?" letter can be seen in the article COOS pastor "wants a nation of righteous Christians". In it I pointed out how badly COOS misread the scientific data.

But here is a completely thorough job by C K Ho. He went to the library and looked up what he could find of the references that COOS made in their two letters. He has written in his report his observations about what those scientific studies actually said. Two other persons (who do not wish to be named, because they are not out) went with him and read the same reports. They performed the necessary quality control review of C K Ho's essay.

To put it bluntly, C K Ho's findings are quite damning (the worst indictment, I have highlighted in red). They show up how scientific findings are misused.

But even outside of the inaccuracies, there's a certain tone in the letter "The truth about being gay" which is worth pointing out. Many would be amused by the outrage expressed by COOS: they are having so much sex! Sensible people who aren't so prudish (the type without their heads in the sand) might ask, "so what? And if that's what people want to do, and they are consenting adults, who are you to say they mustn't?"

Moreover, if they are having so much recreational sex, having so much fun, how does one say, that they are "anything but gay" (paragraph 2 in that letter)? These fundamentalists don't get it: Sex can be enjoyable!

© Yawning Bread 

Report by C K HO

Dear Sir,

I refer to the two letters from Church of Our Saviour ("COOS"), "The Truth about being Gay" and "Born Gay? There is no Scientific Proof", Upon a thorough read, I have taken upon myself to obtain most of the original articles cited by the author(s) in the two letters and have proceeded to analyse these scientific papers.

1. In their article, Theo G.M. Sandfort et al's (2001) findings of high incidences of mental health illnesses amongst the homosexual population concurred with previous studies. This was as stated in the letter "The Truth about Being Gay". However, an examination of the data showed that there was no consideration of the psychosocial factors that may influence the results, in this case, the manifestation of psychiatric disorders in the homosexuals. Any study of sound rigour minimises all possible extraneous factors. If this is not possible, a good study will perform statistical correction to ensure minimal skewing of the results due to these factors. Sandfort et al failed to do as they merely assume that such psychosocial factors that may increase the prevalence of psychiatric disorders in the homosexual population do not exist or is insignificant in the Netherlands, a fundamental flaw in scientific researches as the assumption is not validated.

   In a study by Safren and Heimberg (1999), not quoted in the COOS letters, the researchers found that without correction for the psychosocial variables, the homosexual subjects studied did have a higher prevalence of psychiatric disorders. However, upon statistical correction for the psychosocial variables, there were actually no statistically significant differences between the homosexual and the heterosexual population. This construct is more valid as it takes into account the effects of the psychosocial variables, suggesting that the higher prevalence rates as a result of psychosocial influences, and not the orientation per se.

   In the study by Sandfort et al, the classification of homosexuality was based on the respondents having had sexual acts with someone of the same gender in the preceding year and not based on their orientation. The underlying danger of such a classification is the homosexual population studied may include subjects who have yet to reconcile with their homosexual orientation. In using such a classification, Sandfort et al would have an inherent selection bias error in their research as the subjects that are classified as homosexuals based on behaviour and not orientation may possess psychiatric disorders due to the inability to reconcile with their sexual orientation.

   Bailey et al (2000) explained in his study that current researchers "tend to define sexuality psychologically rather than behaviourally" (pg 524). The reason is because sexual behaviour and sexual orientation may not correlate perfectly as expected due to many factors, particularly social. This demonstrates the error in applying Sandfort et al’s research to all homosexuals.

2. In the article by Leigh et al (1993) quoted in the letter "The Truth about being Gay", the author(s) claimed "79% of homosexuals reported being sexually unfaithful in the previous year with 10% among married heterosexuals and 23% among cohabiting heterosexuals". A thorough critical analysis of the paper failed to reveal any data reported comparing homosexual and heterosexual sexual behaviour. In fact, Leigh et al clearly stated in their paper that "Data are not presented separately by sexual orientation......" (pg 1402)

   Furthermore, the statistics pertaining to infidelity as reported in the original COOS article was not considered at all in the Leigh et al study. The closest questions were one on frequency of sexual intercourse in the last 30 days, 12 months and 5 years and another on the number of sexual partners in the last 30 days, 12 months and 5 years. Again, the statistics presented in the paper were shown comparing, men, women, races, marital status, age group and education level. There were no categories comparing homosexuals and heterosexuals with regards to any of the questions. The authors made a note that a high number of sexual partners is not indicative of sexual infidelity as the participants may be practising "serial monogamy", supported by the fact that same study showed that 85% of the participants who indicated that the partner that they had sex with in the last 12 months was their primary partner. It is of interest therefore the source of information with respect to the statistics for infidelity amongst the homosexuals as cited by the author(s) is without basis as a careful dissection of the study by Leigh et al (1993) does not contain the figures as quoted.

3. In the reference by Bell and Weinberg (1978) quoted in the letter "The Truth About Being Gay", the author(s) cited that "74% of homosexuals reported having more than 100 partners". The study was done 30 years ago and the extrapolation that the homosexual population's sexual behaviour then and now is the same is not validated. Contrary to that, in a different long-term prospective study carried out more recently, Johnston et al (1999) looked into the sexual risk behaviour of homosexual men in 1984/85 and 1995/96. They found that the percentages of homosexual young men who had more than 10 partners in 1984/85 was 55% and in 1995/96 was 16%, showing clearly that the homosexual male population’s number of partners have reduced very significantly from the ‘80s to the ‘90s.

4. 4) In the letter "Born Gay? There is No Scientific Proof", the author(s) states that the concordance rate for both of the identical twins being homosexual was only 52% and 38% in two separate studies by Bailey et al (1991, 2000), far from the expected 100%. Mentioning only identical twins the letter failed to state that the same studies compared the incidences of homosexuality in adopted brothers, non-identical twin siblings (dizygotic twins) and identical twin siblings (monozygotic twins). Bailey et al (1991) found that the figures were 11%, 22% and 52% respectively for adopted brothers, non-identical twins and identical twins, suggesting a strong genetic linkage, unlike the conclusion drawn by the author using the same reference.

   In a subsequent similar study carried out by Bailey et al (2000), the concordance rates for similar homosexual orientation amongst identical twin for men were 37.5% on a lenient criterion of homosexuality and 20% on a strict criterion. In comparison to non-identical twins, the statistics were 6% and 0% respectively. The authors clearly stated in their article that though the results were statistically less than the earlier findings in 1991, the results do not show that there is no genetic hereditability. In addition the authors continued to clarify that this second study was largely limited by having insufficient number of non-heterosexual subjects, particularly homosexual subjects, causing a poor ability to achieve high statistical power, a feature crucial in all scientific studies to reach a conclusion based on the results. Again, this conclusion arrived by the researchers differed from the author(s) from COOS.

   It is now known widely that the physical presentation (phenotype) of an individual is an extremely complex interaction of the genes, alone or in combination with other genes, intra-uterine environment and the external environment. The contention for the absence of a single gay gene simplifies the genetics of human sexuality. Many of our phenotypic expressions are a result of the additive effect of cumulative genes.

   It is with much regret that the scientific papers cited in the two letters by the author(s) from COOS, have been highly misrepresented to provide a simplistic view of a highly complex phenomenon. Deliberate selection of studies that merely represent one point of view without considering contradictory or alternative studies lacks in intellectual integrity. The papers selected should also be methodically analysed to ensure good research methods, crucial in ensuring accurate and valid results. These papers should be analysed as an entity, not in parts, to understand the results of the hypotheses that the researchers have made and seek out to prove or disprove.

   We thank you for your kind attention.

   Yours humbly
   C. K. Ho

Footnotes

None

Addenda

None