Loading ...
Sorry, an error occurred while loading the content.
 

RE: [evol-psych] data for the 'armchair'

Expand Messages
  • Leif Edward Ottesen Kennair
    Surely, like any mathematical or statistical method, meta-analysis is as good as the data processed - garbage in, garbage out... And surely it is not rational
    Message 1 of 7 , Jul 1, 2002
      Surely, like any mathematical or statistical method, meta-analysis is as good
      as the data processed - garbage in, garbage out... And surely it is not
      rational to be opposed to mata-analysis or even critical of meta-analysis as
      such (like questioning algebra or correlation studies might not make sense) -
      as long as the method is used as intended, and claims not supported by the
      method are avoided.

      Cheers,

      Leif Edward

      >===== Original Message From Ralph L Holloway <rlh2@...> =====
      >I'm the guilty party, not Irwin Silverman, who thinks that meta-analysis
      >ought to be very carefully examined, and that yes, I do think some of the
      >m-a analyses I've seen are a "crock". But I never intended to mean they
      >should be ignored.
      >
      >Ralph L. Holloway
      >Dept. Anthropology
      >Columbia University
      >NY, NY 10027
      >212-854-4570
      >Fax= 212-854-7347
      >Web Page www.columbia.edu/~rlh2
    • Pierre Tremblay
      ... meta-analysis is as good as the data processed - garbage in, garbage out... And surely it is not rational to be opposed to mata-analysis or even critical
      Message 2 of 7 , Jul 2, 2002
        --- Leif Edward Ottesen Kennair <leiedoke@...>
        wrote:
        > Surely, like any mathematical or statistical method,
        meta-analysis is as good as the data processed -
        garbage in, garbage out... And surely it is not
        rational to be opposed to mata-analysis or even
        critical of meta-analysis as such (like questioning
        algebra or correlation studies might not make sense) -
        as long as the method is used as intended, and claims
        not supported by the method are avoided.

        Greetings Leif!

        This "garbage in, garbage out" in a meta analysis came
        to my attention recently in a meta analysis of "sexual
        abuse of boys" studies, and one the these studies that
        I was familiar with had produced one of the highest
        Cohen d values for suicidality.

        However, I knew the study quite well: 750 young adult
        males, about 16% reporting child sexual abuse, and 6
        out of 8 serious suicide attempters reported to have
        been sexually abused. A bit of 2X2 Table math here
        (using approximate numbers) produces 6 attempters in
        120 (6:124) sexually abused boys compared to only 2
        attempters in 630 (2:628) non sexually abused boys.
        [Pearson's= 19.112 (p= 0.0000), Odds Ratio 95% CI:
        3.0313<15.2<76.1537] All of this looks good until one
        more piece of information is added:

        5 out of the 6 male suicide attempters reporting child
        sexual abuse are also reporting being either
        homosexual or bisexual (meaning that 62.5% of all the
        serious male suicide attempters were gay/bi), and this
        information should produce a big "Oops!" with respect
        to anyone's interpretation that "sexual abuse" is
        causally implicated in male youth suicidality.

        The "Oops!" occurs because those knowledgeable about
        the research on gay/bi male youth often report not
        only that the "child sexual abuse" is generally not
        implicated in their suicidality, but that they are at
        high risk, compared to heterosexual males, for
        attempting suicide by a 6 to 14 times factor, and that
        they are also more at risk for having been sexually
        abused by at least a factor of 6. Some gay boys
        attempt / commit suicide, for example, because they
        are acting out the "better be gay than a fag" socially
        learned belief.

        All of this, of course, means that gay/bi males are
        great overcontributors to the counts for both "child
        sexual abuse" and "suicidality" in male youth
        population samples, and that not soliciting "sexual
        orientation" in "child sexual abuse of boys" research
        should be deemed a super 'criminal' act.
        Unfortunately, most of the research that has explored
        the negative results of child sexual abuse on boys
        have not solicited "sexual orientation" information
        from study subjects.

        Given that gay/bi males are such overcontributors to
        both "child sexual abuse" and "suicide attempter"
        counts, it would therefore be important to further
        explore the relationship may mean when a particular
        suicide attempt count is associated with the report of
        sexual abuse, and especially when sexual abuse is
        define as an adolescent boys reporting that he had sex
        with a male 5, 10, or even 40 years older than
        himself.

        Luckily, way back in 1991, Gabe Kruks reported that
        gay/bi adolescent boys on the streets are often
        looking for such a man to love them and that, each
        time they think they have such a relationship and they
        are dumped, they become more and more suicidal. In
        other words, for these gay/bi boys, it is not what so
        many researchers would believe to be causal in the
        suicidality of boys; the opposite applies. That it
        is, at least for some boys, it is not having what the
        researchers believe to be harmful that is implicated
        in reports of serious suicidality.

        In a survey of many studies that sought to explore the
        negative effects of sexual abuse on boys since 1991, I
        have not yet encountered a caveat related to the
        information reported by Kruks. Nor have I seen a
        caveat that warns about the overcontribution of gay/bi
        males to "suicide attempt" and "child sexual abuse"
        counts, and that for this reason, the results /
        interpretation may be more than highly flawed. There
        are also many other serious problems with "child
        sexual abuse" research, but I will not go into these
        flaws.

        The above information was supplied to illustrate how
        social scientists in one field of study have been
        producing some of the 'garbage' that ends up producing
        more 'garbage' in meta analyses.

        Cheers!

        Pierre

        PS. I once had an interesting conversation with a
        Metis woman (more like many conversations) who was a
        counselor at an Aboriginal centre in Calgary that
        specialized in treating addictions and other related
        problems. She noted that many very troubled Aboriginal
        male youth at high risk for many problems (including
        suicide) were not the ones who had been sexually
        abused and had been harmed physically; it was the ones
        who reported 'sexual abuse' experiences that they had
        enjoyed (before they had been educated to believe that
        they should not have enjoyed these same-sex
        experiences), and that the memory of having enjoyed
        such experiences had serious (highly negative)
        homosexual self-labeling implications. That is, it was
        these boys thinking that they were likely "gay" that
        was killing them, as it also applies for many other
        non sexually abused boys who kill themselves or
        survive their suicide attempt(s). These boys, however,
        would not dare tell anyone (maybe with one exception)
        about what had been killing them and coroners are
        generally incapable of reporting more than: the
        suicide victim was drinking alcohol, using drugs, and
        depressed - these being the risk factors often thought
        to be the 'reasons' why males committed suicide. But
        why, really, would such males kill themselves?

        In many cases, I have wondered why social scientists
        have not been asking the questions that would lead to
        some improved understanding of negative outcomes such
        as youth suicide. The answer often happens to be that
        their socialization (often of a highly homophobic
        nature), combined with their professional education,
        often precludes being able to ask the questions that
        should be asked, and in ways that individuals will
        reveal the truth about what is happening in their
        minds.

        A while back on this message board there were postings
        related to Bruce Rind's meta analysis that indicated /
        suggested that the effect of child sexual abuse on
        boys may not be what most professionals had been
        educated to believe, in terms of being harmful. The
        troubles at the APA were major, in part because this
        issue had also reached the ears of those in
        government. Even worse, however, is the response of
        'mental health' professionals to anyone (a researcher
        or a client) reporting that what is called "sexual
        abuse" may also be enjoyed. In other words, it is very
        dangerous for anyone, including clients, to ever
        present a reality, including a personal reality, that
        is at odds with what these 'mental health'
        professionals believe as a result of their 'education'
        often based on incredibly flawed research -- or even
        scandalous that is not recognized for what it is.

        A good example of the high lethal nature of 'mental
        health' professionals is related to the court mandated
        'therapy' that Alan Turing received for what middle-
        to upper-class lawmakers deemed to be "criminal" by
        lawmakers, and middle- to upper-class psychiatrists /
        psychologists believed about homosexual people: that
        they should not exist! The one year of hormone therapy
        began with Turing's conviction in 1952, the other
        option being prison, and by June 1954 he had committed
        suicide, thus acting in accordance with the basic
        beliefs of 'mental health'(?) professionals. On this
        web page,
        http://www.csh.rit.edu/~jon/text/papers/turing/, the
        following is reported: "The treatment had severe,
        negative effects on Turing. He became impotent and his
        breasts grew larger. More damaging, however, were the
        effects on his nervous system. The treatment also
        caused waves of depression and despair."

        Cheers!

        Pierre




        ______________________________________________________________________
        Post your ad for free now! http://personals.yahoo.ca
      Your message has been successfully submitted and would be delivered to recipients shortly.