I. ADVOCACY RESEARCH
The Vermont Psychiatric Association (VPA) and associates prepared a brief in support of
the homosexual couples arguing that they have a right to marry in the State of Vermont. As
is required, the brief included an explanation of their reason for offering evidence in
"This brief is submitted to assist the Court by collecting, summarizing and
discussing the social science research and literature about same-gender couples and the
children they parent... The VPA is committed to providing the courts accurate and up to
date information from credible scholarly research regarding issues surrounding gay
and lesbian individuals, relationships, and parenting." (VPA)
According to the VPA brief:
"social science research demonstrates that committed loving relationships between
two people of the same sex are comparable to committed loving relationships between two
people of different sexes. Marriages between same sex couples would be functionally
equivalent to marriages between different-sex couples." (VPA)
A substantial proportion of the "social science research" provided to the
Court in this brief was material designed and produced to be presented as evidence in
cases similar to that of the plaintiffs by various professionals who wish to forward the
cause of same-sex marriage, adoption, or custody. This kind of "advocacy
research" contrasts sharply with impartial research designed to forward the general
knowledge about a particular problem. While this does not invalidate their findings, it
does suggest that the evidence presented be evaluated with due suspicion.
Charlotte Patterson has written and co-authored a number of articles on the subject on
homosexual parenting, some of which are referenced in the VPA brief. In almost every
article she acknowledges her desire to change social policy. In one article she explicitly
encourages others to use their work to forward social change:
"First psychologists and other social scientists can work to expand the body of
research on lesbian mothers, gay fathers, and their children, and can make efforts to
ensure that the results of research become available through appropriate publications.
Second, psychologists and other qualified social scientists can interpret the research
findings for legal purposes by testifying as expert witnesses in court. Third, the
research findings can be brought to the attention of the courts in the form of amicus
"Uncontroverted scientific evidence may not influence the courts as much as some
social scientists would like, because the social, moral, and religious values of judges
and legislators often hold sway over all other considerations... because lawyers often
have little or no training in the area, they often fail to appreciate the value of social
scientific evidence, believing instead in the superiority of moral and philosophical
It is clearly Patterson's belief that social scientific evidence should trump social,
moral, and religious values. This assumes that the former consists of "facts"
and the later mere opinion, but Patterson's own work shows a lack of respect for facts.
She makes use of totally unjustifiable estimates of the number of "lesbian- and
gay-parented children in the United States." According to Patterson and Redding (1996) "Estimates of the numbers of children
of gay or lesbian parents range from 6 million to 14 million." This distorted figure
appear a number of times in the studies produced to support the VPA brief . Its use
demonstrates the authors' the lack of respect for simple mathematical calculation and the
recycling of invented information through footnoting, which characterizes the much of the
research in this field.
Patterson (1992) made the same claim in
previous article citing additional sources and with the following footnote:
"... such estimates can be based on extrapolations from what is known or believed
about base rates in the population. According to Kinsey, Pomeroy, and Martin (1948) and others (see Blumenfeld & Raymond,
1988), approximately 10% of the 250 million people in the United States today can be
considered gay or lesbian. According to large-scale survey studies (e.g.
Bell & Weinberg, 1978;
Saghir and Robins, 1973), about 10% of gay men and
20% of lesbians are parents. ... many have more than one child. Calculations using these
figures suggest that there are about 3-4 million gay or lesbian parents in the United
States today. If, on average, each parent has two children, that would place the number of
children of formerly married lesbians and gay men at about 6-8 million."
The Kinsey study was not projectable to the entire population and was so acknowledged
at the time. Oft quoted figure of 10% referred to men who had been engaged in mainly
homosexual activity for 3 years since age 15. The percent of the Kinsey male sample
considered to be predominantly homosexual since adulthood was 4%. The figure 10% was never
applied to women. Furthermore, the 10% figure for homosexual men has been discredited by
those tasked with estimating HIV infection rates. After extensive research, they knew the
percentage of men who have sex with men who were HIV positive and projected the number of
AIDS cases based on the 10% figure. When the expected number of cases failed to appear,
they discovered the flaw in the 10% figure and recalculated the percent of homosexual men
in the population downward. New research suggests the actual rate is under 3% for men and
under 1.5% for women, or a total of 2% for the adult population. A number of homosexual
activists have admitted using the 10% figure as a public relations tool. No serious
researcher should use such a discredited statistic. In addition, there is a more
fundamental error, the total number of homosexual parents is calculated by multiplying the
entire population by 10%.
The VPA brief claims that: "Research also demonstrates that children of same-sex
couples are not detrimentally affected by being raised by same-gender parents..." The
research presented by the VPA brief does not demonstrate this because the research
presented is so fatally flawed that it cannot be used to prove such a sweeping
II. INVALID STUDIES
In 1993 Belcastro et al published "A Review of Data Based Studies
Addressing the Affects of Homosexual Parenting on Children's Sexual and Social
Functioning." The writers conducted a search of the published literature and
"identified fourteen data based studies which addressed some aspect of homosexual
parenting and its effects on children." Each study was evaluated "according to
accepted standards of scientific inquiry." According the abstract of their article:
"The most impressive finding was that all of the studies lacked external validity,
and not a single study represented any sub-population of homosexual parents. Three studies
met minimum or higher standards of internal validity, while the remaining eleven presented
moderate to fatal threats to internal validity. The conclusion that there are no
significant differences in children reared by lesbian mothers versus heterosexual mothers
is not supported by the published research data base."
Of the fourteen studies cited by Belcastro et al as lacking external validity,
eight are cited as reverences in the VPA brief. The following is a list of these
references and Belcastro's analysis of their validity:
Golombok, S. et al. (1983)
Children in Lesbian and Single Parent Households: Psychosexual and Psychiatric Appraisal.
"The study lacked external validity. The matched control group was compromised in
several ways... Nearly half the women in the experimental group had received psychiatric
care as compared to 18.5 percent of women in the control group (p < .05). Lesbian
families had nearly twice as many female children as male children, with the converse true
of the control group." (Belcastro1993)
Green, R. et al. (1986) Lesbian mothers and
their children: A comparison with solo parent heterosexual mothers and their children. Archives
of Sexual Behavior. 15: 167.
"The researchers concluded that there were no differences between the children
raised by homosexual versus heterosexual mothers. Yet the authors reported that daughters
of lesbian mothers experienced more "cross-dressing" (p < .05). In addition,
the authors reported "daughters of lesbian chose traditionally masculine jobs
significantly more often than heterosexual daughters (p < .05)."... Regarding the
second component of sexual identity, that is, gender role behavior, the authors dismissed
significant statistical differences daughters of lesbian and heterosexual mothers by
stating, "... while there is less adherence to traditionally sex-typed standards,
they (lesbians' daughters) are still similar to many other same-age girls."
"...the authors conceded that their subjects were too young to demonstrate a
sexual orientation, yet go on to conclude, '... it is likely that a comparable number of
predominantly homosexual children will emerge from these groups as the general
Hoeffer, B. (1981) Children's Acquisition of
sex-role behavior in lesbian-mother families.
Belcastro points out that the original Hoeffer study published in 1979 "lacked
external validity." In addition:
"Regarding the conclusions of the study [Hoeffer 1979] the author's bias is
reflected in the final interpretation of the results. For example, commenting on the
finding that boys of lesbian mothers are more gentle than boys of heterosexual mothers the
author states, "...thus boys of lesbian mothers seemed to be slightly ahead of their
heterosexual counterparts on learning valued traditionally feminine traits." Later on
the author concluded, "... a positive finding was that the majority of boys' (70%)
ideal self sex-role trait profile was androgynous." However the finding that twice as
many boys of heterosexual mothers scored sex-typed masculine than boys of lesbian mothers,
and 40 percent of the boys of lesbian mothers scored sex-typed feminine as opposed to zero
percent of the boys of heterosexual mothers was treated as unremarkable. "(Belcastro 1993)
The follow-up publication of the author's dissertation was referenced in the VPA brief.
Hoeffer (1981) concluded: "More striking
than any difference between the two groups of children on acquisition of sex role
behavior, or between the two groups of mothers on encouragement of sex-role behaviors,
were the similarities."
Belcastro points out that this version "eliminated the following statistically
significant results from the original study":
(1) 50 percent of the girls of lesbian mothers scored as sex-typed masculine as
compared to zero percent of the girls of heterosexual mothers;
(2) girls of lesbian mothers had higher scores on male-valued traits than girls of
heterosexual mothers (the converse was true for female valued traits);
(3) boys of lesbian mothers had higher scores on female-valued traits than boys of
heterosexual mothers; and
(4) 100 percent of the lesbian mothers reported males peers as being moderately or
strongly influential on their sons, while only 60 percent of heterosexual mothers did so.
Kirkpatrick et al.(1981) Lesbian
Mothers and Their Children: A Comparison Study
"The study lacked external validity. The study sample (n = 40) attracted an
inordinate amount of emotionally disturbed children due to the investigators offering
mothers a complete evaluation and feedback of their child's psychological profile in
exchange for their participation in the study. This resulted in the study sample
consisting of 60% moderately or severely disturbed subjects. .. the conclusion that
"... no difference in gender development between children brought up by lesbian
mothers and those raised by unmarried heterosexual mothers," is not supported by the
design of the study." (Belcastro 1993)
Lewis, K. (1980) Children of Lesbians: Their Point
of View .
"The study lacked external validity" and "internal validity"
because of aberrant testing conditions." (Belcastro
Miller, B. (1979) Gay Fathers and their Children
"The study lacked external validity" and "internal validity."
"Of the 48 children of sufficient age to determine sexual orientation in the
sample, 14 were arbitrarily selected for study."(Belcastro 1993)
Puryear, D. (1983) Familial Experiences: A
Comparison Between The Children of Lesbian Mothers and the Children of Heterosexual
"The study lacked external validity... The results, reported by the author,
yielded a statistical difference in self and family view between children of lesbian
mothers and children of heterosexual mothers. However, in the summary of the work, the
author suggested that the impact of divorce (or separation) on the child was a greater
influence on this difference than the mother's sexual orientation. This conclusion
conflicted with the study's design since mothers in both heterosexual and homosexual
groups were divorced or separated, and this variable was not factored out of the
analysis." (Belcastro 1993)
Rees, R. (1979) A Comparison of Children of Lesbian
and Single Heterosexual Mothers of Three Measures of Socialization.
"The study lacked external and internal validity. The families with multiple
children in their custody were allowed to self-select one of their children into the
study." (Belcastro 1993)
Belcastro et al concluded:
"Another mutual limitation of these studies was one identified by Rees (1979), namely, lesbians' political and legal desire to
present a happy, well-adjusted family to the world."
"Finally, based upon the researchers' interpretation of the data and at least in
one case censorship of the data, more were biased toward proving homosexual parents were
fit parents. A disturbing revelation was that some of the published works had to disregard
their own results in order to conclude that homosexuals were fit parents. We believe that
the system of manuscript review by peers, for minimum scientific standards of research was
compromised in several of these studies."(Belcastro
III. UNPROJECTABLE RESEARCH
A meta-analysis of a number of studies should always be viewed with suspicion. Since
different researchers rarely use the same criteria, the meta-analysis may combine apples
with grapes. When the studies combined in a meta-analysis lack external or internal
validity or are flawed because their authors ignore their own data, the entire sample is
polluted. Many of the studies referenced in the VPA used non-projectable samples -- and
admit that they are non-projectable in the original study. Yet the same studies are used
to make projections and sweeping generalizations, which are presented as
For example, in the first study referenced in the VPA brief, Allen, M., Burrell, N. (1996) "Comparing the Impact of Homosexual and
Heterosexual parents on Children: Meta-Analysis of Existing Research," is a
meta-analysis of a number of studies which were reviewed by Belcastro and found to be
lacking in external validity or internal validity and therefore not projectable. (Green, 1986; Hoeffer, 1979,
1981; Puryear, 1983; Rees, 1980). The first
line of the abstract reveals the authors' interest in influencing the courts'
determination of "custody and visitation on the basis of the "best interests of
It should be noted that Allen & Burrell admit the limitation of meta-analysis:
"Meta-analysis fails to offer conclusive proof for any claim." However, in a
footnote in the VPA brief which references the Allen & Burrell meta-analysis the
authors assert: "A meta-analysis can help overcome the limitations in individual
studies, such as limited sample size, by cumulating objective data from all reported
This footnote is the reference for the statement that: "the children of gay or
lesbian, or same-sex parents are as happy, healthy and well adjusted as their counterparts
with heterosexual or different sex-parents."(VPA) Not only does a meta-analysis of
invalid studies not prove this conclusion, in this particular case most of the studies
meta-analyzed by Allen & Burrell do not compare the children of homosexual parents to
children living with both biological and married parents, but with the sub-optimum
situation of children living with single mothers -- in many ways the most disadvantaged
children in our society. For example: Harris and Turner(1985) compared 23 gay and lesbian
parents with 16 heterosexual single parents; Hoeffer, (1981) compared 20 lesbian with 20
heterosexual single mothers; Kirkpatrick (1981)
compared 20 children of lesbians with 20 children of heterosexual single mothers; and
Kweskin (1982) compared 22 homosexual with 22 heterosexual single mothers.
The VPA brief concludes that "such research also confirms that the children of gay
and lesbian couples are at least as happy, well-adjusted, intelligent, and morally
developed as those raised by heterosexual parents" Since the research presented does
not compare these children to children raised by their married, biological parents, the
conclusion is invalid.
IV. PROBLEMS IN LESBIAN FAMILIES
The research on children raised by homosexuals referenced in the VPA does provide a
insights into the problems associated in these families, but this information is ignored
by the VPA. A study by Karen Lewis of 21 children of lesbians she saw in therapy is
included in the VPA brief in defense of their position. Lewis's article included comments
by the children aged 9-26 whose mothers were newly declared lesbians. While Lewis is a
supporter of homosexual parents, the comments made by the children should raise concerns
about how homosexual mothers deal with their children's emotional problems. For example,
according to Lewis:
"...the mother's discomfort in telling children about her homosexuality...
prevented the children from asking questions necessary to quell their own fears and from
talking more about their anger and other feelings."
"The younger children emphasized the need for secrecy and the isolation it imposed
"Initial hostility came primarily from the older boys and was directed toward the
lover or hidden in the guise of concern for siblings... furious at his mother's lover...
for 'trying to boss me around.'"
"Several younger children used their mother's gayness as an excuse for not being
close to other children."
"It is easier for children to show overt hostility to an unrelated person."
"... child's ambivalent feelings denied rather than aired."
"Several girls thought they might turn to women if they did not have a satisfying
relationship with a man. One added, 'that's what my mother did.' She said, in regard to
her dating, if she complained to her mother about boys, 'she would tell me to try girls.'
This response not only sets up the daughter to fulfill her mother's prophecy; it also
denies the girl maternal support in problem-solving."
According to Lewis, the children's reactions to their mothers' announcement of her
homosexuality included "brief experimentation with homosexuality." The children
asked: "Will I or won't I be gay when I grow up?" Some were adamant that they
did not want to be gay. Two girls became pregnant which Lewis viewed as their way of
affirming their heterosexual identity.
In only one family did all the children express overt anger at their mother. These
children felt that their mother neglected them in favor of her lover and that an orphanage
would offer more love and affection. Their mother viewed men extremely negatively and
openly expressed the wish that her daughters would be gay and her sons would be
straight," so they would all have a woman."
Lewis drew the following conclusions:
"Most of these [youngest] children simply denied any ambivalence toward their
mother. The anger, hurt, grieving -- the normal feelings one would expect in reaction to
any crisis, developmental or situational -- were missing."
"Some of the children were in family therapy with a gay therapist, but the
therapist was seen by the children as their mother's co-conspirator. "
"It appeared that the children did not feel ambivalent toward their mothers.
Ambivalence is a normal reaction to significant changes in an individuals life... these
children spoke only of their acceptance of their mother's lesbianism. The other side of
ambivalence was missing -- anger, hurt grieving."
The Lewis study suggests that lesbian families may suffer from a particular problem -
namely the denial that there is a problem. It is healthy for children to be able to
complain, voice fears, get angry, and express hostility and still feel loved and accepted.
The unwillingness of the children in Lewis's study to show any ambivalence in a situation
where such feelings would be expected is not healthy. Contrary to what is claimed for
these families, the lack of expressed problems does not show that these families are
superior to heterosexual families, but that they may have deeper problems.
The evidence provided by the Lewis study must also call into question all the studies
which rely on questionnaires completed by homosexual mothers and their children, since
only in a therapeutic setting would an impartial observer be able to discern whether the
family was denying problems or actually problem free. If a gay therapist is seen by the
children as a co-conspirator with the mother, one can legitimately ask: Are the children
likely to answer honestly when interviewed by researcher whose political objective is to
prove the normality of homosexual families.?
What cannot be determined from the questionnaires is whether the homosexual women deny
the problems to themselves or merely deny problems on the questionnaire. While the second
is understandable given the intention of the survey's (to make the case for homosexual
parenting), the first poses a real danger to the mental health and development of the
The negative attitudes of lesbians to research has been noted by other contexts. Saghir
and Robbins, authors of Male and Female Homosexuality: A Comprehensive Investigation,
"In recruiting male and female homosexuals for the study there was a sharp
contrast between the attitudes of the males and the females toward the interviewer and the
research project. The male homosexuals were usually eager to cooperate without much
hesitancy or suspicion. It was a more frustrating experience dealing with women
homosexuals. For one thing, it was more difficult to find them and when solicited for
participation in research they tended to be cautious and suspicious of the motivations and
aims of the researchers. The female homosexual group that was approached was closely knit.
They appeared very secretive and needed a great deal of reassurance and elaboration before
it was possible for the researcher to gain their confidence and cooperation."
While it can be argued that female homosexuals have reasons to be suspicious, this also
makes research material gained from questionnaires distributed to small, self-selected
groups of homosexual women for the purpose of determining their fitness as parents highly
suspect to say the least. They have no motivation to tell the truth if that truth is less
than 100% positive and there is no way to determine the truthfulness of their responses.
V. EVIDENCE IN FAVOR OF THE TRADITIONAL FAMILY
The VPA delivered volumes of research to the Court. The pile was impressive, but each
study must stand on its own. To claim that their pile of research proves the assertions
made in the brief it is the equivalent of claiming that 1 + 0 + 0 + 0 + 0 + 0 + 0 =
Against this what sort of evidence is there? Literally mountains. Were we to collect
all the studies which show the positive value to the child of living with his or her
biological, married parents and the negative effects of all other arrangements, we would
be forced to rent a truck to deliver them to the court. Even a basic bibliography of
relevant references would tax the court's patience.
For example, one book Henry Biller's Fathers and Families: Paternal Factors in Child
Development includes a bibliography of almost 1,000 separate articles or books on the
positive effects of fathers on children. These studies were not done as part of a
political campaign, but serious scholarship done to increase our knowledge of child
The research which shows that children, in general, do better when they are brought up
by their biological father and mother doesn't claim that every child brought up in a
biological family succeeds and that every child separated from one or both biological
parents fails. Such as assertion would be foolish. Children are extremely resilient. Even
in the face of terrible tragedy, life finds a way. Nor should such findings be viewed as
insulting to those single parents, stepparents, grandparents, and adoptive parents who are
raising children. Indeed, such research may help them to realize that the reason they feel
their task is harder is because it is.
The goal should be preventing preventable tragedies and encouraging what is best for
the child. It is one thing for adults to respond heroically to tragic circumstances and
make sacrifices for an existing child. It is quite another to create children who are
predestined to be disadvantaged.