Home Page
Comments to: -pierre@youth-suicide.com
Educ. Index
MHN Badge
Search
for

Get a Search Engine For Your Web Site

To: Gay Page Award. Want one?
 
Caveat Alert For Studies
Soliciting Homo-Info
 

Web Page Summary: In 1998, the result of two studies indicated that a serious problem has existed in demographic studies based on sexual orientation, but also in studies soliciting data much more taboo than the now more acceptable- but still taboo and very threatening to many - "homosexuality" information: information related to men who have been - or are - sexually active with female and/or male children younger than the age of 12 or 13. In an attempt to get related information from men in random surveys, Dr. Bagley began using computer technology to solicit related information in the early 1990s. The method proved to be effective and was even more effective at having young adult male subjects report on other lesser taboo homosexual activity between consenting individuals and their often stigmatized and closeted homosexual/bisexual  identity - or self-labelling/identification. The Bagley & Tremblay (1998) demographic results were accompanied by this statement: "Studies producing 200 to 800 percent underestimates for male homosexual realities should be unacceptable as social science research" (Journal of Homosexuality, 36(2), p. 14). The Turner et al. (1998) study confirmed what was indicated by the Bagley & Tremblay (1998) results. From the abstract: "As a group, however, there was a "fourfold" increase (400%) in males [age 15 to 19] reporting homosexual activity when the computer-audio system was used: from a 1.5% to 5.5% positive response rate.."  The implications of these results are monumentally important, especially with respect to the results of all studies comparing attributes (such as suicidality) of youth identified by their reported (acknowledged or admitted) "homosexual" attributes.
 

Caveat for studies soliciting "taboo" information (socially disapproved behaviors, including illegal behavior) from adolescent and young adult males. The "pencil & paper" method produces great underestimates.

In 1994, Bagley et al. reported on the "at risk" nature of males who were "sexually abused," defined to have experienced "unwanted sexual acts."  The study of a stratified random sample of 750 18- to 27-year old males had used a portable computer to solicit "taboo" information [See Bagley & Tremblay  (1996) - published as Bagley & Tremblay, 1998 - for details]. The indications were that this method for soliciting sensitive taboo information related to child sexual abuse, for example, was more effective than other methods previously used. One result of using this method was that, as adults (reported in Bagley et. al., 1994), 1% of males (8/750) reported having had sex with children less than 13 years of age; this is a highly "taboo" illegal activity and males are also unlikely to report related desires, but 4.5% of males (34/750) did report having sexual desires for children. This is not information adult males would readily reveal to anyone (probably not to a researcher in a "face-to-face" situation, nor to anyone requesting such information on the telephone) and they may also be quite reluctant to disclose related information on paper.

Therefore, the data soliciting method used by researchers may make the difference between a study reporting that a certain type of "taboo" behavior is 'non-existent' in a human group (asking men "face-to-face" if they have has sex with children, or even asking men related questions on the telephone), and another study reporting related data. A major research question must therefore focus on the reliability of research studies reporting the extent of "taboo" behaviors in society, especially for the studies reporting the percentages of teenage or adult males homosexually active with (1) older males (may be intergenerational), (2) plus/minus same-age males, (3) younger males (may be intergenerational), and pre-puberty males (may be pedophilia). All such activities are "taboo" and the "taboo" degree will vary depending one one's immediate environment - such as one's culture or subculture - and life circumstances), noting here that homosexuality was decreed to reflect "mental disorder" in North America until 1973-4. Homosexual behavior also has a history of being illegal, and the illegality of such activities still exists in about 22 American states. The decriminalization of male homo-sex behaviors only began after 1960 in North America.

On the basis of the "admission" results related to men sexually interested in children in the Bagley et al (1994) study, the demographic part of the study (Bagley & Tremblay, 1998; available on the Internet by 1996) was predicted to most likely have decent reliability with respect to its effectiveness at soliciting less taboo information, such as men reporting being either homosexual or bisexual (Many males in these categories are still guarded with respect the ones to whom  such 'identity" information will be revealed.), and with respect to men reporting being homosexually active (having sex at least once with a male in the last six month). These men may quite reluctant to let anyone know about such activities, although their male sexual partner(s) may "know," but they may have been in the "anonymous" category, and their reluctance may be great if these homo-active males are married to a female and are fathers.  See Bagley & Tremblay (1996)  for the results indicating that about 10% of males (both on the basis of self-identification and of being currently homosexually active) may be deemed homosexual or bisexual. The study results also challenged the 1-3 percent estimates of recent studies done using face-to-face interviews; they were deemed to produce highly unreliable "male sexuality" estimates in the [gross] "underestimate" category ranging from about 400% for homosexual behavior and up to 800% for self-identification as homosexual or bisexual. In this respect, Bagley & Tremblay stated: "Studies producing 200 to 800 percent underestimates for male homosexual realities should be unacceptable as social science research" (Bagley & Tremblay, 1998, Journal of Homosexuality, 36(2), p. 14).

In May, 1998, the following study was published and it reports on the response differences between two methods used to solicit taboo and illegal behavior information from adolescent males ranging in age from 15 to 19 years:

Turner CF, Ku L, Rogers SM, Lindberg LD, Pleck JH, and Sonenstein FL (1998). Adolescent sexual behavior, drug use, and violence: Increased reporting with computer survey technology. Science Magazine, 280(5365-8), 867-73.

Abstract Access:  Access to the abstract may be made via the Science Magazine web site (http://www.sciencemag.org), but registering at the site is required - free.  Use this URL (http://intl.sciencemag.org/) for faster access in these countries: Australia, Brazil, China, France, Germany, Hong Kong, Israel, Japan, The Netherlands, Russia, Singapore, South Korea, Taiwan, and the UK. A document delivery service is also a available and a fee is charged to access the full text of the article. When in the site, use this link to immediately access the abstract. (http://www.sciencemag.org/cgi/content/abstract/280/5365/867) or (http://www.intl.sciencemag.org/cgi/content/abstract/280/5365/867) if you are at the international web-site. (PubMed abstract, Full Text.)

Research Highlights:

Sample: A random survey of 1672 males (number used for analysis) aged 15- to 19-years old living at a fixed address in the continental United States from February to November, 1995.

Study: Subjects were asked a number of questions, including questions relating to same-sex activity and illicit drug use. This was done using a pencil and paper method, and also via computer, supplemented by a verbal rendition of the questionnaire heard through headphones.

Results: The difference in "the effectiveness" of the two method for collecting sensitive data was remarkable. For example, "no male who was behind in school reported any of the six male-male sexual behaviors when completing the paper SAQ. In the audio-CASI condition, 6.2%... (answered affirmatively on one or more of these questions).  As a group, however, there was a "fourfold" increase (400%) in males reporting homosexual activity when the computer-audio system was used: from a 1.5% to 5.5% positive response rate; the homosexual behavior with the greatest reporting difference (eightfold, adjusted) was to the question "Ever had receptive anal sex with another male (his penis in your rectum or butt)": 0.1% to 0.8%.  For illicit behaviors such as "ever shared needle," the difference was about tenfold (adjusted): from 0.1% to 1.1%. For "ever taken street drug using a needle: from 1.4% to 5.2% - a fourfold increase.

Discussion: .It was noted that differences in response response rates generally did not occur with the two data intake methods when soliciting information not deemed to be illicit or taboo, thus leading to the conclusion that, at least for adolescent males 15- to 19-years of age, something 'special' occurs in their mind (in terms of withholding taboo information) when it is requested asked for this information via a paper-and-pencil questionnaire, and the difference is most significant for males behind in their grade level. The implications are that these boys - in addition to being like other boys in terms of withholding sensitive information - may also have difficulty reading and this problem would be overcome with the audio part of the computer-based data soliciting method. Furthermore, using expressions like "having a penis in your butt" is also a way to solicit information in a most non-ambiguous ways; males are not likely to misinterpret the question.

THE GREAT IMPLICATIONS FOR THE CDC's YOUTH RISK BEHAVIOR SURVEYS (YRBS)   and   THE RELATED PUBLISHED  1998  ANALYSES CONFIRMING THAT SELF-IDENTIFIED GAY/BISEXUAL MALE YOUTH   or  MALE YOUTH REPORTING BEING HOMOSEXUALLY ACTIVE  are  AT (STATISTICALLY SIGNIFICANT) GREATER RISK FOR A NUMBER OF PROBLEMS, INCLUDING SUICIDE ATTEMPTS:

The YRBS are pencil-and-paper questionnaires, meaning that they it would be one of the least effective method to use (the worst probably being face-to-face interviews)  for identifying adolescent males who have been engaging in illicit activities (such as street drug use) and highly taboo (possibly illegal) behaviors such as homosexual activity.  On the basis of the Turner et al. (1998) study summarized above, an informed conclusion would be that the Youth Risk Behavior Surveys are grossly underestimating male (and possibly female) homosexual activity in adolescence. Therefore the result reported by Faulkner & Cranston (1998): Massachusetts's 1993 YRBS study(homosexual activity reported by 3.5% of male and female adolescents (106/3054) and Durant et al. (1998) (1995 Vermont YRBS study: 2.6% of males reporting homosexual activity) would be underestimates. If, for example, the 4-times underreporting factor from the Turner et al. (1998) study is used, the "2.6%" estimate for boys being homosexually active in Vermont would lead to the informed speculation that about 10.4% of adolescent boys (age 15 to 19) are homosexually active in Vermont, but most (75%) did not identify this fact on the pencil-and-paper questionnaire given to them for the YRBS study. This "10.4%", however, is higher that the 5.5% estimate reported by Turner et al. (1998). How could this be? There are many factors which can influence study results such as, for example, the fact that the Vermont study is a school sample and the Turner et al. study is a random sample requiring living at a fixed address. Gay adolescents, however, are at risk for being street youth and they may also be living with a man who, when contacted by the ones doing a random survey, may not revealed that the requested adolescent male was residing in his home. Such life situations, if they applied to some of the surveyed households, would lead to underestimates of homosexually active males in the adolescent population. On the other hand, a similar "underestimate" would also apply to school samples given that homosexually oriented youth are at greater risk for not being at school - thus increasing the likelihood of not being present the day the survey was administered, thus resulting in an underestimate for school samples.

There are many factors which will affect estimates for homosexually active male adolescents, and others which will affect estimates of homosexual- and bisexual-identified males. Furthermore, all such samples have been used for statistical analyses to establish, for example, that gay/bisexual male adolescents are significantly more at risk for suicide attempts than their heterosexual-identified counterparts. The Remafedi et al. (1998) study, for example, left out male adolescents "not sure about their sexual orientation" from the analysis. Therefore, although it could be said that 2% of gay/bisexual males with a lifetime attempted suicide rate of 28% would not account for 30% of the male youth attempted suicide problem, the "2%" would represent a small subset of the estimated 10 to 12 percent of homosexually oriented males who, let's say by the age of 18 to 27 (average age of 22.7 years, as reported by Bagley & Tremblay, 1997), would then account for about 25% of the males who reported self-harm activities in their lifetime, and 62.5% of males reporting a deliberate attempt at self-killing. A number of reasons account for the great differences between the spectacular Bagley & Tremblay (1997) "suicide attempt" results and the  more modest results resulting from studies of adolescent male:

1. In studies of adolescent male samples where "being homosexually active" is used as the study's sampling and analysis criteria, males who are not homosexually active but already self-identify as gay or bisexual (and are having related problems, such as attempting suicide), are either excluded from the sample or they are placed in the "heterosexual" group simply because they report having had "sexual intercourse" with a female and not with a male.

2. In studies of adolescent male samples where "self-identifying as homosexual/gay or bisexual" is used as the study's sampling and analysis criteria , males not in this category - but are homosexually active - and may later keep on being homosexually active with or without self-identification as homosexual/gay or bisexual and report that they were having serious problems (even including suicide attempts), will either be left out of the analysis or be placed in the "heterosexual-identified" category because they identified as heterosexual at the time. This would happen because these boys are desperately attempting to make themselves believe they were not homosexual or even bisexual, in spite of the behavioral evidence. Many men in many cultures are homosexually active, but only in dominant "insertive" ways; they behave sexually with a male as they would with a female, and identify as 100% heterosexual. In other words, their definition of homosexual or bisexual is not the one assumed by researchers. Another example of this happens when boys have somehow accepted the common belief that homosexuals are only men who have sex with boys; they therefore cannot report identifying to be homosexual. Boys who also commonly believe that gay (or homosexual) boys are effeminate will not report being homosexual (or even bisexual, meaning that they are effeminate only a part of the time) even if they are having sex with other boys and even men.

3. Adolescent male samples studied may have an average age of 15 or 16 year if the sample ranges from grade 8 or 9 to grade 12, meaning that about half of the gay/bisexual males who attempt suicide during their adolescent lives have not yet done so. Only a study of older males, such as the one carried out by Bagley & Tremblay (1997) (but even better studies may be designed), will render a decent 'picture' - or estimate - of what was the "suicide attempt" situation was during adolescence for homosexually oriented males compared to heterosexual males; the Bagley & Tremblay 1997 results also includes a few years into into young adulthood (to the age when they participated in the study: 18- to 27-years, average = 22.7 years, therefore an average of 3 years beyond adolescence) for most of the sampled males..

To date, the Bagley & Tremblay (1997) results form the best estimate of the proportion of North American adolescent male suicide attempter who are homosexually oriented (62.5%), and the study also replicated the Bell & Weinberg (1978) suicide attempt relative risk factor for homosexual males compared to heterosexual males. This estimate was obtained using a stratified random sampling process and "being homosexually oriented" was determined on the basis of subjects being either currently homosexually active (in the past six months) and/or by self-identifying as "homosexual" or bisexual", with an approximate 80% overlap for the two categories. A suicide attempt was defined to be "an attempt at self-killing," which may be different than the "suicide attempt" criteria used in adolescent samples. Often enough, self-harm activities (which are not be attempts at self-killing) are listed as suicide attempts - even by adolescents. Furthermore, the Bagley & Tremblay study used computer technology (the same as the Turner et al. (1998) study, but not including the audio via headphones) was used to solicit relevant "taboo" information from subjects, meaning that the likelihood of not identifying "homosexually oriented males" was reduced, thus improving the reliability of results.

The Turner et al. (1998) study therefore serves as a serious caveat for all pencil-and-paper surveys involving the request for "homosexuality" information from adolescent (and probably young adult male) subjects studied and the subsequent analysis of the "at risk" status of homosexually oriented adolescent males compared to their heterosexual counterparts:

The "at risk" male adolescent population being identified may only represent about 25% of the targeted population which could have been identified had a better data-soliciting/intake method had been used.  The additional caveat is that, no matter what method is used to solicit "homosexuality" information from adolescent males (and probably young adult males), the ones responding affirmatively to the requested "homosexuality" information will always represent an underestimate of "homosexually oriented" males in the study sample (Bagley & Tremblay, 1996).
Bagley C, Wood M, Young L (1994). Victim to abuser: mental health and behavioral sequels of child sexual abuse in a community survey of young adult males. Child Abuse & Neglect, 18(8), 683-97.   (A PubMed "abstract" link with a document delivery service.)
 

Email:   Pierre Tremblay: ----- pierre@youth-suicide.com ----- (403) 245-8827
Home Page
GLBT Education Index
Top Of Page

Visitor Numbers
Tracker
Total Hits
Web Counter
Cell Phone Reviews


excellence
I got this Award:
 If you want one click on the image