... sum”;

Ami?»
him!

. 1. iii: . ‘r‘..A.-:....|1
l. :1.ifltu€l3.fs:uilvzz ..

1.51!

X ‘ .
“ ,dfi'
,ggpm
'33! v

.mfi
.
:9;

.

if
v

.HIIM. flit)... .0 35‘
_ . : . 3.21!
5.}. an a . y;

:1

13:». 1
Ti? (:31

5!... L... .
31...... [flittiiéi

“F33: . HM
. t}? .nv . .
51$

. 3...... Sx‘hr .
.EZE ”62:..-

5:5?
5:39:

:1! 11216 2:. .
,zp.ai.... :\

 

THFSiq

TATE VE

iliililii’liliiiil ”Him lifliilliiifliiil

3 1293 01389 8709

 

 

 

 

 

 

This is to certify that the
dissertation entitled

Considering Parenthood: A Predictive

Model of Lesbian Women's Decisions To Parent

presented by

Nancy M. McCrohan

has been accepted towards fulfillment
of the requirements for

Ph 0 D 0 degree in PSYChOlogy

 

Date 2-26-96

 

§54///

/ ,, ~-——-
wzg’ép/

M

 

Major professor

William S. Davidson, III

 

MSU is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

 

LIBRARY
Michigan State
University

 

 

 

 

PLACE ll RETURN BOX to remove We checkout from your record.
TO AVOID FINES return on or before one due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

APR 2 7 2010 l

 

 

 

 

 

 

 

 

 

 

 

 

 

——__:_J
-——u ‘j

usu leAn Affirmative AotioNEquel Opportunity Inetituion
W

 

 

CONSIDERING PARENTHOOD:
A PREDICTIVE MODEL OF LESBIAN WOMEN'S DECISIONS TO PARENT
By

Nancy Marion McCrohan

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Psychology

1996

ABSTRACT

CONSIDERING PARENTHOOD:
A PREDICTIVE MODEL OF LESBIAN WOMEN'S DECISIONS TO PARENT

By

Nancy Marion McCrohan

This research examined attitudes toward becoming a parent for lesbian women who
were not currently parents. Data were gathered from surveys distributed at lesbian events and
through a lesbian mailing list. About one-third of the one-hundred sixty-three women in the
sample desired children, and about one-quarter intended to have children. Almost half of the
women in the sample had, in the past two years, seriously considered one or more methods of
becoming a parent. Nearly three in ten had considered pregnancy and nearly one-quarter had
considered adoption. More than half of the women had discussed ideas about becoming a
parent with others.

The majority of the women reported that it would be helpful to have interventions
such as support groups for lesbian mothers, ways to get community support for lesbian
mothers, and information on how to become a parent. The degree of expected support from
others varied, but sisters, partners, mothers, and friends with children were expected to be
most supportive if the participant chose to become a parent.

A path model of lesbian women's decisions to parent was analyzed with the use of
LISREL sofiware. Fishbein's theory of reasoned action was used as a framework for
exploring the multivariate relationships. Most of the relationships expected by TORA were
evident in the results. Specifically, attitudes predicted intention, and intention predicted
behaviors engaged in which would lead to parenthood. Variables measuring subjective norms,
however, did not fit in the path model. A post-hoc model suggested that length of time with a

partner and affiliation with lesbian women who are mothers or desire motherhood may be

influential in increasing behaviors leading to parenthood, and may be useful to examine in
future works. Future directions include research focusing on referent groups and normative
messages, replication of the path model findings, dissemination of findings, and advocacy-

based interventions with lesbian women in the areas of family and parenting.

"Even without children,
the potential for mothering
is an issue that affects all lesbians.
Women possess a deep connection with fertility,
if not in the physical realm
then on a spiritual and metaphorical level"

(Evans, 1990, p. 49).

"[T]he revelation of similarities
between lesbian [mothers] and heterosexual mothers
can threaten to nourish a trend
that accords more value to mothers
than to women who have no children,
regardless of their sexual orientation”

(Lewin, 1993, p.191-192).

ACKNOWLEDGEMENTS

There are so many people who helped me in this endeavor. It would have been
impossible without them. Thank you to the participants - you know who you are - for
Sharing your lives with me, and having faith and trust in me. A special thanks to the people in
the community that made this research possible: Margy, Sandy, Laura, and the other women
of the Lesbian Connection; Terry Grant, Susan Frazier, and Yvonne LeFave of Goldenrod
Distribution; Cheryl VanDeKerkhove and Cindy Lehmkuhle of the Real World Emporium;
Ann and Shea of the Detroit Women's Coffeehouse; Karen Kresnak of the Lesbian and Gay
Community Network of Western Michigan; MiChigan Pride, Inc.; Jennifer Sowle of
Awakening Sisters; Julie Enszer of Affirmations Lesbian/Gay Community Center; Patty
O'Donnell of Friend's North; Georgia Davidson of the Lansing Area Lesbian Mother's
Support Group; Judy Pigg Behrendt of Lavendar Morning; John Monahan of The Triangle
Foundation of Michigan; Suzzette of Changing Times Productions; Marsha Young of
Author's; Kelly Smith of A Woman's Prerogative Bookstore; and Lynden Kelly of Common
language Bookstore.

Many thanks to my committee members (current and former) for their enthusiasm and
feedback: Bill Davidson, Tracy Dobson, Linda Jackson, Robert Caldwell, and Ellen
Strommen. And a special thank you to committee member Deb Bybee who taught me a lot and
worked with me through every phase of this project. Also many thanks to my very
enthusiastic and hard-working assistants: J .P. Machuta, Michelle Rousseau, Ryan Smith, and

Hans Heikkinen. Thanks to the people at the Institute for Public Policy and Social Research

who taught me so much: Marcus Cheatham, Larry Hembroff, and Karen Clark (proofreader-
extraordinairel). Thanks also go to the many people at IPPSR who showed their interest in my
work and encouraged me time and time again.

Thanks to my family of origin and to Suz' family for their love and encouragement.
Many thanks to my friends-family-Of-choice, who consistently supported me in this work (for
so long!): Joann-Mom Jahner, Rebecca Aughton, David Canales-Portalatin (support partner
extraordinairei), Julie Larson, Teri Sexton, dawn kirby, Holly Angelique, Kelly Hazel, Deb
Weise (who knew when to ask and when not to), Jennifer Stanley, Joanne Basta, all the
Daves, and the women in. the Lansing community. Thanks also to Becky Campbell and David
Lounsbury for the technical help and moral support. Thanks again to Karen McCrohan,
Leighann Holzhouse, Teri Sexton, Judy Sowle, and Jennifer Pigg Behrendt for the much
needed help in distributing and/or collecting surveys.

To my partner Susan Putman - words cannot express the deep appreciation and
gratitude I have for your enduring love and support - thanks to you and the rest of our
family: the famous Bear, the late great MacNaughton, the late Iggy, Murphy Brown, Scooter,
Miss Marple, Professor Hill, Spliters, Momma, Momma-two, Tom, Felix, Hunk, Pebbles,
Halley, Little Boy (alias Little Joe), Motley, Pancake, and Gracie. I would not have wanted to
do this without you!

Finally, this work is dedicated to the Universe and to my guides, with much

appreciation for my life and my path.

vi

TABLE OF CONTENTS

INTRODUCTION ............................................ 1
Overview ............................................. 2
Values in Social Science ............................... 4
Target Population and Literature Coverage ................... 4
Why the Frameworks are Important ........................ 6
Lesbian Women Choosing Motherhood ........................... 8
Lesbian and Mother as Ideologies in a Cultural Context ........... 8
Prevalence of Lesbian Mothers and Lesbians Who Want Children ..... 13

Value of Considering Parenthood and Perceptions of Mother as an
Identity ..................................... 15
Why Lesbian Women Choose to Parent ..................... 20
Methods to Become a Parent ............................ 23
Parenting in a Lesbian Couple '. . .......................... 28
Support from the lesbian Community ....................... 30
Fertility Decision Research .................................. 35
Fertility Decision Models .............................. 35
Theory of Reasoned Action ............................. 38
Norms .......................................... 42
Normative Number of Children ...................... 42
Normative Pressures and Reference Group ............... 44
Conscious Decision-Making Regarding Parenthood .......... 48
Structural Factors: Demographic, Material, and Attitudinal .......... 50
Summary ............................................. 55
Components of the Decision Model ........................ 56
Justification for the Research ............................ 58
Research Questions ....................................... 59
Demographics and Affiliation with the Lesbian Community ......... 59
Desire and Decision to Parent ........................... 59
Scale Development .................................. 59
Model Development .................................. 59
METHODOLOGY ............................................ 61
Setting ............................................... 61
Procedure ............................................. 61
Sampling ........................................ 61
Data Collection ..................................... 63

vii

Research Participants ...................................... 64

Characteristics of the Participants ......................... 65
Characteristics of Women Not Included in the Sample ............. 69
Survey Instrument ........................................ 71
Pilot Testing ...................................... 71
Pretesting of the Survey ............................... 72
Scale Development ....................................... 72
Attitude Toward Parenting .............................. 72
Intentions and Behavior ............................... 76
Normative Component ................................ 77
Demographic and Family Variables ........................ 80
Model Development ...................................... 87
Overview of SEM ................................... 87
Components of the Model .............................. 88
Evaluation of Model Fit ............................... 89
RESULTS ................................................. 93
Introduction ............................................ 93
Descriptive Data from the Survey .............................. 93
lesbian-Related Characteristics ........................... 93
Parenting-Related Characteristics .......................... 97
Subjective Norms: Perceptions Regarding Societal Assumptions and
Treatment .................................. 104
Concerns about Lesbian Parenting ........................ 106
Parenting with a Partner .............................. 107
Subjective Norms: Supportive People ...................... 110
Norms: Affiliation with Lesbian Mothers .................... 113
Helpful Interventions ................................ 114
Structural Equation Model Results ............................. 116
The Conceptual Model ............................... 116
The Statistical Model ................................ 121
Results of the First Model ............................. 122
An Alternative Model ................................ 126
Results of the Alternative Model ......................... 130
Comparison of Models ............................... 131
DISCUSSION .............................................. 135
Descriptive Analyses ..................................... 135
Measurement Conclusions ................................. 139
Model Analyses ........................................ 142
Methodological Limitations ................................. 145
Implications for Future Research .............................. 147
Future Directions ....................................... 149

viii

BIBLIOGRAPHY ............................................ 15 1

Appendix A: Example of an Initial Contact Letter to a Community Organization ..... 159
Appendix B: Cover Letter of the Lesbian Connection Mailing ................ 160
Appendix C: Survey Instrument ................................... 161

Table 1.
Table 2.
Table 3.
Table 4.

Table 5.

Table 6.
Table 7.
Table 8.
Table 9.

Table 10.
Table 11.
Table 12.
Table 13.
Table 14.
Table 15.
Table 16.
Table 17.
Table 18.
Table 19.
Table 20.
Table 21.
Table 22.

Table 23.
Table 24.
Table 25.

LIST OF TABLES

Demographic Characteristics ............................ 68
Intercorrelations Between Attitude Toward Parenting Subscales ....... 73
Reliability Analysis of Items in Attitude Subscales ............... 74
Reliability Analysis of Items for Final Scales for Attitude, Intention,

and Behavior Constructs ............................... 82
Reliability Analysis of Items for Final Scales for Subjective Norm

Construct ........................................ 84
Lesbian Related Characteristics: Lesbian Identity ................ 95
Lesbian Related Characteristics: Attendance at Lesbian Activities ...... 96
lesbian Related Characteristics: Disclosure of Lesbian Orientation ..... 96
Parenting Related Characteristics .......................... 99
Methods of Parenting Seriously Considered .................. 102
Activities Engaged in Related to Becoming a Parent ............. 103
Subjective Norms: Perceptions of Societal Assumptions ........... 105
Concerns About Lesbian Parenting ....................... 107
Parenting with a PMer ............ , .................. 109
Subjective Norms: Expected Support From Others .............. 112
Norms: Affiliation with Lesbian Mothers and Desirers ........... 113
Helpful Interventions in Thinking About Parenthood ............. 115
Observed Data for Structural Equation Model ................. 124
Standardized Results of Model # 1 ........................ 125
Correlation Matrix Implied by the Structural Equation Model # 1 ..... 125
Variables Used in Attempted Modifications to the First Model ...... 128

Correlations Between the Variables Used in Attempted Modifications to the
Model and the Dependent Variables

Observed Data for Structural Equation Model ................. 133
Standardized Results of Model # 2 ........................ 134
Correlation Matrix Implied by the Structural Equation Model # 2 ..... 134

Figure 1.
Figure 2.
Figure 3.
Figure 4.

LIST OF FIGURES

Original Conceptual Model: Lesbian Women's Decisions to Parent . . . 118
Revised Conceptual Model: Lesbian Women's Decisions to Parent . . . 119
Statistical Model: Lesbian Women's Decisions to Parent .......... 120
Statistical Model # 2: Lesbian Women‘s Decisions to Parent ....... 132

xi

INTRODUCTION

The purpose of this literature review was to explore the literature on lesbian women
and fertility research to get a picture of what we know about lesbian women's choices to
become parents or to not become parents, and the extent to which the decision factors
considered in previous research might illuminate lesbian women's experiences in considering
parenthood. What we discovered is a dearth of information regarding lesbian women's
decisions to parent.

Because of the lack of information about lesbian women, one of the main goals of this
research was to provide data to describe the sample, and their desires and intentions regarding
parenthood. A further goal was to create measures of lesbian women's attitudes toward
parenting, subjective norms regarding the appropriateness of parenting, intentions to parent,
and behaviors engaged in that would lead to parenthood. The attitude toward parenting
measure contained items concerning motherhood identity, nurturance, creating a family, and
material conditions. There were several measures of subjective norms, including expected
support from others.

A final goal was to build a path model predicting lesbian women's intentions to parent
or not to parent, based on the theory of reasoned action (Fishbein, 1979). Because this
research was exploratory, the purpose of the Fishbein model was not to confirm the model per
se, but to use the model as a framework for exploring the multivariate relationships and to
describe the components of the decision to parent. A full description of the model follows the
literature review. Briefly, the model was a structural (path) model, and was used to examine
the use of attitude toward parenting and subjective norms as predictors of the intention to
parent, and the use of intention to parent as a predictor of engaging in behaviors which could

lead to parenthood.

2

Overview

There have always been lesbian mothers, although this has long been thought of as a
contradiction in terms in popular culture. Many of today's lesbian mothers became mothers in
the context of a previous heterosexual relationship and later "came out," or disclosed, their
lesbianism. Since the early 1980s there has been a "lesbian baby boom," with many lesbian
women deciding to have children after coming out as lesbians (Achtenberg, 1990; Brown,
1992; Gil de Lamadrid, 1991; Pies, 1987b). There has been a recent and growing interest in
this lesbian baby boom in the lesbian pOpulation, popular culture, and in social research. Little
data are available, however, beyond the fact that a baby boom is occurring (Brown, 1992).
Most of the initial information on this phenomenon was in the form of stories (See Hanscombe
& Forster, 1981).

Lesbian women becoming mothers are much like other women, concerned with daily
tasks, health care, childcare arrangements, household preparedness, and so forth. Lesbian
women are also different from heterosexual women facing parenthood, however, in many
ways. The issues involved in lesbian parenting are diverse, but they all stem from the fact of
the woman's lesbianism: relationships with family of origin; how to share parenting with
another woman; how to have both partners recognized as mothers; and many decisions about
the method of achieving motherhood. lesbian women become mothers through adoptive,
foster, biological, and co-parenting means. Each method has its advantages, disadvantages,
and decision factors to be considered.

Lesbian women considering motherhood spend a great deal of time and energy
thinking, processing, and deciding whether to have children, about the methods of achieving
this, and the impact of their decisions (Frank & Brackley, 1989; McCartney, 1985; Pics,

1989). One obstacle that must be dealt with is the right to parent: how they, as lesbians, will

3

be fit to parent. There are questions about the fitness of lesbian women as parents and there
have been psychological studies which conclude that the children of lesbian women are as
psychologically sound as the children of heterosexuals (See McCandlish, 1987 for a review of
studies; Pics, 1989). Further, there are studies which conclude that children of gay men or
lesbian women are no more likely than the children of heterosexuals to become homosexual,
have a confused sexual identity, or exhibit confused gender role behavior (DiLapi, 1989;
Pollack, 1987; Steckel, 1987).

The existence of lesbian families and lesbian parents is controversial. The issues
involved and the factors which influence these families need to be examined, in order to
effectively respond to the needs of all the parties involved. One increasingly popular means of
becoming a lesbian parent has been the use of alternative fertilization to achieve pregnancy.
Society has begun to notice and question this practice. Some physicians are evaluating the
wisdom of performing AF on single heterosexual women and lesbian women, in part due to
concerns about the ethics of single parenthood or lesbian parenthood (Fletcher, 1985; Perkoff,
1985; Strong & Schinfeld, 1984). The issue of access is also a matter of concern in adoption
and foster parenting for lesbian women.

We are still far from fully understanding the breadth and depth of the problems and
joys of the alternative fertilization family (Corea, 1985). Corea reported that lesbian mothers
in her study believed it would be beneficial to have information about their experiences
gathered, possibly by one of the participants. This group also would be wary of researchers,
and scrutinize any researchers before participating in a study. In summary, there are many
issues involved in the creation of lesbian families that warrant attention, and very little

information.

V l in i l 1

Value choices are implicit in the daily work of community psychology professionals.
Values can shape the choice of problems to study, the framing of questions, and the
methodologies we use (Heller, Price, Reinharz, Riger, Wandersman, & D'Aunno, 1984).
Specifically, reformist values are important and unique aspects of the field of community
psychology (Heller et al., 1984).

Community psychologists urge us not to deny the importance of our values, but to
acknowledge that science and values are not incompatible (Heller et al., 1984; Rappaport,
1977). Furthermore, because it is inevitable that values affect our science, social scientists are
obligated to explicitly state our biases and values, and to try to understand their impact upon
our work (Heller et al., 1984).

In short, it is false to claim objectivity in scientific work (Heller et al., 1984; O'Brien,
1989; Rappaport, 1977). Objectivity in any endeavor involving human beings can never be
completely attained, as if we live in a cultural, historical, sociological, psychological vacuum.
Scientific writings which claim objectivity are not objective, but instead silently accept and
promote current mainstream values. This author's subjective value orientation is that
lesbianism is not pathological, and that there is nothing morally or socially unacceptable in
being a lesbian mother. I believe that no mere label (eg. lesbian mother, grandparent, nuclear
family) can convey the experiences of the person(s) or the social network within which they
exist.

I E l . l I . C

Adult lesbian women who are not currently parents is the population of interest in this

research. Other populations are discussed in the literature review, as appropriate, in order to

highlight the experience of lesbian women considering parenthood. These other populations,

5

however, are not the main focus of this paper, and inclusion of them in a comprehensive
manner is far outside the goals of this manuscript. For example, lesbian motherhood is
explored to the extent that it sheds light on the position of lesbian women considering
parenthood. This review does not provide an extensive analysis of the entire experience of
lesbian motherhood per se, nor of gay fatherhood.

Single heterosexual women (not married nor in a committed relationship with a man),
as single mothers or as users of alternative fertilization (AF) as a means of achieving
motherhood, were examined because they share commonality with lesbian women seeking AF.
They look similar in that crucial common factor of making reproductive choices that create
families which do not include a father/ husband. The literature is limited for each of these
groups, and the populations tend to be confused or not clearly defined. In research, lesbian
women are sometimes hidden in "single women" categories, sometimes they are purposely
included in a group with single women, and sometimes lesbian women are single women,
without a male or female partner. Therefore, single women are discussed at certain points in
the literature review to clarify issues involved in lesbian parenting.

There is very little research on the decision-making process of lesbian women
contemplating motherhood, or the factors which may influence them. Therefore, the literature
review pulled information from various areas in order to illuminate the experiences of lesbian
women considering parenthood. Types of sources include theoretical works; exploratory,
qualitative, and quantitative empirical works; and clinical works. Areas discussed included
cultural institutions and norms; lesbian identity; lesbian mothers; interpersonal relationships
for lesbian women; the role of the lesbian community; and heterosexual fertility decision-

making processes and models.

6

Wh he Fr mewor Im II

This review of lesbian women considering parenthood included information about the
cultural and interpersonal context in which lesbian women contemplate becoming parents in
order to more fully appreciate the individual decision-making experience. The importance of
looking at the larger picture is emphasized in community psychology. Community psychology
is multifaceted in nature, and community psychologists often have an ecological orientation,
focusing on the interaction of people and their environments (Heller et al., 1984). The
framework of Bronfenbrenner (1977) and the multifaceted, multidisciplinary goals of
community psychology fit well together. Bronfenbrenner outlines the ecology of human
development, and he conceives of the changing relation between persons and environment in
systems terms. The ecological environment consists of a nested arrangement of structures,
each embedded within the next (Bronfenbrenner, 1977). Bronfenbrenner outlines socio-
cultural, social-structural, family, and individual aspects of the ecological environment. In a
later extensive review article, Bronfenbrenner (1986) reveals how the ecological model is
particularly useful in understanding the development of healthy families, noting the importance
of looking at levels larger than interpersonal dynamics. There is some overlap between the
levels, but they should be viewed as nested, with the individual levels as the innermost,
followed by familial, the structural, and the cultural as the outermost level (Bronfenbrenner,
1986). What this nesting implies is that the more individual levels of analysis cannot be
considered without an understanding of the larger influences of cultural beliefs, norms, and
prescribed roles upon an individual.

The justification for this current research rests upon Bronfenbrenner's
conceptualization of the ecological environment. As noted, in this conceptualization, each level

of the ecological environment impacts the levels nested within that level, and thus, of course,

7

the individual level is impacted by the familial, structural, and cultural levels. As discussed in
the literature review, there are many author's who believe that the cultural and structural
context are quite influential upon lesbian women, and that this influence is in a direction not
necessarily experienced by others. These author's were quite concerned that this is not
experienced in a positive manner. Thus, it was revealed that this cultural influence has
implications for very personal decisions about the creation and composition of one's family. In
short, we must know more about lesbian women's decisions to parent because of the
anticipated impact of cultural beliefs regarding parenthood and lesbian parenting upon
individual level decisions.

Although the bulk of the survey focused on individual level data, the survey also asked
questions that tapped information from all four levels of Bronfenbrenner's conceptualization of
the ecological environment. For example, questions about societal assumptions represented the
socio-cultural (ideas, values) level; questions concerning the legal treatment of lesbian
mothers, and methods of becoming a parent represented the socio-structural (values
institutionalized in systems) level; questions about the creation of a family and the changes
parenthood could bring to a partnership represented the familial (relationship dynamics) level;
and the attitude questions represented the individual level. This data was useful for descriptive
purposes.

In terms of modeling, however, one cannot measure and model all possible factors, at
so many levels of decision making all in one study. The second framework, the theory of
reasoned action (TORA), helped to paint a picture of the decision factors involved in lesbian
women's decisions to become parents. This was quite useful because it used a limited number
of constructs in a network to provide a comprehensive, yet concise description of the decision

to parent. The TORA model is basically intended to measure more individual level variables,

8

as opposed to the multi-level framework of Bronfenbrenner. The TORA framework, however,
does focus on subjective norms and suggests that these are derived from socio-cultural
messages and the desire to comply with these messages. TORA also suggests that measured
attitudinal variables are based upon beliefs and are influenced by external variables. Thus,
with the TORA, the higher levels are implied, but are not necessarily measured.

Thus, the chapters presented here begin with an overview of the cultural context in
which lesbian women make decisions about becoming or not becoming parents, and move
toward a description of the relationships between the behavior, intention, attitude, and norms

relevant to parenting decisions.

lesbian Women Choosing Motherhood
i 'n l r 1

In this section, we will take a brief look at the ideologies regarding motherhood and
lesbianism, as well as identities of woman, mother, and lesbian. The purpose is not to
delineate an exhaustive review of these concepts, but to highlight the cultural context in which
lesbian mothers and potential lesbian mothers exist. Decisions regarding highly "personal”
choices about partners, families, and childrearing are situated in a specific psychological,
cultural, historical, structural setting, which gives meaning to individual choices.

DiLapi (1989) presents a conceptual framework built on the idea that ”societal values
determine the appropriateness of motherhood" (p. 101), and that a hierarchy of
appropriateness exists. In this hierarchy, the two main criteria of appropriateness for
motherhood concern sexual orientation and family form. The most appropriate mother is the
married heterosexual woman (correct form and sexual orientation); the marginally appropriate

mother is the single, teen, disabled, or foster mother (incorrect form but correct orientation);

9

and the least appropriate mother is the lesbian woman (incorrect form and incorrect
orientation). Other criteria in this hierarchical framework include fertility, method of
conception or parenthood, and the consciousness of the decision to parent.

DiLapi (1989) suggests that this hierarchical model is a good tool to describe a
"system of unequal distribution of power and resources supporting motherhood" (p. 108). An
example of resource distribution is seen in the use of new reproductive technologies; options
to deal with fertility problems are available mostly to heterosexual married couples.
”Mainstream institutions such as social service agencies, medical facilities, and legal systems
control who has access to these alternative forms of parenthood. In this way, the illusion of
the ‘traditional nuclear family' remains intact" (p. 111).

Motherhood is not just a biological possibility for some women; it is also an ideology
affecting our entire culture (Gordon, 1976). This compulsory motherhood (Gordon, 1976)
reflects social beliefs that motherhood is an integral part of being female; that femininity itself
depends on being a mother. It includes the idea that all women want to be mothers, that all
women should be mothers, that all women will become mothers. Motherhood operates as an
institution, defining women's value in society based on the ability to bear children. This
mothering role is considered a critical role in the creation and maintenance of the family unit
(DiLapi, 1989). However, the expectation that women will be mothers has also led to a
narrowing of women's roles outside the home. The mothering role is considered a main reason
for the second-class status of women (Chodorow, 1974).

Homophobia is rage, fear, and hatred of homosexuals, affecting both the person who
feels this, and those with whom the person interacts (DiLapi, 1989). DiLapi demonstrates how
homophobia is institutionalized in the United States by noting that in most of the US.

discrimination of homosexuals is legal; that homosexual behaviors are illegal; and that national

10

civil rights legislation does not address homosexuals. "In the case of lesbian motherhood,
institutional homophobia denies lesbians access to their reproductive rights, including freedom
of sexual expression, information on health care, and parenting options" (DiLapi, 1989, p.
106).

A companion ideology to homophobia is heterosexism. Heterosexism is the idea that
everyone should be, and is assumed to be, heterosexual (DiLapi, 1989). A heterosexist attitude
implies that heterosexuality is not perceived as an institution, but just assumed to be every
person's innate orientation. But some authors urge a closer look at heterosexuality as an
ideology. "[Hleterosexuality, like motherhood, needs to be recognized and studied as a
political instimfion" (Rich, 1980, p. 637). Heterosexism, like homophobia, is believed to be
an additional barrier to lesbian motherhood, and is "reflected in legislative efforts to control
behavior and deny access to services" (DiLapi, 1989, p. 106).

In summary, social beliefs and values are not just beliefs, but are translated into
actions and institutional policies. Authors looking at the ideologies of compulsory
motherhood, homophobia, and heterosexism reveal how these ideas become institutionalized;
how the values become ingrained into the actions, policies, and operations of systems such as
the courts, social service agencies and the medical industry. This is expected to have serious
consequences for reproductive freedom for lesbian women.

Another aspect of the cultural context in which lesbian women consider parenting is
their invisibility as families. lesbian households are not alternatives to the family, but
alternative kinds of families (Bozett, 1987; Pies, 1987a; Rothman, 1989). Lesbian households
fulfill many or all the functions of a family (Ainslie & Feltey, 1991; Bozett, 1987; Schneider
& O'Neill, 1993) and also act as social support for members of an oppressed minority group

(Ainslie & Feltey, 1991). A lesbian family is achieved by self-identification and mutual

ll

agreement, rather than through ascribed roles gained via biological or legal definitions. Ainslie
and Feltey (1991) note that Marciano's concept of "wider families" - families other than
kinship and marriage based - apply to lesbian families.

Because lesbian women create various configurations of families, they are sometimes
seen as breaking down the recent tradition of nuclear families, and there is a dominant social
belief that lesbian women (and gay men) should not be parents (Pies, 1987a). lesbian women
are seen as having no family, or not interested in family (Achtenberg, 1990), or not valuing
family (Achtenberg, 1990; Bozett, 1987). There is no recognition that millions of lesbian
women are parents.

There are those who argue that the lack of recognition of lesbian families must
change, for developmental reasons (Slater & Mencher, 1991), and for social policy reasons
(Schneider & O'Neill, 1993). In the social realm, there is a lack of recognition - lesbian
families have not had the benefit of rituals to mark and validate milestones and passages.
Slater and Mencher (1991) point out that rituals to validate and sustain family life include an
enormous variety of activities and events: engagement and marriage activities - rings, last
name changes, legal and financial changes; births; baptisms; Bar Mitzvahs; Bat Mitzvahs;
anniversaries; funeral services recognizing the next-of-kin. Lesbian family life, "completely
empty of images of normal progression", is allowed only individual level rituals, such as
birthdays or new job celebrations, that don't relate directly to the family as a unit (Slater &
Mencher, 1991, p. 373).

Since the integrity of their bond is not recognized, the bond is seen as without stages,

overriding themes, or temporal progression. There is no act, commitment, or length of

time spent together that moves such a couple into a sanctioned social status. The

couple must fight both active and passive disregard throughout life (Slater & Mencher,

12

1991 , p. 376).

Even when a couple spends a lifetime together, the partners are seen as "unmarried
women who never had a family" (Slater & Mencher, 1991, p. 376). Because of the perceived
lack of validation, or progression as a family unit, the length of time a couple is together often
takes on special meaning, as an indication of their legitimate bond (Slater & Mencher, 1991).
The duration of the relationship takes the place of external conventional validation.

A social policy perspective echoes the developmental perspective: legal recognition of
lesbian partnerships is missing, can impede family unity, and would be beneficial to the couple
and to the larger society. Schneider and O'Neill (1993) suggest that spousal benefits to same-
sex partners is in the best interest of gays and lesbians and society as a whole, that this is a
social policy matter, and that denying same-sex spouse benefits is in opposition to the social
policy of ensuring the well-being of citizens; acknowledging partners in this way "promotes
security, stability, and other values which families ideally represent" (p. 23).

In summary, there are a variety of issues involved in preparing for parenthood, but the
most problematic are the questions focused on the right to even have or raise children.
lesbian women have difficulties in addressing this issue, questioning their right to have
children. These issues reveal the influence of societal norms and cultural biases in the
individual level decision-making process. The internalized homophobia that lesbian women
experience consists of the criticisms and doubts about one's self, life, and peers (Pies, 1989).
Pies (1989) asserts that internalized homophobia cannot easily be avoided, and causes lesbian
women to question their desires for parenthood. Gays and lesbian women are seen by the
media and others primarily by their sexuality, not their skills, strengths, and personal
characteristics. This makes it harder to see them as parents, involving a holistic perception of

the person.

13

Thus, lesbian women deal with many concerns in evaluating motherhood. Pics (1989)
reminds us that lesbian women face the same concerns as do heterosexuals, including time;
money; emotional support; response from friends and family; and having a healthy child.
These concerns are compounded by societal assumptions and biases about lesbian women,
especially about lesbian women who choose to become parents. Robinson and Pizer (1985)
agree that lesbian women face special problems in conceiving mostly because of the
discrimination that they face.

v I c f i n M e i n W 'l II

To begin the analysis of the decision to become a lesbian mother, there are several
basic questions that are commonly asked and should be addressed immediately. The questions
are: (1) how many lesbian mothers are there?;~ (2) how many lesbian women would like to
become mothers?; (3) why should lesbian women even ponder the role of mother and how is
this role perceived?

As to the W, lesbian women have always been parents in our society,
visible or not (Falk, 1989; Pollack, 1990). It is estimated that there are 3 to 4 million lesbian
mothers in the United States, (based on an estimate that 10-20% of women are lesbians, and
that 20-30% of lesbians are mothers) (Pollack, 1987). Some of the children are from previous
heterosexual relationships, whereas others are adopted, fostered, or conceived through
reproductive technologies, such as AF. Some lesbian women raise their children as single
parents, and some raise their children in the context of lesbian relationships.

Because so many lesbian women became mothers in the context of previous
heterosexual relationships, the W becomes how many lesbian women want to
become parents after they are "out" as lesbians? Pies (1987b) reports that more than 300

lesbian women have attended her "Considering Parenthood" groups, which are usually

l4

composed of 8 to 15 women, for a six week period. She reports that about 15% of each group
decides to have children, and the others decide not to have children or are still considering
their options at the end of the group (Pies, 1987b).

Ryan and Bradford (cited in Shaw, 1989), in the first national survey of lesbian health,
report that 28.7% of their sample had been pregnant, and an additional 22.8% desired
pregnancy. In a statewide (Michigan) health survey consisting of 1,681 lesbian women,
1.74% of the sample had had artificial insemination (Bybee, 1990). However, a larger number
were interested in becoming pregnant: 25% of the total sample said they would try AI, and an
additional 9% said they would consider sex with a man to become pregnant.

Zeidenstein (1990) did a descriptive study with 20 lesbian women regarding the effect
of disclosure of lesbian identity upon gynecological and childbearing needs and experiences.
She reported that 70% of the lesbian women in her study wanted to become mothers; 15% had
considered parenthood and decided against it; and 15% had never considered parenting.
Twenty percent of the women in her sample had children after coming out as lesbians, or
were pregnant at the time of the study. It is likely that the rate of 70% desiring children is not
truly representative of the desires of the general lesbian population, but was due to selective
participation of lesbian women interested in the topic of gynecological and childbearing needs.
Furthermore, 55% of the total sample had considered or pursued adoption, and 60% of the
total sample had either considered or pursued AI in order to have children (Zeidenstein,
1990).

Johnson, Smith and Guenther (1987) reported that 58.8% of the 1,921 lesbian women
in their 1980 health study had considered having a child after coming out to themselves. Of
those who had considered having a child, 61% had considered AF, 37% had considered

having sexual intercourse with a cooperative man, and 15% had considered having intercourse

15

with an unsuspecting man in order to conceive. Sixty-two (62%) percent of those who had
considered having a child had considered adoption. Very few of those women who had
considered having a child actually pursued this (10.3%), and about one-quarter of those had
successful results. It appears from the article that "considered having children" was not
defined for the respondents. The data from this question are very interesting, but there is no
information about the outcome of that consideration (ie. considered and decided to not have),
or the strength of current desires. The sample was one of convenience, with participants from
women's music festivals, and the group was mainly white, young, and well educated.
However, it was the largest sample to date, and included a wide range of ages, occupations,
and geographical areas (Johnson et al., 1987).

To recap, it is estimated that there are millions of lesbian mothers in the United States
today, a great many of whom became mothers in the context of a previous heterosexual
relationship. The research cited varies in the estimates of the proportion of lesbian women
who are considering parenthood. As a whole, however, the research suggests that a notable
percentage of lesbian women are considering parenthood and becoming parents. Clearly

lesbian motherhood is not an oxymoron.

 

The W is partly based on the assumption that lesbian women are not
mothers: why should lesbian women even bother to think about parenting?; and how is this
identity/role perceived? Perhaps we should be asking why should they not consider parenthood
- some authors suggest that it is very important to consider and ponder, regardless of the final
decision.

Evans (1990) made the unusual call for all lesbian women to look at the issues of

motherhood and to make conscious choices regarding this role. She insisted that to not

16

examine the meaning and desirability of any role is psychologically harmful. She did not
suppose that all lesbian women would choose motherhood or choose childlessness, but that
each must examine her desires, motives, and the meaning of the role for herself.

Evans (1990) looked at mothering issues for all types of lesbian women - those who
want children, those who have children from previous heterosexual marriages, those who are
not mothers, those who are partners to lesbian mothers. Evans reviewed literature on lesbian
motherhood and found nothing on the needs of lesbian women from a developmental
perspective, nor any literature with a developmental framework that analyzed options
regarding decisions about motherhood. Evans looked at the issues for parenting, noting that
choices in this realm offer the potential for growth and for loss. She emphasized that
parenthood decisions be informed ones that are made quite consciously.

In the literature regarding lesbian parenting, Evans (1990) and Pies (1988) provide
unique contributions to decision-making and identity aspects of parenting, with unusual
attention to childless lesbian women. Being a childless (or childfree) lesbian offers many
opportunities, often unrecognized, for growth. Both Evans and Pics insist that it is important
for lesbian women to make conscious choices about parenting, and emphasize that the best
final decision belongs to each individual. "Even without children, the potential for mothering
is an issue that affects all lesbians. Women possess a deep connection with fertility, if not in
the physical realm then on a spiritual and metaphorical level" (Evans, 1990, p. 49).

Evans (1990) noted the differences between the role of mother and mothering energy.
She reported that the role of mother is one which has been culturally distorted to mean a
person defined by the needs of others, in contrast with the more spiritual construct of
mothering energy:

When mothering is recognized and understood to be both part of and larger than the

17

actual act of being a mother, this energy is powerful and creative. To own this

heightened vision the lesbian woman without a child must individually reckon with

what femininity really means to her, and how she chooses to express her fertility. If
this question gets bypassed or its importance denied, a primary source of her

development gets lost (Evans, 1990, p. 49).

Evans (1990) discussed the therapeutic tasks for helping lesbian women explore the
meaning of motherhood, which include clarifying and supporting the mother role; accepting
and grieving losses (either loss of traditional family or the lack of children); supporting the
spiritual exploration of the feminine; developing alternative creative expression; and exploring
the family of origin. Evans (1990) insisted that motives for parenting decisions need to be
examined and that the family of origin is the most productive area for information and clarity.

Perhaps it is so important for lesbian women to examine the role of mother (Evans,
1990) because the identities of lesbian and mother are viewed as opposites, and are both so
strong. Rich (1979) discusses these two powerful sources of identity - motherhood and
lesbianism - and how the two main components of these sources are experience and
institution. These salient identities, lesbian and mother, are seen as opposites in both
heterosexual culture and in the lesbian culture (Rich, 1979). Others note the difficulties of
merging motherhood and lesbianism in concrete ordinary ways (Pollack, 1987).

The lesbian community is based in part on a sharing of feminist values (Lockard,
1985), and much of the feminist analysis of motherhood is highly critical of the oppressive
nature of mothering in our culture today (Ainslie & Feltey, 1991). Thus, one could expect that
this would lead feminist lesbians to a philosophical reluctance to engage in parenting. It is
commonly believed that lesbian motherhood is a challenge to patriarchal culture and the

nuclear family (Lewin, 1993).

18

Moreover, there are different experiences to be faced by lesbian women as mothers.
"Because a lesbian woman does not fit any of society's stereotypes, she must create a niche
for herself that is uniquely hers. The emotional and political implications of her decisions
regarding mothering are different in many ways from those of heterosexual women" (Evans,
1990, p. 43).

Much of the literature on lesbian motherhood focuses on the similarities and
differences between lesbian and heterosexual parenting. And there is consensus -- lesbian
parents' experiences are the same as heterosexual parents' experiences (Evans, 1990; Pollack,
1990), -— and yet are also different from the heterosexual experiences of parenting (Brown,
1992; Evans, 1990; Pics, 1988; Pollack, 1990; Rohrbaugh, 1988; Stiglitz, 1990). Lewin
(1993) leads us to contemplate how these similarities between lesbian and heterosexual
parenting may serve to reflect and reinforce the gender system in which mothers are more
valued than women with no children.

lewin (1993) is concerned that motherhood and womanhood continue to be mutually
defined, and that lesbians who can choose motherhood can gain access to womanhood through
negotiation. It is important to ask why we have heard about lesbian mothers in the mainstream
recently, and Lewin suggested two viewpoints. In one way, the increase in lesbian mother's
visibility reflects lesbian pride and sense of entitlement or legitimacy. In short, a lack of
shame and secrecy. In the other view, lesbian motherhood is motherhood, reaffirming the
rigid "construction of gender in terms of motherhood and the simultaneous defusing of the
threat to traditional gender categories" which had been achieved by the gay, lesbian, and
feminist movements (lewin, 1993, p. 192).

Lewin (1993) discussed the process of her research evaluating the stresses and coping

of single heterosexual and lesbian women. She hoped to find that the lesbian women in her

19

study looked "normal", but she found more than that - she unexpectedly found them quite
ordinary. She found "creativity directed toward the complexities of negotiating identity rather
than toward delineating bounded behavior or institutions particular to lesbian mothers" (p.
182). Lewin (1993) found that heterosexual and lesbian mothers were essentially the same in
defining motherhood. She found that lesbians' stories "focused on establishing claims to
motherhood and on affirming motherhood as a central identity" (p. 190). Lesbian women,
however, shared the meaning of motherhood without assuming biological inevitability. She
perceived that being both lesbian and mother, they moved between resistance and
accommodation of gender identity.

This research would indicate that lesbian women do not focus on being "lesbian
mothers" per se, but rather on being "mothers". If so, this may mean that for some lesbian
women, "lesbian" and ”mother" are not contradictory experiences; or that the identity of
mother is stronger than that of lesbian; or that in discussing motherhood they are focused on
the mothering experience per se and not perceiving any mothering experiences as being
lesbian-mother-experiences.

In conclusion, research is needed regarding the proportion of lesbian women who
identify with the role of mother or potential mother. Do lesbian women who identify as
potential mothers all strongly desire to be mothers? Or are there some who see themselves as
potential mothers but decide to not have children? Motherhood is more than an identity; it is
also an experience in nurturing others. It would be useful to know more about lesbian
women's attitudes about nurturant behavior toward children. Do mother identity and attitudes
toward nurturing children correlate for lesbian women?

Ideology and identity are important aspects of the context in which one contemplates

becoming a parent. It would also be useful to know more about lesbian women's perceptions

20

of societal messages regarding the appropriateness of lesbian motherhood, and concerns about
being a lesbian mother. As descriptive information, this research project collected information
about the proportion of women who want children, perceptions of societal assumptions about
lesbian motherhood, and the perception of difficulties associated with lesbian motherhood. In
addition, items related to mother identity and nurturance are included in the attitude toward
parenting scale.
Wh le i 0 en P en

Now for a final question - why do lesbian women become parents? The limited
research available on lesbian women who strongly desire parenthood offers some information
about the consciousness of the decision to become a parent and reasons for methods chosen.

Hanscombe and Forster (1981) did groundbreaking work in their interviews with
lesbian women in England and Wales. Based on their qualitative data, they report that many
lesbian women would like to be mothers. They note that "lesbians are just women," with the
same physical, psychological and fertility needs as other women (Hanscombe & Forster, 1981,
p. 15). McCartney (1985) looked at motivations for becoming parents, and related a quote
from one of the subjects which she thought typified the experience of the sample, " ‘ Nothing
is missing in my life, I just want to expand it' " (p. 323). Pics (1989), who has worked with
lesbian women for many years doing research and parenting workshops, found that lesbians
choose to become parents for some of the same reasons as heterosexuals: they want children,
they want to bear a child, they want to create the experience with a partner, or they want to
make a fulfilling childhood for their child.

It appears that alternative fertilization is a common method of conception for many
lesbian women and single women. Harvey, Carr, and Bernheine (1989) did a study of 35

lesbian mothers, using self-report questionnaires. They questioned the women about the

21

reasons for biologically parenting, as opposed to adopting. They found the reasons included:
wanting to experience pregnancy and childbirth (54.3%), the importance of early biological
bonding (22.9%), the lack of adoption alternatives (17.1%), and the desire to raise a newborn
(5.7%).

McCartney (1985) did a study of 12 single women and lesbian women seeking
alternative fertilization, and one aspect of the study was the preference for AF over other
options. She found that some women in the sample had infertility problems and therefore were
reluctant to involve any man not invested in the pregnancy. Further, the women did not want
any man with whom they had a loving but temporary relationship to feel obligated to care for
the child; they wished to prevent any custody lawsuits; they did not want to have intercourse
with strangers or acquaintances; they had difficulty adopting as a single person; they were
cognizant of the greater stresses involved with raising children who were more readily
available for adoption; carrying the pregnancy gave them greater control over the pre-natal
environment; and the semen used in alternative fertilization was screened for health problems.
The lack of eugenic concerns in the responses is noticeable; even though the women saw
advantages in biological parenting, there is no indication that they were seeking "superior"
genetic offspring.

The decision by lesbian women to use AP to conceive a child is a conscious choice
and a lengthy process (Frank & Brackley, 1989; McCartney, 1985; Pies, 1989; Robinson &
Pizer, 1985). Lesbian women seeking motherhood have not only given the method of
conception a great deal of thought, but have contemplated the whole concept of parenthood
thoroughly. Women seeking AF work very hard in planning ahead and seriously considering
the many factors involved in being a parent, and these decisions " . . .reflected the conscious

choice to pursue a strongly prized goal in life with the recognition of accountability for the

22

decision" (Frank & Brackley, 1989, p. 158). The women have given much consideration and
"...focused attention to the personal, social, psychological, ethical, and practical
considerations" involved in becoming an autonomous mother (Pies, 1989, p. 139).

Frank and Brackley (1989) summarized that the women they studied were fulfilling
their dreams or goals, but also acknowledged possible future problems. Corea (1985) noted
that many lesbian women face hostility and fear and they worry about what the child may
experience. The women in McCartney's study (1985) had given thought to the child's
environment, and to childrearing behaviors. Mothers have also planned how to tell the child of
their AF conception (Frank & Brackley, 1989; McCartney, 1985).

In summary, the lesbian women in these few studies desire parenthood, they appear
highly motivated and planful. They provided reasons for the methods chosen. The sample
sizes in the research cited here are generally small, and the generalizability to the larger
lesbian population is unclear. It would be useful to gather more information. Are most lesbian
mothers this strong, this certain, this motivated? Do most lesbian women take such a long
time to decide on a course of action? What is the true percentage of lesbian women who
strongly desire motherhood?

Most of the research cited focused on lesbian women who are mothers or who are
interested in motherhood, but there is less information available about lesbian women who do
not choose motherhood. To what degree have other lesbian women given this role conscious
and prolonged thought? In the current research, data were gathered regarding whether
participants had thought a lot about their decisions and if they saw parenting as something that
took a lot of planning. Data were also gathered on the types of methods considered for
becoming a parent, and what activities had been engaged in that might lead to parenthood. In

addition, the model of the intention to parent itself is an overall indicator of why lesbian

23

women want to parent -- the attitude items are enlightening as to the frame of reference and
frame of mind of the participants.
Me h B m Pa 11

Lesbian women considering parenting must make a wide range of decisions regarding
the best way to have children. To become pregnant, or to adopt?; which partner should
become pregnant?; to have a known or unknown donor?; to inseminate at home or with a
physician?; to use ovulation regulating drugs or not? The list goes on and on. We know very
little about the proportion of women who use each method to have children. A brief overview
of the potential difficulties with AF and adoption illustrate the wide range of decisions that
lesbian women considering parenthood must make. There is a need to know what routes
lesbian women take in order to become parents, what are the perceived difficulties of various
methods, and whether these are obstacles to becoming a parent.

Some of the decisions to be made come with legal implications. The use of alternative
fertilization (artificial insemination) by lesbian women to become pregnant is common for
lesbian women choosing to have children. It is a situation that raises many questions; of great
concern to many is the question of sperm vendors' (donors') rights, and how a legal
recognition of these adversely affect a lesbian woman's ability to create a family unit that is
acceptable to her. One of the difficult tasks faced by the women desiring AF is the choice of a
sperm donor or vendor (Pies, 1987a). "[T']here are specific and historical reasons why lesbians
choose unknown donors" (Pies, 1989, p. 147) which include protection from legal battles; the
fear of losing custody; the desire to be the primary parent; not wanting to parent with a
stranger; and the lack of suitable vendors.

In addition, decisions and choices would be easier if recognition of lesbian families

was assured. The legal suggestions for changing alternative insemination and adoption

24

practices illustrates the usefulness of such recognition. Changes in legal process and
clarification of legal rights, with emphasis on acknowledgement of lesbian created families,
would be useful to lesbian women, lesbian families, their children, and society.

The Supreme Court has acknowledged the state interest in protecting legitimate family
relationships and the family unit (See Mid/1m. 1971; W
Sum, 1972), but it has not defined family unit nor defined the nature or boundaries of
its interest in protecting the family (Reproductive technology, 1985). Polikoff (1990) asserted
that the Court can "preserve the fiction" of a one-mother-one-father home, or it can recognize
the reality of diversity, but it cannot make family life uniform for everyone (p. 473). Polikoff
(1990) claimed that the courts need to serve the best interests of the child, which means the
currently functioning family is the starting point, the context in which the child's needs are
examined. She asserts that without consideration of the context, the courts will maintain
inaccurate ideas about family homogeneity, which will be harmful to children where this form
is different.

The situations that arise from the use of alternative fertilization by lesbian women to
create families without sociological or legal fathers can become complex and frustrating for all
parties if there is any conflict of understanding with a known sperm donor regarding paternal
intentions or role. While it is beyond the scope of this review to go into detail about the legal
arguments of such cases, it is useful to note here that several legal scholars are coming to
similar conclusions about ways to prevent such legal conflicts; conclusions which focus on a
recognition of "alternative" families. Central to a clear understanding and fair treatment for all
parties is clarity in familial expectations for situations involving known vendors. Several
commentators have endorsed the idea of a legislative provision for pre-insemination

agreements to outline the intended rights and obligations of the parties (Bryant, 1990;

25

Donovan, 1983; Kaiser, 1988). Authors frequently suggest that contracts should be written if
paternal rights are to be expected. Without a written agreement, the vendor (donor) would be
understood to have waived his rights. There is no guarantee, however, that contracts would be
enforced by a court. Agreements waiving paternal rights and obligations could be honored if
all have freely consented. Paternal rights can be permanently waived in adoptions, and it may
be possible to do this in insemination situations also.

There is a need for legislation which allows and protects the use of alternative
fertilization at home instead of in a medical setting (Spallone & Steinberg, 1987). Having
alternative fertilization in a medical setting does not guarantee adequate vendor screening
(OTA, 1988); nor offers a clear assumption of a waiver of paternal rights (See mg“
mm. 1989); nor meets a need for insuring the woman's safety (Spallone & Steinberg,
1987). Recommendations also include a legal recognition of various family structures, such as
a two-mother home (McCandlish, 1987; Polikoff, 1987).

In recent years, openly lesbian women (and gay men) have had the options of adoption
and foster parenting (Achtenberg, 1990). Florida and New Hampshire, however, explicitly
prohibit lesbian women from adopting (Gore 8; Frank, 1993). In practice most states
discourage adoption by lesbian women; this is unfortunate because of the clear evidence of
lesbians fitness to parent and because there are inadequate numbers of qualified adoptive and
foster homes for children (Achtenberg, 1990).

In states where adoption by lesbian women is allowed, it is only as a "single parent"
adoption (Achtenberg, 1990). In most communities "joint adoptions" (two women are both
legal parents) are not available to lesbian women (Gil de Lamadrid, 1991). There have been a
few joint adoptions, all in Northern California, which were represented by the National Center

for lesbian Rights (Achtenberg, 1990). The first joint adoption of a child by a lesbian couple

26

was in 1986 in California and joint adoption by lesbian women is still considered "extremely
rare" (Gore & Frank, 1993). As of 1989, there were only five joint‘adoption cases
successfully completed in the United States (Achtenberg, 1990).

In recent years there have been increased efforts on the part of lesbian women to gain
"second-parent adoptions. " These are cases in which a nonbiological co-parent adopts the child
of the lesbian mother, without a termination of the biological mother's rights. As of 1989,
there were only ten second-parent adoption cases successfully completed in the United States
(Achtenberg, 1990). Second-parent adoptions have been granted in Alaska, California,
Minnesota, Oregon, Washington, and Washington, DC. (Gil de Lamadrid, 1991). The courts
generally do not allow second-parent adoptions for lesbians, as the court does not recognize
the lesbian co—parent as related to the child by blood or marriage (Gil de Lamadrid, 1991). Gil
de Lamadrid (1991) reports that the few second-parent adoptions granted are under special
circumstances, and does not expect them to become widely available in the future. Further,
the author only recommends pursuing this option if there is ”overwhelming evidence of
stability for the child" (Gil de Lamadrid, 1991, p. 28). The author defines stability as:
longevity of relationship (5 to 10 years minimum); income or employment stability; adequate
age of the child to verbalize the parental relationship; absence of a father; and general stability
of the home.

Under most adoption statutes, second-parent adoptions are not quite the same as step-
parent adoptions which are granted routinely (Gil de Lamadrid, 1991). The mother's partner
cannot adopt the child without limiting the rights of the biological (or initial adoptive) mother
(Gil de Lamadrid, 1991). Second-parent adoptions are considered non-agency, independent
adoptions. In this situation, a state-licensed social worker is assigned to do a home study.

These studies usually include visits to the home and interviews with all the family members.

27

The social worker completes a report, including recommendations, and judges often follow the
recommendation of the social worker (Gil de Lamadrid, 1991).

In most states, in considering adoptions, the legal standard applied by the courts is the
"best interest of the child." However, there is wide variability in the amount of discretion used
by the judge in applying this standard. Adoptions can be denied for lesbian women on a page
basis, that is, based on the presumption that lesbians are unfit by virtue of their lesbianism
(Gil de Lamadrid, 1991). While adoption statutes may appear in some states to be open to
lesbian adoptions, the wording may preclude second-parent adoptions (Gil de Lamadrid,
1991).

In part three of her series of articles on lesbian parenting, Brown (1993b) talks about '
custody cases for previously-heterosexual lesbianwomen. In these court cases, "in just about
any state in any year there are judgments against lesbian and gay parents based predominantly
on their sexuality, but there are also judgments in their favor on the premise that sexual
orientation alone is not sufficient basis for custody decisions" (Brown, 1993b, p. 20). Brown
reports that according to the National Center for Lesbian Rights, the long-term trend for
lesbian women involved in custody battles is positive, despite the mixed results we see now,
and that the outcomes depend mostly on the individual judges and their fears. "The more
people understand gay and lesbian families, the less afraid they will be of us" (Brown, 1993b,
p. 21). It is likely that the gradual changes seen in the courts dealings with custody cases will
also have an effect on cases for joint and second-parent adoptions.

In summary, alternative fertilization situations with known sperm donors are arenas
for conflict due to the lack of recognition of lesbian families. The lack of recognition of
lesbian families is even more striking in the area of adoption. The solutions offered in all

these cases illuminate this issue; the solutions are focused on understanding and recognizing

28

lesbian couples and lesbian couples with children as real families.

In addition, the discussion of the obstacles and potential legal conflicts faced in
becoming a parent illustrates that there are many difficult choices to be made for a lesbian
woman to become a parent. We have very little quantitative information about the proportion
of women who have tried or are willing to try various avenues for becoming pregnant or
adopting. While a detailed listing of the pro‘s and con's of each possible method was beyond
the scope of the current study, collecting information on what methods are considered and
what behaviors are being engaged in are first steps in understanding the choices these women
are making, and what choices they consider to be acceptable for themselves. In the current
research, lesbian women who are not mothers were asked what methods of becoming a parent
they have seriously considered in the past two years, and what activities they have done that
could lead to parenthood in the last two years.

Understanding the lesbian couple relation is important; a good foundation (including
intimacy, loving, communication, and problem-solving skills) for the couple is critical to the
success of the parenting relationship (Pies, 1987b). Acknowledgement that the couple
relationship affects parenting, and is affected by parenting, are likely to be a considered by
lesbian couples as they approach the topic of parenting.

The dynamics of a lesbian couple, however, can differ from heterosexual couples,
based on two differences: the dynamics of having partners of the same gender, and the
dynamics influenced by living in a homophobic culture (Mencher, 1990; Stiglitz, 1990). To
say that lesbians are affected by having a female partner, and by being lesbians in our society
sounds obvious. But let us examine the meaning of these differences. Recent feminist

psychology has put forth ideas about healthy human development, questioning the male-

29

oriented frameworks which view separation and autonomy as the hallmarks of mature
development. Instead there is a focus on the pr0position that women grow through connection,
and that development requires mutual, authentic relationships (Mencher, 1990). A scrutiny of
lesbian relationships can provide a unique opportunity to examine women's psychological
development.

The strength of a lesbian relationship is partly the result of the intense intimacy that
two women can share. Several authors view this intense intimacy as potentially positive or
negative (Mencher, 1990; Stiglitz, 1990). Fusion, merger, and enmeshment are used
interchangeably to denote a common dynamic for lesbian couples (Mencher, 1990). Fusion is
described as a state of psychic unity, where ego boundaries are crossed, or boundaries are
unclear, and the characteristics of fusion are intense intimacy, a lack of separation, or
overidentification (Mencher, 1990). Fusion is in part a bonding that serves as a united front
against homophobia; it is an attempt to validate a stable bond in contrast to the lack of
recognition experienced in the larger culture. Mencher (1990) suggests that, based on her
interview data and contrary to the notion that fusion is limiting in terms of identity, "intense
intimacy creates the trust and safety which foster self-actualization and risk-taking" (p. 4).
Fusion dynamics and behaviors may not be problematic, and may be normative, growth-
promoting, constructive, and an expression of relational strengths.

A thorough discussion about the dynamics of lesbian relationships, and debate as to
whether these are growth-enhancing or destructive, is not necessary here. It is useful,
however, to note the context in which lesbian couples exist and the unique ways in which
being lesbian can affect the couple relationship, and hence, the manner in which lesbian
women approach the idea of parenthood. The intimate lesbian couple relationship is like all

intimate relationships: it is important. In addition, lesbian couple relationships are also imbued

30

with intense meaning. They may be a protection from homophobia, an affirmation of identity,
and a realm of intense intimacy. Partnerships are for some lesbian women the only family they
have —- if the family of origin is unaccepting of their identity.

Thus, it would be useful to know the concerns for the couple relationship vis-a—vis
parenting: the perceived parenting desires of partners and the expectation of changes in the
relationship or the bond brought on by parenting. Items regarding the desire to share the
parenting experience with a partner and other questions related to one's partner were asked in
the survey and were intended for inclusion in the attitude toward parenting scale.

WWII!

A discussion of lesbian families would not be complete without an examination of the
lesbian community. There is much discussion about the role of the lesbian community as an
extended family, and about the importance of community support for lesbian women and for
lesbian mothers. Barnhart (cited in Lockard, 1985) said the community is a partial alternative
form of family unit for community members. Baran and Pannor (1989) reported that lesbians
in their sample got strength from the lesbian community which appeared to be their large,
extended family. Many authors refer to the lesbian community as a network of friends and
lovers, "an extended-family who help each other emotionally, financially, and in other ways"
(Richardson, 1981, p. 284). Slater and Mencher (1991) also viewed the lesbian community as
positive, equating community with the social networks, cultural events, and the increase in
lesbian women coming out publicly. "This sense of commonality with others is the unique
opportunity offered by a lesbian community." (Slater & Mencher, 1991, p. 381). Pearlman
(1987) discussed the "deeply human urge" to belong or feel a part of a group. Pearlman stated
that the need to belong is especially important for lesbian women, and that the lesbian

community provides that sense of belonging, the sense of affirmation or acceptance missing in

31

the larger culture, and helps to maintain lesbian identity.

Pearlman (1987) marveled at the diversity of lesbians beyond the similarity of gender
and the decision to act on their preferences to relate emotionally and sexually to women. She
pointed out the diversities in race, ethnicity, class, politics, education, work, living styles, role
identity, coming out differences, sexual behavior, and so forth.

What is special to lesbians (and gay men) as an Oppressed group is that there is an

attempt to create bonding and community without a shared historical and cultural

experience (unlike other oppressed groups, blacks or Jews, for example), and in spite
of enormous diversity. That both community and lesbian culture have been achieved,
including a sense of lesbian history, tradition, and humor, is no small accomplishment

(Pearlman, 1987, p. 316).

Lockard (1985) reviewed literature on the lesbian community, scrutinized a particular
community, and formed a more specific definition of the phrase 'lesbian community'. Lockard
related that the lesbian community is not the same as the lesbian population, and that it is
distinguished by four features. She stated that the lesbian community is composed of (1) social
networks of lesbian women, (2) their group identities, (3) shared subcultural values, and (4)
an institutional base where the members can interact. Lockard proposed that the lesbian
population is made up of women who identify as lesbians, and that a lesbian is anyone who
claims to be.

Lockard (1985) discussed her participant observation experiences in a lesbian
community in the Southwest region of the United States during a five year period. She noted
the important resources provided to the community by the small number of elder lesbian
women. She noted the importance of ethnicity as a factor in community participation, and

reported that social networks were segregated according to racial lines. Lockard explains that

32

only a few Mexican-American women were active members in the mainly white lesbian
community, and that rather than acting as a bridge between racial groups, they were
marginalized in the Mexican-American social networks. "Cultural and language differences,
and a continuing history of prejudice and discrimination have created a barrier that is not
overcome by the bond of a common sexual preference." (Lockard, 1985, p. 93).

Citing data from her own study and other studies, Lockard (1985) concluded that the
ethnic communities of minorities offer a family structure and ethnic identity that is important
to the minority women, but this also makes it more difficult to come out as lesbians. They
"risk the loss of that family and ethnic community support for the less certain support of the
white lesbian community [therefore minorities] interact primarily with each other in their
own social networks" which lack the institutional base of the white lesbian community (p. 93).

Thus, not everyone in the lesbian population is part of the lesbian community, nor
comes to see members of the community as family members. Sharing "lesbian" as part of
one's core identity may influence the degree to which the lesbian community is one's referent
group. So, what is the role of the community for lesbian women? What does the community
offer, and what are the implications for supporting lesbian women who choose to parent and
those who choose not to parent?

There are tensions in the lesbian community due to the desire for "sameness", so
important for keeping a common identity. But this need for sameness conflicts with the need
for individual identity (Pearlman, 1987). This need for sameness has been described as
"mirroring", indicating that one finds mirror images of oneself in others (Krieger, 1983).
Differences can be uncomfortable in the lesbian community, and Pearlman wonders if
differentness is seen as a tool to erode cohesiveness.

lesbian mothers are being criticized by some lesbians for putting their energies and

33

time into mothering activities which are so closely associated with heterosexuality. Some
lesbian community members do "criticize their [lesbian mothers] decision to have children as
selling out to the heterosexual myth" (Brown, 1992, p. 26). Part of the concern is that lesbian
mothers will not have the time and energy for political work.

Regardless of whether lesbian mothers look "different" from other lesbian women,
some claim that being a lesbian parent is a very political act (Ainslie & Feltey, 1991; Brown,
1992). Pies says that "being a parent is one of the most political statements you can make"; if
it wasn't so many people in the GOP would not be so upset about it (quoted in Brown, 1992,
p. 25). Ainslie and Feltey reported that the women they interviewed saw motherhood as a
vehicle for creating positive social change by raising politically aware children. The mothers
further felt that coming out and creating a positive lesbian identity made them better mothers,
and helped them to appreciate their children‘s uniqueness too.

Perhaps the criticisms are notable because the community is so political. Perhaps some
lesbian mothers, because they are politically-minded, are aware of being invisible as parents
and are not inclined to silence. Pearlman (1987) notes that members of the lesbian community
have high expectations and idealized notions of what lesbian women and the community have
to offer. It is not common for child-care to be provided at concerts, political meetings, and
festivals in heterosexual culture, but lesbian women are expected to offer this, and to be more
sensitive to issues. The expectations are different.

The lesbian community is not a perfect haven; lesbian women can still experience
rejections, alienations, and conflicts in this context (Pearlman, 1987). Some lesbian mothers
perceive that they have been invisible in their dual roles in the lesbian community, just as they
are in the larger society, but they are beginning to find increased support in lesbian mother

support groups (See Lesbian Mothers' Group, 1987). Some lesbian mothers seek out other

34

lesbian mothers when they feel rejected by or invisible to other lesbian women (Brown, 1992).
Often this is in the form of a support or discussion group. Lesbian mothers gain a great deal
from interacting and sharing with other lesbian mothers. They feel a sense of increased
validation and self-confidence; they see more clearly what are lesbian mother issues versus
mother issues (Lesbian Mothers' Group, 1987).

Ainslie and Feltey (1991), reviewing other studies, report that lesbian women active in
lesbian feminist communities feel more autonomy, strength, self-esteem, self-acceptance, and a
more positive lesbian identity, compared to more isolated lesbian women. Ainslie and Feltey
reported the following themes in their study: the centrality of the lesbian community to the
experience of motherhood; the significance of motherhood as political work in lesbian lives;
the definition of motherhood as experience; and the expansion of the definition of motherhood
by the lesbian community to include nonbiological co-parents and members of the lesbian
community. Ainslie and Feltey summarize that "lesbian feminist communities play a central
part in the restructuring of lesbian families" (p. 82). The community provides support and a
comfortable social context. lesbian women are more likely to have close friends than
heterosexuals, which may indicate the existence of stable networks providing strength and
consistency for the children (DiLapi, 1989).

One area rapidly changing in the lesbian community is the growing acknowledgment
of the nonbiological co-parent as a parenting figure and an integral part of the immediate
family unit of the lesbian family. The lesbian community is important in validating the
co-parenting role in part because of the lack of acknowledgement of the co-parent by the
larger culture (Ainslie & Feltey, 1991). Acknowledgement of the co—parent and validation of
the family as a family is important for the child's well-being too (Ainslie & Feltey, 1991;

Brown, 1992).

35

In summary, the lesbian community is in part a network of friends with a shared
identity, and it can serve as a family for lesbian women. Affiliation with the lesbian
community is correlated with positive integration of lesbian identity. The attitude toward
lesbian motherhood in the community is mixed, and may even be in a state of transition.
Lesbian women who are mothers have felt invisible in the community and in the mainstream.
This is changing, however, and a critical factor in their changed experiences and perceptions
is related to changing networks. Lesbian mothers who interact with lesbian women similar to
themselves feel more comfortable and supported. But there is also the idea that the community
in general is becoming more supportive.

Thus, there is much to yet learn about the perceptions/expectations of lesbian women
regarding the support of the lesbian community and of individuals within it. In the current
research, data were collected on: expected support from the lesbian community, expected
support from friends, the importance of community support in parenting, the importance of
friends' support in parenting, whether they had close friends who are parents or who have
parenting interests, and whether they believe that most lesbians assume that lesbians do not

become mothers.

Fertility Decision Research
E '1' E . . I l l !
There has been a cultural shift toward low fertility in recent decades (Oakley, 1986;
Silka & Kiesler, 1977; West & Morgan, 1987). Overpopulation and related issues are in part
causing attitudes to shift to a preference for lower fertility (Straits, 1985). There has also been
much research directed at understanding fertility trends by examining individual decisions

regarding fertility. Decreased fertility is attributed to the increased direct costs of children,

36

increased opportunities for women outside of the home, and the decreased cultural support for
parenthood (Straits, 1985; West & Morgan, 1987).

Fertility decision-making research has focused on a variety of aspects of (heterosexual)
fertility: voluntary sterilization (Miller, Shain, & Pasta, 1990; Miller & Shain, 1985); intent to
relinquish to adoption (Tennyson, 1988); use of reproductive technologies (Johnston, Shaw, &
Bird, 1987); influence of HIV (Sunderland, 1990); nonmarital childbearing (Linn, 1991;
Oakley, 1985); prediction of fertility (Beach, Campbell, & Townes, 1979; Neal & Groat,
1980; Townes, Beach, Campbell, & Wood, 1980); intentional childlessness (Oakley, 1986;
Silka & Kiesler, 1977); and even the decision to be a lesbian parent (Kenney & Tash, 1992;
Pics, 1989). Much of the research in fertility decisions has been focused on the overall
motivation or desire to have children (Beach, Townes, Campbell, & Keating, 1976; Beckman,
1987; Currie, 1988; Dunn, 1988; Oakley & Schechtrnan, 1980; Ory, 1978; Rabin, 1965;
Straits, 1985; Townes, Campbell, Beach, & Martin, 1976; Voss, Knaub, & Eversoll, 1984;
and Wood, Campbell, Townes, & Beach, 1977).

The factors examined in understanding influences on fertility decisions are varied. The
factors examined have been at multiple levels: cultural, structural, familial, and individual.
More specifically, the factors include: norms and sanctions; cultural support for having
children; ideal family size; direct costs and indirect costs of children; value of children; socio-
demographic variables; family of origin variables; changes in relationships or dynamics;
impact on careers; and competition with extra-familial roles.

There are several models used in fertility research to understand or analyze individual
fertility decisions. Decision models in reproductive decisions evolved from more general
decision models. Decision models have focused on various aspects of decision-making,

including a concern with the conditions under which decisions are made, and the quality of the

37

decision-making process (Janis & Mann, 1977). Most fertility models follow frameworks
using a cognitive, rational, orderly progression of assessments.

Oakley (1985) studied premarital childbearing decision-making, looking at when the
sample made childbearing decisions, the degree of discussion with one's partner about
childbearing, and the amount of change over time (retrospectively) in the childbearing
decisions. Oakley suggested that it is likely that different decision models are valid, but each
model would apply to different subsets of the population. For example, she suggests that there
are one-time single-decisions, sequential decisions, and tentative decisions strategies. Oakley
suggested that single-decisions may apply for those who intend to have no children.

Dunn (1988) similarly suggested that there is not one decision-making model that fits
all persons. In Dunn's (1988) analysis of very young mothers' decisions regarding a second
child, she looked at the decision-making from several perspectives. Dunn (1988) examined
ecological interrelationships and socio-cultural influences, including social, political,
economic, household and individual resources and needs. From her findings, Dunn
conceptualized several fertility "styles" of the young mothers.

In fertility decision studies, a commonly used model is the normative model (see Fried
& Udry, 1980; Ory, 1978). According to this theory fertility decisions depend heavily on
childbearing norms, reinforcing social sanctions, and influences of social-structural factors on
the incorporation of norms. Factors relating to the constructs of the normative model include:
ideal family size, age when decisions to parent are made, passive versus active decisions, and
social sanctions and attitudes.

Another approach to understanding fertility decisions is the structural model, which
looks at outcomes based on socio-demographic factors and psychosocial attitudes (see Currie,

1988; Ory, 1978; West & Morgan, 1987). Factors considered important according to the

38

constructs of the structural model include: family order or position, childcare responsibilities
experienced in childhood, work histories, contraceptive availability, overpopulation, familial
roles and extra-familial roles, religion, and sex roles.

Many studies utilize a subjective expected utility model (SEU) as a framework to
understand the factors involved in fertility decisions. The factors of the SEU model commonly
examined in fertility research include: the affective value and expectation of an occurrence,
the normative beliefs and the motivation to comply with these, and the confounding variable of
social desirability (see Fishbein, 1972; Townes et al, 1980). The SEU is derived from the
social exchange theory model, but adds another factor -- the intention to act (see Fishbein,
1972).

WE

Fishbein (1979) questioned the use of so many different models to explain behavior in
different content domains. He urged psychologists to develop general and comprehensive
theories of behavior. In short, he suggested that a limited number of constructs within a
theoretical network could be used to predict and explain various behaviors. Fishbein and his
colleagues called their efforts to create such a comprehensive theory the "theory of reasoned
action" (TORA) (Ajzen & Fishbein, 1980; Fishbein, 1979). This is based an the assumption
that people are rational, with the capacity to process information. While information is not
assumed to be complete, nor the processing perfect, data is assumed to be used in a reasonable
way to arrive at a behavioral decision.

The utility and comprehensiveness of the TORA is illustrated by its use in
investigating a variety of research topics. The TORA has been successfully used as a
framework to investigate attitudes toward contraceptive use (Chan & Fishbein, 1993; J accard,

Hand, Ku, Richardson, & Abella, 1981; Jaccard, Helbig, Wan, Gutrnan, & Kritz-Silverstein,

39
1990; and McCarty, 1981), attitudes toward a Park Service Policy (Bright, Manfredo,

Fishbein, & Bath, 1993), intentions regarding safe sex practices (Fishbein et al., 1992;
Fishbein et al., 1993), and prediction of intention to parent (Davidson & Jaccard, 1976;
Fishbein, 1972; Jaccard & Davidson, 1975; Loken & Fishbein, 1980).

The theory of reasoned action maintains that behavior is determined by beliefs,
outcome evaluations, and motivation to comply with perceived norms. The main questions put
forth by the theory of reasoned action concern (1) what are the determinants of intentions, and
(2) to what degree do intentions predict behavior. According to the TORA framework,
intentions are based on the (1) attitude toward the behavior and (2) subjective norms (Ajzen &
Fishbein, 1980; Fishbein, 1972; Fishbein, 1979; Fishbein & Ajzen, 1975). More specifically,
the attitude component is the attitude toward performing the behavior. Attitude is assumed to
be made up of (a) beliefs about the outcomes of a behavior and (b) an evaluation of these
outcomes. Subjective norms are (a) beliefs of what others expect one to do, or say one should
do, and (b) the desire to comply with what others think. In other words, the theory of
reasoned action says you do something when you think it will turn out well and think that
others want you to do it. These are the driving forces behind intentions. As to the second
aspect of the theory of reasoned action, evidence suggests that when appropriate means of
intention are obtained, they are highly predictive of overt behavior (Davidson & Jaccard,
1976; Fishbein, 1979; Loken & Fishbein, 1980; Townes et al, 1989).

This theory of reasoned action suggests a rather parsimonious model of behavior.
Fishbein (1979) says that once one has a good measure of attitudes, prediction of intention is
not much improved by measuring other attitudes, personality variables, or demographic
variables. These external variables do not describe why a person engages or does not engage

in any given behavior, in and of themselves. They cannot be used to M predict behavior.

40

These external variables, however, could be explored as factors that contribute to differences
in people's beliefs, outcome evaluations, and motivations to comply, ie., they may determine
intentions. Thus, Fishbein proposed that the theory of reasoned action provides a framework
in which to appropriately study external variables.

In the current research, the theory of reasoned action (TORA) was used as a
framework in which to understand and predict lesbian women's fertility decisions. Lesbian
women are a unique subgroup in many ways. Given the assumptions of non-motherhood, and
other lesbian-specific issues, the use of the TORA was expected to be most helpful in
understanding lesbian women's decisions in a general and comprehensive manner. It was not
anticipated that all lesbian women would make the same choices, but that differences in
choices could be predicted by the direction of attitudes, norms, and intentions.

One of the tasks in the current study was to ascertain how to measure the attitudes of
lesbian women. For this task we explored the meaning of parenthood - the pros, the cons, as
well as the issues that may be more lesbian-specific. Some of the factors found to be
illuminating in previous research were used to create the attitude scale. Thus, the attitude
component of this research was based on factors considered relevant to fertility decisions.

The items used in this research to represent attitude were assumed to be caused by
beliefs about the consequences of actions and the evaluation of these outcomes. No effort was
made to measure the beliefs about outcome and evaluations more specifically and directly, as
was done by others (See Loken & Fishbein, 1980). Such an endeavor was outside the goals of
the current research.

Recall that in addition to the attitude component, subjective norms are a component of
the intention construct, according to TORA. Efforts were not made to ascertain the perception

of a general norm of lesbian parenting by directly asking if specific, important referents (cg. ,

41

person x or person y), believed the respondent should or should not become a parent. Instead,
respondents were asked if they thought that specific people in their lives would be supportive
if the participants chose to become parents. We did not directly ask if they were motivated to
comply with other peoples opinions, but asked them how important it was to have the support
of others. In addition, respondents were asked if they thought that most peOple assumed that
lesbians did not become mothers, if they thought that most lesbians assumed lesbians do not
become mothers. Thus, expected support from others was one of the subjective norm variables
hypothesized to predict intentions to parent.

The use of the TORA has been successful in examining and predicting intentions to
parent in several studies (Davidson & Jaccard, 1976; Fishbein, 1972; Jaccard & Davidson,
1975; Loken & Fishbein, 1980). In each of these studies, the two components - attitudinal
beliefs and normative - adequately predicted childbearing intentions. In addition, in both
Loken and Fishbein's study and in Jaccard and Davidson's study, the attitudinal components
were more important determinants of intentions to have a child than were normative
considerations. Furthermore, the TORA model proved to be comprehensive - external
variables (eg., SES in Jaccard & Davidson, or work variables in Loken & Fishbein) had an
effect on intentions only indirectly - to the extent that the external variables affected the
attitude toward having a child and subjective norms. Knowledge of the external variables did
not improve the prediction of intention beyond that accounted for by attitudes and norms.

In the next section, decision factors in empirical research in considering parenthood
are discussed in greater detail. Specific information about the methodologies and samples of
the studies that are cited will be given as each work is introduced. Here is some general
information about the studies overall. Research in the area of fertility or childbearing decision-

making is usually based on theoretical models of reproductive decision-making. The research

42

frequently is focused on norms, sanctions, structural (ie., practical, material) conditions,
attitudes, behaviors (eg. birth control usage), and intentions.

Most of the research described here is quantitative, and is collected via mailed
surveys, questionnaires, or structured or semi-structured interviews. About half of the target
groups are heterosexual, married couples, and half are heterosexual individuals. About two-
thirds of the studies collecting data from individuals collected data from women.

Some of the research projects' samples are randomly drawn, some are self-selected,
and those that deal with intentional childlessness generally recruit respondents. About one-
quarter of the samples are drawn from university or community college populations. The
samples are generally drawn from urban areas, and at least half are demographically
representative of their area. Most of the samples are predominantly caucasian, middle to upper
income levels. Sample sizes range from 14 to 76 for the 3 ethnographic and exploratory
studies cited here. Of the other studies, sample sizes range from 102 to over 1,000
individuals. Two of these are longitudinal studies.

limes
W

In Ory's (1978) work, she operated from the framework that the documented
preference for 2-4 child families reflects several interacting norms and sanctions that act to
regulate the number of children. The goal of her research was to evaluate the usefulness of
both normative and structural explanations of fertility decisions. Ory worked from the unusual
perspective of understanding voluntary childlessness. The sample consisted of 84 purposive
parents (P) and purposive nonparent (NP) couples. Data were collected using semi-structured
questionnaires, plus in-depth interviews with 10 couples. In evaluating the normative aspects

of fertility decisions, Ory found that over 94% of respondents perceived a general family size

43
preference in the United States of 2-3 children.

Ory (1978) also measured the respondents' desired number of children at 9 life stages
of the respondents' lives, retrospectively. She found that all respondents had internalized a
desire for children at an early age. Respondents who became parents had desired a larger
number of children at each life stage in comparison to the respondents who became NP. The
Ps' desires were all within the 2-4 child range at each life stage. Ory found that the NP men
and the P men did not differ in their desires at the life stages; the female P and female NP
differed in their desires starting in childhood. In Ory's study, about 40% of men and 25% of
women reported that they did not have a desired number of children or did not remember
having a desired number of children when they were children themselves. Ory notes this as
evidence of the saliency of parenting for women as compared to men.

In factoring the respondents' current ages, the older NP and P looked the same in
their desires at earlier life stages. The younger NP and P look different at the life stage
"before marriage". Ory interpreted this as an indication that the older NP were deciding to be
childless via postponement, whereas the younger NP were more likely to get married with a
mutual decision to have no children.

West and Morgan (1987) conducted a public opinion survey, via phone interviewing,
using a random sample. The sample consisted of 438 adults and was demographically
representative of the East coast metropolitan area in which the calls were made. The purpose
of the survey was to examine the existence of a normative threshold for family size. West and
Morgan had several research questions that they were examining: what is the ideal number of
children per family; what is the amount of consensus on an ideal number of children among
varying social groups; reasons for the decrease in ideal number of children (compared to two

decades before); and subgroup differences in the importance of reasons for the decrease.

44
West and Morgan (1987) found that the ideal number of children was 2.24. It is

unclear from the publication of the findings whether the ideal is supposed to represent the
respondents' personal ideals, or the respondents' perceptions of a generalized normative ideal.
There were no significant differences in the ideal number of children data by demographic
groupings based on age, sex, education, or one- versus two- income families. There were
some statistically significant differences in perception of the ideal number of children between
groups in current family size, marital status, religion, and race. A minimum of two-thirds of
respondents in all demographic groups (current family size, marital status, religion, and race),
however, still reported 2 to 3 as an ideal range of number of children for a family.

'v P u f r

There were several studies that examined. the relationship between normative pressures
and fertility planning. Fried and Udry (1980) found evidence that expected and experienced
normative pressures are related to parity. The authors explored a very specific meaning of
norm: a group is said to have a norm about a particular behavior if the members of the group
approve or disapprove of the behavior and reward or punish persons based on their behavior.
Norms per se do not control behavior; there are other conditions which must exist in order for
norms to have an effect. First, people belonging to a group must perceive the existence and
strength of the norm. Secondly, the group members must find the rewards and punishments
associated with the norms salient.

Fried and Udry (1980) asked respondents if they had experienced encouragement or
discouragement about having children. Over half of respondents at various parity levels, sexes,
and races, reported receiving encouragement (which range from 52% of Black men to 72% of
Black women). However, there was also a finding that a surprising number of parity zero

respondents received discouragement from having a child.

45

Fried and Udry (1980) asked respondents if they expected persons in their social circle
to be more approving or more critical if they should become pregnant soon. Most respondents
did not expect a response. Of those expecting a response, people at lower parities anticipated
positive reactions from people in their lives. Expectations of disapproval was almost absent in
the lower parity, but increased with parity. At the higher parities (2 +) fewer respondents
expected any response from others, either positive or negative. The authors concluded that the
positive normative expectations (and the absence of negative expectations) at lower parities
would lead to a positive influence on fertility.

Fried and Udry (1980) asked respondents if they thought that others' opinions would
influence their decisions to have a (another) child. The majority of respondents in all groups
(results were analyzed grouping men and women, Black and white), said they would not be
influenced at all. At higher parity, people were more likely to say they would not be
influenced at all. Fried and Udry suggested that persons at parity two or more would feel no
pressure or influence. An alternate interpretation of this finding is that persons at parity two or
more have already attained the parity norm, so that further pressure is not likely to occur.

Silka and Kiesler (1977) had the unusual goal of gathering information on intentionally
childless couples. They had a self-selected sample of heterosexual married couples responding
to advertisements. They collected data via questionnaires and interviews from couples
intending to have no children; couples delaying childbearing; and couples not yet mutually
decided on having children. The respondents in the sample were representative of the
population from which they were drawn. In Silka and Kiesler's (1977) study, 75% of those
who intended to remain childless were advised/encouraged by others to have children; other
groups (those unsure of intentions and those intending to have children later) experienced this

encouragement less.

46

Ory (1978) measured social sanctions by the: degree and type of stereotyping of
couples with nonnorrnative family sizes; expected societal reactions to chosen family Size; and
perceptions of social pressures to conform to normative family size. Of the nonparents (NP),
83% expected unfavorable reactions to their choices from others, and more than two-thirds Of
NP felt pressure to conform to having children. NP and parents (P) both perceived a negative
cultural attitude toward childlessness. Both NP and P assumed that NP were seen as selfish,
but they differed on the other reasons for the negative attitudes toward NP. The nonparents
assumed that NP were perceived as troubled or immature, whereas the parents assumed NP
were perceived as missing out or having incompatible activities.

Ory (197 8) concluded from nonparents' comments that reference group support
appeared to be an influential factor in the decision to have no children. "Many nonparent
couples indicate that they countered pronatalistic pressures by minimizing social contacts with
parents and friends, seeking new friends among the childfree, or joining organizations... that
provided support and acceptance for the decision to remain childless. " (Ory, 1978, p. 538).
Ory concluded that the nonparents in her study redefined the social meaning of parenthood,
and that they actively sought a referent group that valued nonparenthood.

Ory (197 8) stated that research is needed to identify and refine both normative and
structural components of fertility decisions. Further, she remarked on the importance of
"social psychological factors mediating the impact of the normative" influences: (1) the
respondents definition of the situation, and (2) selective referent group participation (p. 538).
But NP choices of association and redefinition of the social meanings of parenting indicates
that NP "attitudes and behaviors are not anormative but may be motivated by adherence to
variant subcultural norms." (Ory, 1978, p. 538)

In summary, it is firmly established that 2 to 3 is the normative number of children in

47

the United States at the current time for heterosexual married couples. High parity couples
report that they do not feel influenced by others' opinions; they receive no normative
consensus about having additional children. Low parity couples report feeling pressured to
have children, but did not feel they were influenced by such pressures. Non-parents say they
feel pressured from others to have children, and report seeking out a different message from a
different referent group - associating with people who will affirm their decisions to remain
childless.

It is expected that lesbian women will be subjected to conflicting messages about
normative fertility behavior. lesbians, as women, are exposed to the normative expectation of
2-3 children all their lives. As lesLian women, however, they have also experienced the
normative assumption that lesbians don't exist, and if they do, they are non-mothers. "Since
norms pertain not only to specific events but also to specific social positions, subgroups in the
population may differ in their perceptions." (West & Morgan, 1987, p. 162).

There are no data at present that tell us how many children are desired by lesbian
women who intend to have children or who have children. This kind of data, combined with
data about expectations of support for parenting, would be a start to learning if there is a
normative number of children for lesbian women, or if the traditional norm of zero is
beginning to change for lesbian women. We do not have quantitative data regarding lesbian
women's expectations for support of parenting choices. This could provide information about
the normative expectations of various people in the respondents' lives, and the degree to which
lesbian women are receiving mixed messages from different reference groups. In the current
research, data were collected regarding whether participants have lesbian friends with an
interest in parenting, what support was expected from various types of people and from the

lesbian community, and the number of children desired.

48
iu D lion-M kin r in rn

Much of the early research in the area of fertility did not examine fertility so much as
an area of choice, but as a range of decisions within an assumption of parenting; looking at
the timing, the number of children, and so forth. It has only been in more recent decades that
research has focused on whether fertility "decisions" are actually active, conscious, chosen
decisions. People who are intentionally childless are a great resource for examining these
distinctions.

Recall that in Ory's (1978) study, she measured the decision to have children at nine
life stages. She found that Non-parents decide later in life, and are more likely to make post-
marital decisions, as compared to Parents. Looking at all the life stages combined, 92.5% of
NP had made active decisions regarding parenting, whereas 62.9% of P had made active
decisions regarding parenting. Silka and Kiesler (1977) found that the couples who intended to
have no children had thought about whether to have children for the longest period of time (5
years), and had discussed their decisions earliest (about 3 years), but had not discussed their
decision more frequently than the other groups. Furthermore, intentionally childless couples
are committed to their decision, feel happy about it, do not feel they need children for
fulfillment, and do not expect to change their minds (Silka & Kiesler, 1977). In Ory's (1978)
study, the NP group was more likely to define childlessness in positive terms.

Couples who intended to have children did not consciously decide to have children by
way of considering and weighing options; they did not discuss or consider their intentions very
much. Some said they had "always" intended to have children. These data imply that research
on family decision-making " . . . may have over-emphasized the search for deliberation and
choice to the neglect of other factors" influencing actions (Silka & Kiesler, 1977, p. 24).

Oakley (1985) pulled 202 ever-married subjects as a subsample from a larger study

49

(larger study reported in Oakley, 1986) and measured the extent of premarital decision-
making. Oakley used a discriminant analysis to identify characteristics associated with making
decisions about children, and grouped respondents into the following groups:

Deciders (35% of sample, decided to have no children or decided on the number of

children before marriage); Semi-deciders (28% of sample, decided to have children,

but not the number of children); Talkers (23% of sample, had talked about children,
but had not decided before marriage); Not talkers (14% of sample, were not decided
and had not talked before marriage).

Oakley (1985) found that deciding on the number of children before marriage was
associated with characteristics of the rational-comprehensive thinker, but also with non-
normative childbearing expectations (whether below or above the 2-child norm). A majority
(76%) of the Not Talkers expected to have 2 children. Oakley stated that fertility decisions are
salient and therefore should lead to forethought, and yet for some people, childbearing is not
even a factor in mate selection. Talkers (talked about, but not decided about children)
described why they had not decided about the number of children before marriage: 7 %
disagreed with their partners; 58% said it was not important; and 35% cited other reasons
such as feelings of ambivalence or lack of salience.

Oakley (1985) concluded that the Not Talkers accepted the fertility norm and acted
upon it. Further, theories of the decision-making process (eg. Brim et al., 1962, and Janis &
Mann, 1977) "are not very helpful for analyzing issues that never get onto the agenda. The
findings of this study suggest that, for fertility, normative preferences - rather than high
fertility values per se - may play a part for those who do not even talk about childbearing
before marriage" (p. 563). Oakley says that theories and research on decision-making need to

be expanded to include people who are not choosing between alternatives. In short, normative

50

strictures may indeed influence fertility behaviors, but may not necessarily be very helpful in
explaining conscious decision-making.

In another article, Oakley (1986) noted that previous research in related topics
(abortion, vasectomy) indicated that conscious consideration of alternatives may not be
experienced by everyone. Part of Oakley's hypothesis was that people who specifically
considered low-fertility Options would be different in demographic, attitudinal, and other
processing characteristics. Research on other decision-making areas (ie. not fertility decisions)
suggested that people who recognize options are "more educated and less traditional... these
traits are associated with different values and decision processing characteristics" (Oakley,
1986, p. 251).

As noted earlier, evidence suggested that lesbian women tend to make very conscious
decisions about becoming parents, and take a long time to think through the issue. lesbian
women who have children after they "come out" as lesbians are purposive, intentional parents;
it is unlikely in the extreme that they would become accidental parents. However, it cannot be
assumed that lesbian women have not unconsciously accepted the norm of 2 - 3 children. That
they must purposely take steps to become parents is not proof that they have made conscious
decisions to have children; they could be assumers who are taking steps to act on their
assumptions. In the current research, participants were asked if they have thought a lot about
whether to become a parent or not, and if they agree that there is a lot to think about and plan
for before having children, and whether they have talked with others about parenting.

r : i ' ' ' I

In some fertility research endeavors, material, demographic, and attitudinal factors

have been examined to illuminate the differences between parents and intentional nonparents,

and their reasons for their choices. These structural factors (ie. material, demographic,

51

psychosocial attitudes) have been found to be useful in understanding the decision to become a
parent or not. We will briefly look at some findings of research projects. Notice how there is
no clear pattern that the variables in these studies, or any subgroup of these variables, can
predict behavior. But they could be useful in determining the attitudes toward becoming
parents, from which intentions to parent could be predicted.

The respondents in Silka and Kiesler's (1977) study provided information on reasons
to have and to not have children. Those who intended to have no children and the unsure
group were more likely to check every reason on their given lists for not having children than
those who intended to have children. The main reasons given for not having children by the
group intending to be childless were:

time with spouse (100%); Opportunities and freedom (90%); wife's job or career

(81%); avoid responsibility (71%); leisure or travel (62 %); worry/social problems

(52%); economic (48%); manage work load (43%); dislike children or not sure good

parent (29%). Less than 50% of the intend children group gave any of these reasons.

The reasons given for having children were the same for all groups. Reviewing the
data on reasons to have and not have children, Silka and Kiesler (1977) concluded that there is
"no evidence that the intentionally childless thought in qualitatively different ways about
having and not having children than others" (p. 20).

To recap, people who intend to have no children do provide reasons for wanting
children, which are similar to those given by people who desire children. People who desire
children provided reasons for not wanting children which are the same, but fewer in number,
than those provided by intentionally childless persons. This suggests that the reasons for not
having children are more salient, or more definitive reasons, for people who are intentionally

childless as compared to reasons for having children. The number one reason given relates to

52

the partnership, however, most of other reasons are structural.

West and Morgan (1987) looked at the reasons for the decrease in the ideal family size
over recent decades. The respondents indicated the following reasons for the decrease in
family size as "very or somewhat important":

economic costs (91.5%); wives busy working outside the home (85.4%); better birth

control (89.2%); too worried about the future (67.4%); and less interest in children

(40.9%).

Thus, the majority of respondents supported structural conditions (though not
exclusively, notice the "less interest in children") as reasons for changes in ideal family size.
These findings are "consistent with macroeconomic theories of fertility and leaves open the
possibility that changes in structural components [might result in changes in individual level
decisions" (p.170). Van Dusen and Sheldon (cited in Silka & Kiesler, 1977) also attributed the
decrease in childbearing in recent decades to structural reasons such as: later marriages; fewer
marriages; more divorce; increased labor force participation by women; and more educated
women.

Few studies give comparisons between people who are intentionally childless, those
who are postponing childbearing, and those who are undecided (Silka & Kiesler, 1977).
Research on intentional childlessness completed in the 19603 and before is limited in its
usefulness because "childlessness was rare and deviant" (Silka & Kiesler, 1977, p. 16).
Comparing the nonparents (NP) and parents (P) in her sample, Ory (1978) found that NP and
P differ on their family position. First horns are overrepresented in both the NP and P, there
were more only children in the NP group (25.9% vs. 62.9%), and fewer last borns in the NP
group (5.6% compared to 21.1%). The NP and P groups also differed in experiencing

childcare responsibilities when they were children; 62.9% of P and 24.1% of NP had no

53

experience at all in childcare duties (Ory, 1978).

In terms of other structural items, Ory (1978) found that NP were more likely to be
nonreligious. Other studies have not found any relationship between religiosity and parenting
intentions or parenting status (Beckman, cited in Silka & Kiesler, 1977; Silka & Kiesler,
1977). Silka and Kiesler (1977) found no differences in the various intentions groups in:
selfishness; maturity; interest in humankind and community; community activities; investment
in political and social causes; hobbies; nor materialism. Silka and Kiesler (1977) found no
differences in the various intentions groups' backgrounds in terms of coming from "broken
homes".

Nonparent women were more likely to work outside of the home than women who
were parents in Ory's (1978) study. Susan Bram, (cited in Silka & Kiesler, 1977) did not find
that employment differentiated between couples intentionally childless, couples intending
children, and couples with children. In Silka & Kiesler (1977) there were no differences in
education, amount of employment, intentions for further education, education or career
aspirations, intentions to work, desire for a fulfilling job, but the intentionally childless group
had higher job prestige ratings. The intentionally childless women were more likely to be
professionals and to hold traditionally-male jobs. Silka and Kiesler concluded that the
intentionally childless group had higher levels of rewards and success in their employment.

Silka and Kiesler (1977) found no differences in the intentions groups responses to
their Interpersonal Check List in: happiness; satisfaction with life; satisfaction with partner;
happiness of childhood; self-esteem; agreement with partner on values; and attitudes and
interests. There were no sex role differences found between nonparents and parents in Ory's
(1978) study.

Silka and Kiesler (1977) found only one main area in which groups differed. The

54

intentionally childless and the unsure groups both differed from the intending children group in
having: less interest in interaction with people; greater independence; preference for being
alone; and in behavior that reflected these attitudes to some degree - they lived farther from
their parents and engaged in more solitary pursuits.

There are studies that found few relationships between certain attitudinal variables and
parenting intentions / parenting status. Studies have found no relation between desired number
Of children (or current parental status) and nontraditional attitudes (Beckman, cited in Silka &
Kiesler, 1977; Eagly & Anderson, cited in Silka & Kiesler, 1977); approval of women's
liberation (Beckman, cited in Silka & Kiesler, 1977; Eagly & Anderson, cited in Silka &
Kiesler, 1977); or sex role equity (Eagly & Anderson, cited in Silka & Kiesler (1977).

Susan Bram, (cited in Silka & Kiesler, 1977) however, did find in her dissertation
research, differences in attitudes and life style among couples intending no children, couples
intending children, and couples with children. The not intending groups had more egalitarian
marriages and more shared activities. The groups in Bram's study were not different in
background, social pressure, and attitudes about children.

Thus, the ideal number of children is conceptualized as a norm brought about by
social pressures and expectations. The decline in the ideal number of children, however, in the
last few decades has been attributed to structural reasons. Ory (1978) concluded that the
normative model is best for explaining general childbearing expectations, but the structural
model is best for explaining specific fertility outcomes within a norrnatively prescribed range
(eg. cost of additional child; concern for overpopulation; ability to balance work and home).
Thus, it is possible that structural factors may be especially useful in understanding the ceiling
on fertility choices after the first child more than the whether to parent or not choices.

In summary, the evidence that demographic and attitudinal factors relate to parenting

55

decisions, or distinguish between groups with different intentions to parent is mixed. For the
most part, demographic and attitudinal variables are not definitive in distinguishing between
groups of parents and nonparents or those intending parenthood and not intending parenthood.
This lack of congruence is expected according to the theory of reasoned action. None of these
external structural variables are expected to directly predict intentions or behavior leading to
parenthood. They can be useful, however, indirectly - to the degree that they influence beliefs
about outcomes of behaviors and the evaluation of those outcomes as positive or negative.
Thus, they may be of some use in influencing attitudes, and analysis of such relationships
could be pursued in future research.

Given how little we know of lesbian women's decision making, however, the inclusion
of a limited number of demographic and attitudinal variables could be useful. Of all the
possible demographic, material, and attitudinal variables, ones that are more directly related to
the daily care of children and changes in life created by parenthood would be more
illuminating than variables related to general life attitudes and personality. In the current
research, we asked demographic and background questions such as participant's ages, number

of siblings, experience with childcare, and questions about one's partner.

Summary
As the literature review reveals, there are many unknowns about lesbian parenting,
and about what factors are important for lesbian women who are considering parenthood.
There is almost no quantitative literature about how many lesbian women become mothers
after coming out as lesbians, nor how many lesbian women seriously consider parenthood.
There is no quantitative data on the factors weighed by lesbian women considering

motherhood, or the difficulties that may be encountered in making decisions. However, there

56

are sources of information about lesbian women in general, lesbian parenting, lesbian
relationships, and fertility decision-making for heterosexual people. Pulling information from
these empirical, theoretical, and narrative sources provided ideas about factors that may be
important for lesbian women as they contemplate parenting. The theoretical constructs that
were included in the model are summarized below. Lesbian decision-making appears complex,
yet the TORA brings together a limited number of components together in a comprehensive
fashion. The TORA model is proposed as a comprehensive model to understand decision-
making. It sheds light on the seemingly complex and mixed results in fertility research. In
TORA, decisions are based on attitudes and norms which determine intention. Behavior can be
predicted from intention, provided that intention is adequately measured.
0 f q ' i el

To summarize, based on the TORA model, attitude, norms, intention, and behavior
were selected as the components of the decision model. In terms of the attitude scale, the
literature provided many ideas about possible components of this attitude. For example, as
discussed, there has long been a conceptualization of lesbians and mothers as mutually
exclusive roles. In contrast, research on fertility decision-making for fertile heterosexual
couples often begins from a point which assumes parenthood as a role for these people, and
then looks for a range of behavior within this assumption. Because of these concepts, identity
as mother and nurturance were included as components of attitude. Nurturance has long been
assumed to be one of the main satisfactions that draws people to become parents, but in fact,
nurturance may or may not be linked with identity as a mother. Therefore, items concerning
nurturance, and identity as mother, were included as part of the attitude toward parenting scale
in the model.

Structural concerns -- those dealing with material conditions and demographics - were

57

considered as factors in fertility decisions in previous research, with mixed results. While
these kinds of factors were not very revealing in predicting behavior, Fishbein (1979)
suggested they may be useful in understanding attitudes. Moreover, material conditions were
considered salient reasons in choosing mt to parent. Therefore, items concerning material
issues - time, money, responsibility, health care insurance - were included as possible
aspects of the attitude toward parenting scale.

Part of the feedback from the social environment is about the perception of the family
as a real family created by a lesbian couple or lesbians with children. Moreover, the
relationship with a partner is one of those most intimately affected by the choice to parent, and
it is an important relationship. Therefore in the current research, items relating to children as
a completion of the family, and sharing this experience with a partner were included as part of
the attitude toward parenting scale.

The literature on norms suggests that feedback from people in the social network is
important as a message of normative expectations. It appears that the social acceptance and
encouragement of lesbian parenting is at a point of transition. Therefore, this research
included a set of items to measure the degree of expected support from others (including
heterosexual and homosexual friends, family members, and the lesbian community) for a
decision to become a parent, as well as items measuring the importance of support family,
friends, partners, and the lesbian community.

According to TORA, intention is expected to be the mediating step between attitude
and behavior, and between norms and behavior; and neither attitude nor norms are expected to
directly predict behavior. Intention is expected to predict behavior provided that it is measured
well.

In summary, the attitude toward parenting scale included the concepts of mother

58

identity, material factors, nurturance, and creating a family. The intention construct was based
on the degree to which one desired to parent and intended to parent. The behavior construct
was based on the number of activities engaged in which could lead to parenthood. The
proposed model examined attitudes, norms (via support), intentions, and behaviors, and
attempted to predict behaviors from intentions, and to predict intentions from attitudes and
norms.
if i f r R se

It is important to learn more about lesbian women's decisions to parent because of the
impact of cultural beliefs regarding parenthood and lesbian parenting upon individual level
decisions. Parenthood is so universal, the research on motivation includes research that asks
whether this is a conscious decision. Further, there is a long-standing assumption that lesbian
women are childless persons. Homophobia is revealed in questions such as 'what right do you
have to parent' and in the numerous myths regarding the fitness of lesbian parents. All of
these cultural beliefs may influence individual level choices regarding reproduction. '

The need to examine lesbian women's decisions to parent is also revealed by the fact
of limited data on lesbian families and parents. There is a lack of data or need for further
research on: lesbian desire and motivation to parent; decision factors in becoming a lesbian
parent; lesbian decision making processes; the relationships between lesbian identity and
identity as a parent/perceptions of parenting as an option; and the weighing of identity factors,
support factors, nurturance factors, and material factors as decision factors in becoming a
lesbian parent. Moreover, according to lesbian women, there is need for research. Lesbian
mothers see research on lesbian mothers as desirable, but would be wary and scrutinize the

researcher before participating in research (Corea, 1985).

59

Research Questions
De hic Ii ' n i s i i
1. What were the demographic characteristics of the sample? This included basic information
such as race, age, and education. It also included parental status, relationship, living
arrangement, and so forth.
2. What were the lesbian characteristics of the sample? This included the degree of lesbian
identity, degree of disclosure of one's lesbian identity to various people, and degree of
attendance at various types of lesbian/gay events.
'r ° i P
3. What were the desires / intentions in regards to parenthood? How many children were
desired?
4. What behaviors were engaged in that would lead to parenthood? What methods of achieving
parenthood were being considering or tried?
1 e n

5. Did the survey items that theoretically fit into anticipated scales (eg., nurturant, material,
mother identity, and created family items as part of the attitude scale, and expected support as
the norm scale) also fit according to statistical analyses? What items made up the scale and
which did not belong in any scale?

Went
6. Did the anticipated model best predict the intention to parent?

a. What were the relationships between the independent variables (those
predicting intention to parent, and those predicting behaviors related to
parenting)?

b. Did the attitudes toward parenting scale predict the intent to parent?

60

c. Did the norm variable (expected support) predict the intent to parent?

(1. Did the intention to parent predict the behaviors leading to parenthood?

METHODOLOGY
Setting
The setting for this research was the lower peninsula of the state of Michigan. The
information came from a survey that was distributed via: a mailing to lesbian women by a
lesbian organization that publishes a worldwide monthly newsletter, surveys given to lesbian
women at lesbian events and lesbian establishments in the state of Michigan, individuals in
communities who volunteered to distribute them, and placement in lesbian or lesbian/gay

bookstores. The surveys were distributed from May to June, 1995.

Procedure
Semantic
Participants in the survey were drawn from mailings of the survey instrument, through
bookstores, through individual distribution, and through in-person distribution of the survey
instrument at lesbian establishments or events. The purpose of this approach was to include a
wide variety of participants, and include women who may not have been on mailing lists. A
snowball sampling procedure was used to gather participants. The difficultly of sampling
groups that are difficult to contact, or who are subjectively defined by the individuals
themselves, can be minimized by the use of snowball sampling, in which informants are able
to introduce researchers to other members of the group (Burgess, 1982; Bybee, 1990).
Several organizations in Michigan were initially approached with a request to help in
the research, and they were chosen based on three criteria. First, the organizations were
lesbian or lesbian/gay organizations. Second, organizations were from various cities,
representing different geographic areas. Third, the organizations had mailing lists of over 100

people, or were the largest or most nonspecialized of the lesbian newsletters in that area. The

61

62

organizations approached operate for the benefit of the lesbian or homosexual population.
Their primary purposes were social, educational, support, and/or political. The contact
persons in these organizations proved to be very helpful. They provided ideas about ways to
distribute surveys, connected the principal investigator to other contacts, and allowed the
principal investigator access to events being produced in order to distribute and collect
surveys. See the appendix for an example of an initial contact letter from the principal
investigator to an organization.

The surveys were marked with color codes on the side (binding area) of the survey to
determine where the survey was distributed from, so that the return rates the various
distribution points could be estimated. There were a total of 14 distribution points, which
were aggregated into 5 distribution points for descriptive purposes. The success of the 14
distribution strategies ranged from a low of 10% (a bookstore), to a high of 55.5% (an event).
The return rates for the 5 aggregated distribution strategies were as follows: 46.7% from the
lesbian Connection mailing; 30.8% from bookstores, 51% from distribution at events by the
principle investigator, 36% from a lesbian mother's support group mailing, and 50.8% from
in-person distribution by the principal investigator and other people. A total of 638 surveys
were distributed, of which 299 were returned, for an overall return rate of 46.9%. Eight of
the surveys returned were unusable because they were filled out by non-Michigan residents,
they were filled out by more than one person, or they were returned after data entry was
completed.

The sample consisted of surveys returned from the various distribution points: 10.4%
from surveys individually distributed by the principal investigator, 3 % from the lesbian
mother's support group, 3.8% from bookstores, 57.4% from the LC mailing, 24.7% from in-

person distribution at lesbian events, and 0.3 % from an unknown source (a photocopy of the

63

survey was returned).

MM

As noted previously, the surveys were mainly distributed in two ways - via a mailing
and through in—person distribution at social events. Surveys were also placed in bookstores and
distributed to acquaintances by the PI and assistants in Saginaw and Kalamazoo. In-person,
surveys were distributed at a lesbian social party, concerts, and other events. The principle
investigator and an assistant solicited participation in the survey at these places and events. We
encouraged people to fill out the surveys at the time' of the event, or to take them home, in
order to maximize the completion of the surveys. Priority was given to data collection at sites
that were in geographically diverse settings, at settings conducive to data collection, and
settings in which a racially diverse group was expected.

As the surveys were returned, the Pl monitored the proportions of participants
intending to have children. It was determined that additional effort was needed to distribute
surveys to women who were likely to favor becoming parents, in order to have enough
variance in the intention variable. To this end, surveys were mailed to 25 women in a small
local support group for lesbian women seeking to have children, 9 (36%) of which were
completed and returned.

Surveys were mailed to women on the Michigan mailing list of the "Lesbian
Connection" newsletter. The staff of the "Lesbian Connection" newsletter allowed the survey
to be mailed to 368 women, which is one-quarter of the persons on their Michigan mailing
list. The Lesbian Connection (LC) newsletter has the largest lesbian mailing list in the world,
which includes approximately 1,400 lesbian women in Michigan. Because of this extensive
coverage, the LC was the only publisher mailing list necessary for this research. Also, other

mailing lists would overlap with the LC list to a large extent.

64

The surveys and return envelopes were prepared by the principle investigator, and
given to the Lesbian Connection (LC) staff for mailing. Because the mailing list is strictly
confidential, the LC staff attached the mailing labels to the prepared surveys, and sent out the
survey as a separate mailing from their newsletter. LC received payment for their labors and
materials. The mailing included a cover letter from LC staff that explained the purpose of the
mailing and reassured recipients that the mailing list remained confidential. See appendix for a
copy'of the LC cover letter.

The survey was cream colored and contained 10 pages of questions along with a cover
page and back page. The survey was printed double-sided on 11.5 by 17 inch sheets of paper,
folded and stapled, so that it read as a booklet with standard 8.5 by 11 inch size pages. A pre-
paid business reply envelope was included with each survey to maximize returns and ensure
anonymity. In addition, a 10% discount coupon and a flyer listing merchandise were included
with every survey. The discount coupon was contributed by a nationwide disu'ibutor of
women's music that operates out of the Lansing, Michigan area. This was included as an

incentive to participate and token of appreciation for participation in the survey research.

Research Participants
The research participants consisted of 163 adult women who were self-identified
lesbian women living in the state of Michigan. A power analysis determined that a sample size
of 163 participants would provide power in excess of .80 to detect an effect size of R2 = .10
for six predictors, at an alpha level of .05 for the path model developed in this research.
The sample consisted of all the nonparents under the age of 45 taken from a larger
sample (N =291) of surveys respondents. The use of only nonparents under the age of 45 was

dictated by the research questions proposed. The main thrust of the research proposed was the

65

prediction of behaviors leading to parenthood and intentions to become a parent. Thus, it was
decided to include in the analyses only those women who were of an age at which parenting is
typically pondered. A crosstabulation of age by intentions to parent revealed that there was a
dropoff at age 45 in the intention to become a parent. Of women aged 40-44 who responded to
the survey, 15.9% said they intended (or most of the time they intended) to have children. Of
the women age 45 or older who responded to the survey, only 2 (2.9%) said they intended to
have children. Thus, age 45 became the cutoff point.

In terms of examining only nonparents, the primary questions in this research
concerned understanding lesbian women taking up the role of parenthood, rather than lesbian
women who are already parents choosing to have additional children. lesbian women who are
already mothers know the role of parent; they do not need to imagine the role, they do not
need to question whether this is a role with which they can identify. In addition, this research
was focused on how lesbian women approach the decision to parent, afier coming out as
lesbians. The mothers who responded to the survey included those who had children before
and after coming out as lesbians. Thus, for this research, becoming a parent at all and
choosing to have additional children are separate issues. Moreover, one could hardly predict
that which has already happened. Thus, only nonparents were chosen for this analysis.

Table 1 presents the demographic characteristics of the 163 research participants (non-
mothers under 45 years of age). As shown in the table, less than one in five respondents were
age 29 or less (18.4%). Nearly one in four respondents (39.9%) were age 30 to 34 and 41.7%
were age 35 to 44. The mean and median age was 34 and the standard deviation was 5.6
years. The majority of the sample was caucasian women (87.1%), which is similar to the

proportion of white people in the state of Michigan overall (83 %). Although some effort was

66

made by the P1 to include African-american women, the sample included only 3.1% African-
american women, which was less than the overall proportion of African-american people in
Michigan (13.9%).

The women in the sample were highly educated. About one in twenty (4.9%) of the
women had only a high school education, and 21.4% had vocational training or some college.
Nearly three-quarters of the women in the sample had a bachelors degree or a higher degree:
29.4% of the women in the sample had a 4-year college degree, 13.5% had some graduate
schooling, and 30.7% had graduate degrees.

The majority (92.5%) of the women in the sample were employed. The annual
household income ranged from $5,000 to $185,000. The mean annual household income was
$44,704, the median income was $40,000, and the standard deviation was $26,580. Nearly
one in three (28.7%) women in the sample had incomes of less than $30,000, 26.3% Of the
sample had annual household incomes from $30,000 to $49,999, and 35.0% had incomes of
more than $49,999.

Nearly two-thirds (64.4%) of the women in the sample were in monogamous
relationships, and 31.9% were single. The average length of time in the monogamous
relationships was nearly 5 years (56 months), the median was 4 years, and the standard
deviation was about 3.5 years (44 months). Of the women in monogamous relationships,
nearly a third (31.7%) had been in the relationship less than three years. Nearly one in four
(37.5%) of the women had been in the monogamous relationship between 3 and just under 6
years, and 30.8% had been in the relationship between 6 and 16 years.

Thus, the women in the sample tended to have many external resources, such as
education, high income, and employment. The majority were in monogamous relationships

averaging nearly five years, were white, and the median age was 34. Nearly 6 in ten women

(57.7%) lived with a partner or lover.

67

[Characteristic N = 163

68

Table 1.

Demographic Characteristics

 

 

 

 

 

 

Age
19-24 7 4.3
25-29 23 14.1
30-34 65 39.9
35-39 29 17.8
40-44 39 23.9
Race
African American 5 3.1
Caucasian 142 87.1
Other Minority 16 9.9
Education
High School or GED 8 4.9
Vocation or some college 35 21.4
Bachelors degree 48 29.4
Some graduate/professional school ' 22 13.5
Graduate/professional degree 50 30.7
Employment
Employed 148 92.7
Unemployed 2 1.3
Student 9 5.6
Homemaker 1 0.6
' Anmral household income
$5,000-19,999 16 10.2
$20,000-29,999 29 18.5
$30,000-39,999 29 18.5
$40,000-49,999 28 17.8
$50,000-69,999 26 16.5
$70,000-l85,000 29 18.5
Relationship status
Monogamous relationship 105 64.4
Single 52 31.9
Other 6 3.7
3 Length of monogamous relationship, N =104
less than 1 year 12 11.5
1 year to 35 months 21 20.2
3 years to 47 months 17 16.3
4 years to 71 months 22 21.2
6 years to 95 months 14 13.5
18

8 years to 16 years

 

 

17.3

69
r risti f W0 11 N l ' l

The women who were not included in the final sample due to age or parenting status
are described here. Of the 128 women not included in the sample, 43% were parents under
the age of 45, 22.7% were parents over age 44, 31.3% were nonparents over age 44, and
3.1% were missing data on one of these variables. In all, 66.9% of those not included in the
sample were parents. Of those over age 44, 42% were parents.

The mothers under 45 years of age who were not included in the sample had 44.4% of
their 72 children after they "came out" as lesbian women. The mothers over 44 years of age
who were not included in the sample only had 10.2% of their 65 children after they "came
out" as lesbian women.

The women in the two groups not included in the sample and the women in the sample
were similar in many ways. Several tests were conducted to look for differences in these
groups. Analysis of variance revealed that there were no statistically significant differences in
the three groups (sample, young lesbian mothers, and older lesbians) in terms of the degree of
lesbian identity, the annual household income, the degree of disclosure to their families of
origin, and the degree of disclosure to straight friends. In addition, crosstabulation analyses
showed there were no statistical differences between the 3 groups in terms of dichotomized
race or ethnicity, residing with one's partner, nor in having close lesbian friends who had
children. Furthermore, a comparison between the sample and the younger mothers showed no
differences in terms of intentions to have children (older lesbian women were not included in
this analysis).

There was a significant difference between the three groups, however, in terms of
having a partner (71.2 (4, N = 274) = 13.4, p = .009). Almost all (90.7%) of the younger

mothers were in monogamous relationships, compared to 71.2% of the women in the sample

70

and 65.7% of the older lesbian women. In addition, there was a difference in length of the
current monogamous relationship (analysis included only women in a relationship) (F (2, 190)
= 15.5, p = .00). The women in the sample (m = 56 months, n = 104) had the shortest
relationships, the young mothers (m = 86 months, n = 45) had longer relationships than the
women in the sample, and the older lesbian women (mothers and not mothers) (m = 120
months, n = 41) had the longest relationships.

There was also a difference between the 3 groups in the prOportion of women who had
close lesbian friends who were actively trying to have children (71.2 (2, N = 282) = 8.9, p =
.011). Over one-quarter of the women in the sample (27.3%) and in the older lesbian group
(26.9%) said they had close lesbian friends trying to have children, but 48.1% of the younger
lesbian mothers said that they had close lesbian friends trying to have children. A final
difference was in the amount of education (F (2, 287) = 3.7, p = .03). The women in the
sample tended to have bachelor degrees, (m = 6.37, n = 163) whereas the young mothers (m
= 6.42, n = 55) and the older lesbians (m = 6.9, n = 69) were more likely to have graduate
degrees.

In summary, there were few demographic and disclosure differences between the
sample and the 2 groups not included in the final sample. There were, however, other
differences in these groups. The young mothers were more likely to be in monogamous
relationships, and to have close lesbian friends trying to have children, whereas the older
lesbians were more likely to have longer relationships. Both groups (young mothers and older
lesbians) that were not included in the sample tended to have more education than the women

in the sample.

71

Survey Instrument

The survey instrument was created based on a review of theoretical and empirical
literature, pilot testing data, and pretest feedback. The format and questions were developed
by the principal investigator with input from committee members and six pretest participants.
The surveys were strictly confidential and anonymous, and contained closed-ended and open-
ended questions. The survey included sections regarding background information; desires and
intentions about becoming parents; perceptions about cultural assumptions about lesbians as
parents and the legal treatment of lesbian mothers; expected support from others; and what
activities or interventions would be useful to lesbian women in contemplating parenthood.
There was also a section of attitudinal questions with a 5-point response set ranging from
"strongly agree" to "strongly disagree", that focused on several aspects of parenting expected
to be important in understanding choices to parent or not to parent: motherhood identity;
nurturance; concerns about lesbian parenting; material factors; and creating a family. For
those who were interested in becoming parents, there were also questions about whether they
had engaged in activities related to becoming a parent (cg. charting ovulation, contacting an
adoption agency), and what methods they had seriously considered. See the appendix for a
copy of the survey instrument.

2.] I 'n

The pilot testing of the survey instrument was conducted at the Michigan Womyn's
Music Festival, and in the greater Lansing area, in August of 1994. The participants were 28
adult lesbian women who were solicited for participation at the Festival, acquaintances of the
principal investigator, or acquaintances of participants in the pilot study (ie. snowball
participants).

Confidentiality was assured because no names were written on the survey instrument.

72

The investigator accepted some of the completed surveys in person, but did not store or mark
them in any manner that would identify a participant.

The pilot data were used to further refine the content of the instrument, and the model
of decision-making. The data was entered into a systems file and analyzed using SPSS-PC.
Analyses of the pilot data included an examination of the open-ended items and quantitative
frequencies. Reliability procedures were run on sets of items that were expected to measure
constructs of the model.

P e in f e rv

The survey was revised after examination of the pilot data and consideration of the
research questions. The pretest survey was completed by 6 participants who were acquainted
with the principle investigator. These women varied in their intentions to parent. Evaluation of
the pretest survey included an examination of the responses as well as written feedback from
the participants concerning their experiences in completing the task. Additional revisions were

then completed based on this information.

Scale Development
11 in

It was expected that there would be a series of scales that would be useful in modeling
the prediction of parenting intentions and behaviors: motherhood identity, creating a family,
nurturance, and material concerns. These attitude toward parenting survey items were created
based on the theoretical relevance of items and pilot testing reliability analysis of the items.
Most of the anticipated attitude toward parenting scales appeared feasible based on the initial
reliability procedures. These four scales were so highly intercorrelated, however, that they had

to be considered subscales and were combined into one scale representing the attitude toward

73

parenting construct. Table 2 shows the intercorrelations between these attitude toward

parenting subscales. The intercorrelations have been corrected for attenuation.

Table 2.

Intercorrelations Between Attitude Toward Parenting Subscales

Creating a Mother
Family ldentity__

 

Mother Identity

 

 

 

 

Table 3 shows that reliability analyses of the attitude toward parenting subscales.
These internally consistent subscales were not used in the prediction model. They are shown
here to illustrate the attitudinal components initially hypothesized to predict the intention to
parent, and to illustrate the process from which the final single attitude toward parenting scale
was created. Table 3 provides for each subscale the number of items, the variable label
corresponding to the question number in the survey, the item wording, the range of values for
the corrected item-total correlations, the number of cases with no missing data on every item,
the items that did not load in the procedure, and the alpha from the reliability procedure. An
"(R)" indicates that the item was reflected in coding, so that direction of scales were parallel.
A high value on the scales indicates an attitude consonant with parenting, lower values

represent positive attitudes toward not parenting.

74
Table 3.

Reliability Analysis of Items in Attitude Subscales

SUBSCALE: MATERIAL

1. C16 I don't have enough time to be a parent

2. C33 I don't want to interrupt my job-career

3. C40 I would rather spend money on other things

4. C41 I don't want to be tied down with children

5. C43 I don't want the responsibility of having children

Number of cases = 162
Alpha = .91
Range of values for corrected item-total correlations = .65 - .88

Item not included:
C17 It's important to have insurance before having a child

SUBSCALE: MOTHERHOOD IDENTITY

1. C28 Being a parent does not fit with how I see myself

2. C37 1 would not feel comfortable as a mother

3. C49 When I think of the future, I don't see myself having children
4. C11 I have never seriously considered having children

5. C24 1 have never pictured myself as a mother

Number of cases = 160
Alpha = .90
Range of values for corrected item-total correlations = .70 - .83
Items not included:
None

75
Table 3. (cont'd.)

SUBSCALE: NURTURANCE TOWARD CHILDREN

C7 (R) I want to love a child

C9 (R) I want to nurture a child & help her or him grow
C13 (R) Caring for a child is a joyful experience

C27 (R) I want the challenge of raising a child

Number of cases = 158
Alpha = .86
Range of values for corrected item-total correlations = .61 - .79
Items not included:
C10 1 don't like being around children
C14 I don't want to experience pregnancy or childbirth
C46 I find children too stimulating

SUBSCALE: CREATING A FAMILY

1. C8 (R) Having a child would make my family complete

2. C35 (R) Having a child is an experience I want to share with a partner
3. C36 (R) Having a child would increase the bond with my partner

4. C39 (R) Having a child would create a sense of family

Number of cases = 129
Alpha = .85
Range of values for corrected item-total correlations = .63 - .79
Items not included:
C45 My partner doesn't want children

 

76

The listing of scales in Table 4 is a summary of three of the four main constructs used
in building the models predicting behaviors leading to parenting. The 3 components of the
models shown here were attitude toward parenting, intention to parent, and behaviors engaged
in that would lead to parenthood (scales related to norms are shown on another table). As
noted, the subscales of nurturance, motherhOod identity, material, and creating a family were
highly intercorrelated, and therefore a single WW scale was created from
the items that were in these subscales. The attitude toward parenting scale consisted of the
mean of the responses to the items, so that every respondent had a single score for this scale.
A summary of the reliability analysis is shown in Table 4. Attitude toward parenting was a
strong scale; the alpha for this 13 item scale was .95.

n i e i r

The scale for measuring one of the dependent variables, W, was
initially anticipated to consist of one item: "What item below best describes your plans about
becoming a parent? Do you intend to have a child / children / another child?. " This was
followed by five complete sentence response options which ranged from: do not intend, most
of the time do not intend, am not sure if I intend, most of the time intend, and definitely
intend. Conceptually and empirically, however, another item also warranted inclusion: "What
item below best describes your feelings about becoming a parent? Do you want to have a child
/ children / another child?. This was followed by a similar response set, substituting "want" in
place of "intend". Thus, an average of these two items were used to measure intention to
parent. These items correlated .87 and had an alpha of .95, as shown in Table 4.

The other dependent variable, WWW, consisted of 15 items
which described various activities, such as charted ovulation cycle, talked with other lesbians

who are mothers, applied for foster parent status, and so on. Respondents were asked if they

77

or their partner had engaged in any of the activities in the past two years, to which they
responded "yes" or "no." The Behavior variable consisted of counting the number of "yes's"
from all the items. The alpha of this 15 item scale was .70. Unlike the other scales, internal
consistency is not important for the validity or the utility of this particular scale. This scale
merely adds the number of behaviors engaged in. It is not an attempt to measure an
underlying construct.

N rm ' m en

One of the pieces to be explored in this study predicting behaviors leading to
parenthood was that of norms. In this study, it was suggested that asking respondents whether
they expected support from specific others in their lives would be an adequate indication of the
perceived normative messages regarding the appropriateness of parenthood for the
respondents. Thus, W for becoming a parent was a
normative construct proposed in this research. A section of the survey asked what how much
support was expected if one chose to parent, on a five-point scale from "very supportive" to
"very supportive. " The items listed 14 different people, such as specific relatives, fi'iends,
partner, and co-workers, for whom support data was requested.

As shown in Table 5, a reliability procedure revealed that this scale consisting of
items measuring the expected support from various people was not very internally consistent.
Although the level of support from various sources were not consistent with each other, this
did not necessarily mean that there was a problem with the measurement of this support or
that the scale is not useful. It is entirely feasible that the expected support would vary from
person to person. These items were useful as an average of the expected support from many
sources. Therefore, the lower corrected item-total correlations were not an issue for this

particular scale. The alpha for this 13 item scale was .81.

78

The expected support from people items had notable amounts of missing data. The
high rate of invalid responses was not necessarily due to reluctance to answer, however, but to
respondents who reported that they "did not have these people in their lives/not applicable ".
For example, 14.7% said expected support from the mother was not applicable and 1.2% did
not answer. In terms of expected support from the lesbian community, 4.9% said this was not
applicable and 0.6% did not answer, and 30.7% said the expected support from the sister was
not applicable and 1.8% did not answer. The reasons for a person being "not applicable'
were not requested in the survey. It is expected that this could be due to a number of
possibilities, such as not having siblings, deceased relatives, or a disconnection from members
of their family of origin.

Because this initially hypothesized normative component of the model - expected
support from others - did not emerge as a predictor in the model testing, other data in the
survey were examined as potential indicators of a normative construct. Several of the post-hoe
scales were additional efforts at examining the expected support construct. These expected
support scales were made up primarily of items that were subsets of the original expected
support scale, but they measured support from family, support from friends, and support from
lesbian sources separately instead of together. Other scales related to support and created for
use in the model testing were importance of support from others, and affiliation with lesbian
mothers or lesbians who desired to be mothers (another indication of support from others).
The reliability analysis information for these scales is given in Table 5.

The scale for W consisted of three
items. Two of the items measured the amount of expected support of lesbian friends without
children and expected support of the lesbian community. The third item in the scale asked

whether lesbian friends support the decision to have or not have children, with a strongly

79

agree to strongly disagree response set. All three items had a five-point response option. The
alpha for this expected support of lesbian friends and community scale was .73.

The expeeteg guppgrt from family scale consisted of five items. These items measured
the amount of expected support from various members of the family of origin -- father,
mother, sisters, brothers, and other relative. All items had a five-point response Option. The

alpha for this expected support from family scale was .83.

The scale for WW consisted of four items. These items

measured the amount of expected support from heterosexual and lesbian friends, with and
without children. All items had a five-point response option. The alpha for this expected
support of friends was .73.

A scale for W consisted of 3 items measuring whether the
respondent had any close lesbian friends who: had children, were thinking about having
children, or were actively trying to have children. The response options were "yes" and "no"
and the variable was computed by counting the number of "yes's" to these items. This resulted
in a variable with values ranging from 0 to 3. The alpha for this scale was .58

W consisted of 3 items measuring the degree of importance of
support for decision to parent or for decision to not parent from friends, family, and the
lesbian community. The response options for these items, as for the attitudinal items, was a
Likert scale ranging from "strongly agree" to "strongly disagree. " The alpha for this scale
was .77.

Another attempted modification to the model was the use of a variables measuring
perceived societal norms regarding the appropriateness of lesbian parenting. One measure

concerned whether the legal treatment of lesbian mothers was a factor in respondents'

decisions to parent or to not parent. This single item asking if the WM

80

Me was a factor in the choice to parent was coded on a 5-point scale from strongly agree

to strongly disagree. The other two single item variables used in testing the model asked

whether the respondents' perceived that most WW
mothers. and whether most WWW. again using
a five-point Likert agree-disagree response set.
i ' V ' 1

Additional variables were included as modifications to the path model. This included
length of time in a monogamous relationship. Because the initially lgpothesized components
of the model - ie., creating a family, motherhood identity, nurturance toward children, and
material reasons - turned out to be subscales of a more general attitude toward parenting
construct, other data in the survey were examined as potential indicators of those initial
constructs. For example, creating a family was a subscale that dealt with having a sense of
family with one's partner and child, of children influencing the bond between partners. So,
upon further reflection, it was hypothesized that other variables dealing with partnership could
be indicators of the "creating a family" construct. length of time with one's monogamous
partner was deemed an acceptable measurement of family, and was entered into the equation.
This variable was desirous because there was no missing data - those without a partner would
have a valid value of zero, and because the length of time in a relationship takes on a special
meaning for lesbian couples, as a representation of external conventional validation of their
union (Slater & Mencher, 1991).

Another ad hoc adjustment was the consideration of demographic variables. One
demographic variable considered to be salient vis-a-vis parenting was age. Parenthood is
assumed to be a choice made only by those of "child-bearing" age. The question arises as to

whether one is increasingly likely to choose non-parenthood as one ages and is still childless.

81

Similarly, number of siblings was used to modify the model because it has been examined as a
variable related to fertility decisions in previous research.

Thus there were several continuous variables that were used in the attempt to build the
path model that were not scales consisting of several items. One was gge, measured in years,
another was number 9f siblings, with values ranging from 0 to 11. Another continuous

variable was the W with a monogamous partner which was coded

into number of months. Respondents without a partner were assigned a valid value of zero for
this variable. The values for this variable ranged from 0 (35.8%) to 192 months.

To summarize the development of scales, most of the items anticipated to create 4
attitude scales of nurturance, motherhood identity, material concerns, and creating a family
were found to more appropriately form one total scale representing attitude toward
parenthood, due to the high intercorrelations of said items. The intention scale was a mean of
desire and intention variables. The behavior scale was created as anticipated as a count of the
number of behaviors engaged in. Finally, several scales were created to represent a normative

construct. The internal consistency of most of these scales were adequate as shown in Tables 4

and 5.

82
Table 4.

Reliability Analysis of Items for Final Scales for Attitude, Intention, and Behavior Constructs
ATTITUDE TOWARD PARENTHOOD

C16 I don't have enough time to be a parent

C40 I would rather spend money on other things

C41 I don't want to be tied down with children

C43 I don't want the responsibility of having children

C28 Being a parent does not fit with how I see myself

C37 I would not feel comfortable as a mother

C49 When I think of the future, I don't see myself having children
C7 (R) I want to love a child

C9 (R) I want to nurture a child & help her or him grow

10. C13 (R) Caring for a child is a joyful experience

11. C27 (R) I want the challenge of raising a child

12. C8 (R) Having a child would make my family complete

13. C35 (R) Having a child is an experience I want to share with a partner

PPSP‘S‘PPPT‘

Number of cases = 126
Alpha = .95
Range of values for corrected item-total correlations = .65 - .89
Items from subscales not included:
C33 I don't want to interrupt my job-career
C11 1 have never seriously considered having children
C24 I have never pictured myself as a mother
C36 (R) Having a child would increase the bond with my partner
C39 (R) Having a child would create a sense of family

INTENTTO PARENT

1. Bl Describe feelings about becoming a parent
2. B2 Describe plans about becoming a parent

Number of cases = 163
Alpha = .93
Value of corrected item-total correlation = .87

83
Table 4 (cont'd.)

BEHAVIORS LEADING TO PARENTHOOD

F201
F203
F204
F205
F206
F2Q7
F2Q8
F209

9999999.“?

Contacted an agency regarding a home study
Talked with lesbians who are mothers
Talked with lesbians who want to be mothers
Charted ovulation cycles

Contacted sperm bank(s)

Discussed ideas with my partner

Discussed ideas with potential sperm donors
Inseminated under a doctors supervision

F2010 Inseminated at home

10. F2Q11 Used ovulation regulating drugs

11. F2Q14 Discussed ideas with potential "surrogate" mother
12. F2Q16 Other activity mentioned

13. F2Q2 Had a home study for adoption completed

14. F2Q12 Applied for foster parent status

15. F2Q13 Tried to adopt

Number of cases = 90

Alpha = .70

Range of values for corrected item-total correlation = -.01 - .65

 

84
Table 5.

Reliability Analysis of Items for Final Scales for Subjective Norm Construct

 

NORMS: EXPECTED SUPPORT OVERALL

D1
D2
D3
D4
D5
D6
D8
D9
D10
10. D11
11. D12
12. D13
13. D14

PPHP‘MPPPT‘

(R) Mother's support

(R) Father's support

(R) Brother's support

(R) Sister's support

(R) Other relative's support

(R) Partner's support

(R) lesbian community support

(R) Partner's family support

(R) Straight friends without children support
(R) Straight friends with children support
(R) lesbian friends without children support
(R) lesbian friends with children support
(R) Co-worker's support

Number of cases = 30

Alpha = .81

Range of values for corrected item-total correlations = .23 — .73

 

NORMS: IMPORTANCE OF SUPPORT

1. C4
2. C5
3. C6

(R) It is important for me to have the support of my friends in my decision to
parent OR to not parent

(R) It is important for me to have the support of my family in my decision to
parent OR not to parent

(R) It is important for me to have the support of the lesbian community in my
decision to parent OR to not parent

Number of cases = 163

Alpha = .77

Range of values for corrected item-total correlation = .56 - .64

 

85
Table 5 (cont'd.)

 

NORMS: EXPECTED SUPPORT FROM FAMILY

D1 (R) Mother's support

D2 (R) Father's support

D3 (R) Brother's support

D4 (R) Sister's support

D5 (R) Other relative's support

999.“?

Number of cases = 48
Alpha = .83 .
Range of values for corrected item-total correlations = .54 - .75

 

NORMS: EXPECTED SUPPORT FROM FRIENDS

D10 (R) Straight friends without children support
D11 (R) Straight friends with children support
D12 (R) Lesbian friends without children support
D13 (R) Lesbian friends with children support

#UJNI—t

Number of cases = 137
Alpha = .73
Range of values for corrected item-total correlations = .37 - .70

 

NORMS: EXPECTED SUPPORT FROM LESBIAN FRIENDS AND COMMUNITY

1. D12 (R) Expected support from lesbian friends without children

2. D8 (R) Expected support from the lesbian community

3. C22 (R) My lesbian friends will support my decision to have or to not have
children

Number of cases = 150
Alpha = .73
Range of values for corrected item-total correlations = .52 - .60

 

86
Table 5 (cont'd.)

 

NORMS: AFFILIATION WITH LESBIAN MOTHERS

1. A7 Have close lesbian friends who have children
2. A8 Have close lesbian friends thinking about having children
3. A9 Have close lesbian friends actively trying to have a child

Number of cases = 161
Alpha = .58
Range of values for corrected item-total correlations = .34 - .42

 

NORMS: INDIVIDUAL ITEMS

1. C2 Most lesbians assume that lesbians do not become mothers
2. Cl Most people assume that lesbians do not become mothers
3. C3 The way the law treats lesbian mothers is a factor in my decision to parent or

to not parent

 

87
Model Development

WM

Structural equation modeling (SEM) is a comprehensive statistical approach to testing
hypotheses about relations among observed (measurable) and latent variables. SEM begins
with the specificatiOn of a model to be estimated. A model is a statistical statement about the
relations among variables. The parameters that require specification are constants that indicate
the nature of the relation between two variables. Fixed parameters are not estimated from the
data and their value typically is set at zero. Free parameters are those estimated from the data
which the investigator believes to be nonzero. There are various indices of model adequacy,
which indicate the degree to which the pattern of fixed and free parameters specified in the
model is consistent with the pattern of variances and covariances from a set of observed data.

The SEM procedures emphasize comfiwtces rather than cases. It minimizes the
difference between the sample covariances and the covariances predicted by the model, rather
than minimizing the functions of observed and predicted individual values. The fundamental
hypothesis for these structural equation procedures is that the covariance matrix of the
observed variables is equal to the matrix implied by the model. The covariance matrix of the
observed variables is a function of a set of parameters; if the model were correct and we knew
the parameters, the sample covariance matrix would be exactly reproduced. The residuals are
the difference between the observed and implied covariance matrices, and are the foundation
for overall model fit indices. Regression analysis, simultaneous equation systems,
confirmatory factor analysis, and ANOVA are all special cases of the fundamental hypothesis
in structural equation modeling.

The emphasis of the SEM is on systems of linear equations. Relations between all

variables can be represented in linear structural equations. More specifically, it is the

88

structural equations linking the observed, latent, and disturbance (error) variables that are
linear, and not necessarily the covariance structure equations.

The relationships among the variables of interest —— attitudes, norms, intentions, and
behaviors - were examined with the use of a structural equation model (SEM). The model
hypothesized in this study is a path model - it included only observed variables, and no latent
variables. This model tested the linear relations among the variables with the use of LISREL
software. The analysis was essentially a multiple regression with 2 dependent variables
(intention and behavior). With SEM, one can calculate all regression equations simultaneously,
as well as obtain other indices useful in modifying the model and indices evaluating overall
model fit. These data are not available with a multiple regression analysis.

A nested comparison of alternative models is one in which the components in one
model are a subset of components in another model. Such is the case in the current study. In
the current study post-hoe modifications were made to the model (model #1). In post-hoe
modifications, there is potential for capitalizing on chance or the idiosyncracies of the
particular sample. This possibility is more likely with small samples with which the likelihood
of finding a replicable model is quite low (Hoyle & Panter, 1995). Although a priori
modifications are preferred, post-hoe modifications can be quite useful provided they are well-
grounded in theory. The post-hoe modifications attempted in this study were pursued based on
the following reasons: investigating the usefulness of additional variables as indicators of the
constructs under investigation, and comparing the parsimonious first model with a less-

parsimonious alternative model.

89
WW

Although predictive models were tested in this research, the goal was not to confirm
any model per sc. Rather, the goal in this exploratory work was to use the model as a
framework for exploring relationships among variables in the decision to parent. The proposed
SEM was an effort to describe and predict lesbian women's decisions to become parents or to
not become parents. Several questions were addressed: (1) what were the relationships
between predictors (attitude and norms) of the dependent variables (intention and behavior)?:
(2) what was the relationship between attitude toward parenthood and the intention to parent?;
(3) what was the relationship between the subjective norm variable (expected support) and the
intention to parent?; and (4) what was the relationship between the intention to parent and the
behaviors engaged in which lead to parenthood?

The research questions as originally conceived included a query regarding the
relationships between the various attitude toward parenting scales (ie., between mother
identity, material conditions, nurturance, and creating a family). These attitude subscales were
to predict intentions to parent according to the model. As discussed previously in the scale
development section, however, it was clear empirically that attitudes toward parenthood scales
were really subscales of the same construct. Thus, the first question became: (1) what are the
relationships between the total attitude toward parenthood scale and any other predictors (in
this case, subjective norm variables) of the dependent variables in the model.

In the initial conceptualization it was thought that behaviors should include activities
leading to parenthood (cg. contacting an adoption agency) as well as actual current
parenthood. Upon further reflection, however, it was concluded that prediction of a status
already experienced was not conceptually sound. This idea was abandoned in favor of

predicting behaviors leading to parenthood only for women who were not currently parents.

90
Evl 'no lFi

The Maximum Likelihood (ML) estimation method was used to output the data. The
robustness of estimators to violations of assumptions is an important issue for empirical study.
West, Finch, and Curran (1995) assert that ML estimation is still appropriately used even for
smaller sample sizes, when distributions are not substantially nonnorrnal.

The covariance structure hypothesis is that the observed data matrix is equal to the
covariance matrix of the model parameters. Basically all measures of overall model fit involve
functions of the observed covariance matrix and the covariance matrix implied by the
hypothesized model. The fit indices assess the closeness of these two matrices, though the
closeness is measured in various ways. The advantage of overall fit indices is that they
evaluate the whole model and can indicate inadequacies not revealed by the fit of the model
components (eg., parameter estimates).

As recommended by Hoyle and Panter (1995), several indicators of overall fit were
used to assess the validity of the conceptual model. The use of more than one absolute fit
index, at least one type-2 incremental fit index, and at least one type-3 incremental fit index
were suggested. All fit indices used in this study, excepting the chi-square, are descriptive
statistics; they do not allow for testing of statistical significance.

The chi-square variatc and the goodness-of-fit index (GFI) were the measures of
W used in this study. Absolute fit measures concern the degree to which the
covariances implied by the fixed and free parameters specified in the model match the
observed covariances from which the free parameters in the model were estimated. These are
actually badness-of-fit indices - an optimal fit provides a value near zero.

W concerns the degree to which the model in question is superior to an

alternative model, in this case the null model, in reproducing the observed variances. These

91

gauge the goodness of fit, with larger values indicating a greater improvement of the model
over the alternative model in reproducing the observed covariances. Two incremental fit
indices were used to assess the model fit: a type-2 index, the nonnormed fit index (NNFI),
and a type-3 index, the comparative fit index (CFI). NNFI performs well when ML estimation
is used. CFI is a preferred index because the values fall only between 0 and 1, thus providing
a familiar normed range.

In this model, the critical value indicating an acceptable fit for GFI, NNFI, and CF]
was .90 for each. In other words, a value of .90 or greater for any of these indexes indicates
an acceptable fit of the hypothesized model to the data. Critical values for overall fit indexes
in the standard sense are not defined, because the sampling distributions of overall indices of
fit are unknown. An agreed-upon cutoff of .90 for overall fit indexes, however, has become a
standard practice (Hoyle & Panter, 1995, p. 164).

The goodness-of-fit test statistic, T, is used for hypothesis testing to evaluate the
appropriateness of an application of SEM and is equal to the product of minimized F and (N-
1), where N is the sample size. This chi-square test is a simultaneous test of the null
hypothesis that all residuals from the two matrices are zero, ie., that the model fits the data
perfectly. The p-value of the chi-square variatc was used to determine the statistical
significance of the model fit. Values greateuhep .05 indicate an acceptable fit of the
hypothesized model to the data.

The chi-square variatc is the basis for most of the goodness-of-fit indices in the
evaluation of a model. Therefore, it is important to note that the computation of the chi-square
statistic is based in part on the degrees of freedom in the structural equation. In SEM, the
degrees of freedom (df) are the number of free parameters subtracted from the number of

elements on the covariance matrix (variance and covariance terms, ie., (p(p+ l)/2). In the

92

initial model, the number of free parameters was 5 (regression on behavior, regression on
intention, error term for behavior, error term for intention, and the variance of attitude).

There were 6 variance and covariance terms (3(4)/2). The df was thus equal to 1 (ie. 6 minus

5).

RESULTS
Introduction
The nature of the research was exploratory. Thus, one purpose of the present study

was to describe the sample of lesbian women, their characteristics, and their desires,
intentions, attitudes, and behaviors vis-a-vis parenthood. A further purpose was to build a
model as a framework for discussing and understanding lesbian women's decisions to parent
or to not parent. The goal was not to confirm the model per se but to describe the
relationships between the factors in parenting intention and behavior. The components of the

predictive model examined were attitude, intention, and behavior.

Descriptive Data from the Survey
8 ' n-

Demographic information about the sample was provided in the methodology section.
Lesbian-related characteristics of the sample are given in Tables 6-8. The first survey
question asked the participants to indicated their sexual identity, with the use of a scale with a
continuum from 1 to 7, with 1 being completely identified as a lesbian, 4 being identified as a
bisexual, and 7 being completely identified as a heterosexual. As shown in Table 6, 91.9% of
the women in the sample indicated that they identified as a "1" or a "2", 5% identified as a
"3" on the sexual identity scale, and 3.1% identified as bisexual. Two respondents did not
answer this question.

The respondents who reported their identity as a "4" and those who did not answer the
question were kept in the sample for several reasons. It is expected that in none of these cases
did the respondent confuse the intended population of the survey: "lesbian" was in large bold

letters across the front of the survey, the entire introduction used the word "lesbian" no less

93

94

than nine times, and the first 17 questions in the survey pertained to disclosure of sexual
identity, lesbian publications, lesbian events, close lesbian friends, paperwork with a lesbian
partner, and so forth. It is likely that if these questions did not seem applicable (ie. if the
respondent was not a lesbian), the respondent would not be motivated to answer, nor does it
seem likely that anyone would take the time to answer a 20 minute survey that did not appear
applicable to them. Thus, the respondents most likely felt this "lesbian survey" was
appropriate to them.

Secondly, it is suspected that for most of these respondents, this question was a
philosophical or theoretical one, and some data confirms this: 3 of these women reported that
they were in monogamous relationships with women for 4, 6, and 11 years, respectively, and
lived with their partners. They are, in effect, living lesbian lives, even if they are theoretically
bisexual. This is not to argue against their self-perceptions, but rather to suggest that data
from these women were not inappropriately kept: they look like lesbians. For all practical
purposes, living monogamously with a lesbian paMer for 11 years would qualify as a lesbian.
For these reasons, it was deemed safe to include these women in the sample with the
assumption that they are lesbian women.

Other lesbian-related characteristics are shown on Table 7. Participants were asked
how often they attended lesbian/gay (LG) bars, LG social events, and LG political events. The
majority (57.1%) of respondents said they rarely or never attended LG bars, and 35.6% said
they occasionally attended. The majority (79.2%) said they frequently or occasionally attended
LG social events. The majority (57.1 %) reported that they frequently or occasionally attended
LG political events.

The survey also probed the degree of disclosure of lesbian identity to various groups

of people. As shown in Table 8, the majority of respondents (71.6%) reported that most or all

95

of their family knew of their sexual orientation, 90.2% said all of their LG friends knew, and
66.9% said most or all of their heterosexual friends knew their orientation. The respondents
were less forthcoming with co-workers: 49.4% said none or a few of their co-workers knew,
and 36.9% said that most or all of their co-workers knew of their sexual orientation.

In summary, most of the women in the sample were engaged in lesbian/gay social and
political events, but were less likely to attend LG bars. The large proportion of respondents

also reported "being out" to many people in their lives.

Table 6.
Lesbian Related Characteristics: Lesbian Identity

1 Lesbian Identity ‘ —m

Lesbian identity on a scale of l to 7

    

   
     
 

l = lesbian identity 101
2 47 29.2
3 8 5.0

 
   

 

 

   

4 = bisexual identity

96
Table 7.

Lesbian Related Characteristics: Attendance at Lesbian Activities

How often do respondents
attend these activities

N
percent

 

Lesbian/gay social events
N
percent

 

Lesbian/gay political events
N

 

 

 

 

 

Table 8.

Lesbian Related Characteristics: Disclosure of Lesbian Orientation

How many people in these
groups know of respondents
sexual orientation (N / %)
Family

N
percent

Lesbian and Gay friends
N
percent

 

 

" Heterosexual friends
‘ N

percent

 

Co-workers
N
percent

 

 

 

 

 

 

. 97

Prein-Rl h ctrii

Table 9 shows the parenting-related characteristics of the sample. Respondents were
asked if they matted children, as well as whether they intended to have children. Nearly half
(46%) of the sample said they did not desire or most of the time they did not desire to have
children. About one in five (20.2%) said they were not sure if they desired children, and
33.7% said they desired or most of the time they desired to have children.

People were more conservative, however, in terms of intentions to parent. Over half
(53.4%) of the women in the sample said they did not intend or most of the time did not
intend to have children. About one in five (20.2%) said they were not sure if they intended to
have children, and 26.4% said they intended or most of the time they intended to have
children.

Of those who intended to have children or who were unsure of their intentions
(N=73), about three in ten (30.1%) said they wanted one child, 53.4% wanted two children,
and 16.4% wanted three to five children. The mean number desired was 1.9, the median and
mode were 2 children, and the standard deviation was .81. Three women did not answer this
question.

In the survey, women were asked how often they thought about their intentions
(intentions could be to have, to not have children, or undecided) regarding parenthood. Nearly
four in ten (37.4%) women said they thought about their intentions very frequently or
frequently, 36.2% thought about their intentions occasionally, and 26.4% thought about their
intentions rarely or never.

In another question, participants were asked if they agreed with the statement, "there
is so much to think about and plan for before having children." Almost every participant

either strongly agreed (59%) or agreed (37.9%) with this statement. Participants were also

98

asked if they agreed that "I have thought a lot about whether or not I want to have a child. "
The majority either strongly agreed (32.5%) or agreed (35 .6%), 19.6% neither agreed nor
disagree, and 12.3% disagreed with the statement. In addition, as discussed below, many
women talk about the idea of parenting with other people. Thus, it appears that the women in
this sample perceive that having children is a planful undertaking, and a large portion of the

women personally have thought about and/or currently think about their intentions.

99
Table 9.

Parenting Related Characteristics

(Characteristic

Desire to parent
Not desire
Mostly not desire
Not sure of desire
Mostly desire
Yes desire

 

Intention to parent
Not intend
Mostly not intend
Not sure of intention
Mostly intend
Yes intend

 

Frequency of thinking about intentions
Very frequently
Frequently
Occasionally
Rarely
Never

 

», Number of children desired

 

 

 

100

Women were asked if they or their partner had done any activities or discussions
related to becoming a parent in the last 2 years, and if yes, what methods they had seriously
considered and what activities they had done. Of the entire sample, 64.9% reported that they
had done some activity or considered some method related to becoming a parent, 57.7% had
done 1 or more activities, and 46% reported 1 or more methods seriously considered to
become a parent. Table 10 shows the proportion of respondents who said they seriously
considered a method to becoming a parent. Respondents often did more than one behavior, so
the percentages for each are not intended to add to 100%. About one-quarter of the sample
had seriously considered each of the following: pregnancy from an anonymous donor (28.8%),
pregnancy from a known sperm donor (29.4%), or adoption (23.9%). Less common were
those who seriously considered pregnancy from sperm of a partner's male relative (12.9%) or
foster parenting (12.3 %), and almost none had considered engaging a "surrogate" mother
(1 .2%).

Table 11 shows the activities engaged in related to becoming a parent. The most
common activities were communicative ones: about half (50.3%) of the women in the sample
said they discussed ideas with a partner, 39.9% talked with lesbian mothers, 37.4% talked
with lesbians who want to be mothers, and 17.8% discussed ideas with potential sperm
donors.

Some activities related to becoming pregnant: 12.9% of the women in the sample
reported that they had charted their ovulation cycles, 9.8% had contacted sperm banks, 6.1%
had inseminated under a doctor's supervision, 3.7% used ovulation regulating drugs, and
1.8% had inseminated at home. No one in the sample reported trying to adopt or applying for
foster parent status. Three women (1.8%), however, had contacted an agency regarding a

home study, and 0.6% had a home study completed.

101

More than one in ten (11.7%) reported engaging in one or more "other activity".
These other activities were open-ended responses. Three of these other activities were deciding
or discussing NOT having children, four concerned potential guardianship or adoption of
relatives, three concerned reading material, and 6 concerned discussions with various people.
The remainder concerned doing donor insemination, becoming pregnant, preparing for
pregnancy, saving money for invitro insemination, seeing an infertility doctor, and dealing
with such family matters as an attorney.

It is important to note here that engaging in communication with others regarding
parenting is not always done with the positiveness of parenting in mind. As the "other
activities" reveal, some of these talks were confirmation of the idea of NOT parenting. It
cannot be known from the closed-ended questions regarding communication with others how
often this was a confirmation of a decision to NOT parent. The rest of the activities referred
to in the closed-ended questions, however, are clearly those which would lead to parenthood,
and thus would be undertaken only by those seriously interested in pursuing this role for
themselves.

In summary, about one-third of the women in the sample said they wanted children,
but about one-quarter said they intended to have children. About half of the women said they
did not want to and did not intend to become parents. About one in five women were not sure
of their desires and intentions. Nearly four in ten women reported thinking about their
intentions vis-a-vis parenting very frequently or frequently.

Nearly six in ten women in the sample had done some activity (including discussions)
leading to parenthood, and nearly half reported seriously considering at least one method to
becoming a parent. The most common activities were communicative ones, but nearly 10%

had contacted sperm banks, 6.1% had inseminated under a doctor's supervision, and 1.8% had

102

inseminated at home. Nearly 30% of the women said they have seriously considered getting

pregnant, and nearly one—quarter said they seriously considered adoption, in the last two years.

Table 10.

Methods of Parenting Seriously Considered

: Methods SeriOuSIy Considered by Percent
1 Respondent or Partner _of Cases

 

 

 

 

 

l Pregnancy from a known sperm donor 29. 4

' Pregnancy from an anonymous sperm donor 47 28.8 I
Adoption 39 23.9 ‘
Pregnancy from sperm of male relative 21 12.9 ll
Foster parenting . 20 12.3

i. Discussed ideas with potential "surrogate"

 

2

 

1.2

103
Table l 1.

Activities Engaged in Related to Becoming a Parent

Activities Engaged in During Past 2 Years N Percent
of Cases

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Discussed ideas with partner 82 50.3
Talked with lesbians who are mothers 65 39.9
I Talked with lesbians who want to be 61 37.4
l mothers
Discussed ideas with potential sperm donor 29 17.8
Charted ovulation cycles 21 12.9
Contacted sperm bank(s) 16 9.8
1 Inseminated under a doctor's supervision 10 6.1
Used ovulation regulating drugs . 6 3.7
! Inseminated at home 3 1.8
I Contacted an agency regarding a home 3 1.8
’ study
‘ Discussed ideas with potential" surrogate" 2 1.2
E Had a home study completed 1 0.6
' Applied for Foster parent status 0 0.0
Tried to adopt 0 0.0
Other activity 19 11.7

 

 

104

.n' 'v Null- ‘ P 'u 'n "HUI“ _ie I: m it .. 0. ,r'.,n '

Respondents were asked several questions that related to their perceptions of people's
assumptions about lesbians and the way the law treats lesbian mothers. Responses to these
questions are shown in Table 12. The idea of "mother" and "lesbian" as conflicting and
mutually exclusive roles is the underlying idea of these questions. In addition, the literature
suggests that the societal disapproval of lesbians in general and lesbian parenting in particular,
formalized in legal treatment of lesbian mothers, may influence the ways in which lesbians
approach the possibility of parenting.

When asked if they agreed with "Most people assume lesbians do not become
mothers," about eight in ten either strongly agreed (15.3%), or agreed (66.3%). About one in
ten neither agreed nor disagreed (10.4%), 7.4% disagreed and 0.6% strongly disagreed.
When asked whether most lesbians assume lesbians do not become mothers, about one in five
either agreed (18.4%) or strongly agreed (0.6%), about one-quarter (25.8%) remained neutral,
and over half of respondents either disagreed (43.6%) or strongly disagreed (11.7%).

Respondents were asked whether they agreed or disagreed with the following
statement: "The way the law treats lesbian mothers is a factor in my decision to parent OR to
not parent." The responses were as follows: strongly agree, 7.4%; agree, 18.4%; neither
agree nor disagree, 22.1%; disagree, 33.1%; and strongly disagree, 19%. Thus, although over
half did not agree with this statement, over one-quarter (25.8%) agreed or strongly agreed
with this statement.

Further analysis, however, revealed that lesbian women who were unsure of their
desire to have children were significantly more likely to agree with the statement about legal
treatment of lesbian mothers than were lesbian women who wanted children and lesbian

women who did not want children (X2 (8, N = 163) = 17.4, p = .03). Of the women who

105

were not sure of their desires to parent, 45.5% agreed or strongly agreed that the legal
treatment of lesbian women was a factor in their decisions. About one in five lesbian women
who wanted children (21.8%) agreed or strongly agreed with this statement, as did 20% of
lesbian women who did not want to have children.

Thus, about 80% of the respondents supported the perception that people in general do
not see lesbian women as mothers, and about 20% perceived that lesbian women assume
lesbians are not mothers, The results also imply that the (poor) legal treatment of lesbian
mothers is an influential factor for a portion of lesbian women, particularly those who are

uncertain of their desires.

Table 12.

Subjective N orms: Perceptions of Societal Assumptions
Strongly Agree Neither Disagree Strongly
agree % % % disagree
% .
66.3 10.4 7.4 .

' Most lesbians assume 0.6 18.4 25.8 43.6 11.7
lesbians do not become
I mothers

  
       
 
   
     

Questions regarding society

   
      
   

 
  
 

 

f Most people assume
f lesbians do not become
, mothers

 

 

.
___ _._ _____.-__ ______ z.__,._.________4

 

 
 

; The way the law treats
: lesbian mothers is a factor

' in my decision to parent or
i to not parent

 

 

 

 

 

106

C nc rn o s i P re i

There were several questions in the survey that specifically addressed potential
problems with lesbian parenting. These questions and the responses are shown in Table 13.
One statement was "It would be a problem having a child 'outing' me." About six in ten
respondents either strongly disagreed (30.8 %) or disagreed (31.4%), and 26.4 neither agreed
nor disagreed with the statement. Another statement in the survey was "I worry that my
children could be harassed for having a lesbian mother. " Nearly three-quarters of respondents
either strongly agreed (20%) or agreed (53.1%), and 16.3% neither agreed nor disagreed with
the statement. Finally, women were asked about this statement: "I would be very concerned
about a sperm donor's legal rights to a child. " About two-thirds of the respondents either
strongly agreed (29%) or agreed (38.3%), 17.9%‘neither agreed nor disagreed, and 11.7%
disagreed with the statement. Thus, it appears that about 10% of women were concerned about
children disclosing the parents lesbian orientation, but most were concerned about possible

harassment of their children, and about sperm donors' legal rights to children.

107
Table 13.

     
 

  

   

      
      
       

Questions regarding lesbian
parenting

  

 
  

Concerns About lesbian Parenting
Strongly Agree Neither Disagree
agree % % %
% _, _ _ ._ — __.__ .___ _____ ,_
It would be a problem 5.0 6.3 26.4 31.4
having a child "outing" me

I worry that my children 20.0 53.1 16.3 6.9
could be harassed for
having a lesbian mother

I would be very concerned 29.0 38.3 17.9 11.7
about a sperm donor's legal
: rights to a child

 

 

 

 

 

 

 

en' wi

There were questions regarding the recognition of the co-parent presented to the
respondents in the survey. These questions and the responses are shown in Table 14. In these
questions, respondents revealed that recognition for the co-parent (the women who is not
legally or biologically the "parent") is important, yet elusive. One statement was whether
"When a lesbian couple has a child, it is difficult for the non-biological co—parent to be
recognized as a full parent. " About 7 in ten women either strongly agreed (24.5%) or agreed I
(47.2%), 17.8% neither agreed nor disagreed, and 10.4% disagreed or strongly disagreed with
this statement. "Arranging for legal recognition of a co—parent is important if a lesbian couple
has children." was the other statement to which participants responded. Nearly all respondents
said they strongly agreed (61%) or agreed (28.3%) with this statement. less than one in ten
(7.5%) neither agreed nor disagreed, and only 3.1% disagreed or strongly disagreed with this
statement.

Other questions dealing with the relationship with one's pmer vis-a-vis parenting

108

were whether having a child was an experience the respondents wanted to share with a
partner, and whether having a child would increase the bond with their partner. About six in
ten women either strongly agreed (32.9%) or agreed (28 %) that having a child was an
experience they wanted to share with a partner, 18.6% neither agreed nor disagreed, and
about one in five said they either disagreed (13.7%) or strongly disagreed (6.8%). The women
were divided as to whether parenting would be a bonding experience with their partners.
About three in ten women either strongly agreed (8.4%) or agreed (21 .4%), about one-third
neither agreed nor disagreed, and more than one-third either disagreed (22.9%) or strongly
disagreed (13.7 %) that parenting would increase the bond with their partners. Thus, more
than half the women expressed that it was desirable to share parenting with a partner, but only

half as many thought it would be a bonding experience with a partner.

109
Table 14.

Parenting with a Partner

Questions regarding Strongly Agree Neither Disagree Strongly
parenting with a partner agree % % % disagree
%
47.2 17.8 6.1 .

It is difficult for the
nonbiological co-parent to
be recognized as a full
parent

 

Arranging for legal

recognition of a co-parent is
important if a lesbian
couple has children

 

Having a child is an
experience I want to share
. with a partner

 

Having a child would

lncrease the bond with my
artne

 

 

 

 

 

 

109
Table 14.

Parenting with a Partner

Questions regarding Strongly _—Agree Neither
parenting with a partner agree
24.5

It IS difficult for the
nonbiological co—parent to
be recognized as a full
parent

 

Arranging for legal
recognition of a co-parent is

, important if a lesbian
couple has children

 

Having a child is an
. experience I want to share
with a partner

 

. Having a child would
increase the bond with my

 

 

 

 

 

 

110

' ctiv N rms: ive Pe 1

Respondents were asked four questions polling the invariance of support for their
decisions - decisions to parent OR decisions to not parent - from friends, family, the lesbian
community, and from their partners. About half of the women (50.3%) agreed or strongly
agreed that the support of friends was important in their decision to parent OR to not parent.
Nearly half (46.6%) said the same about their families. But the respondents were more
equivocal in regards to the lesbian community; about one-third (35 .6%) agreed it was
important to have the support of the lesbian community, and another third (36.2 %) said they
neither agreed nor disagreed with the statement. Almost all the respondents either strongly
agreed (62.8%) or agreed (28.7 %) that the support of their paMer was important in the
decision to parent or to not parent.

Respondents were asked if they thought that "Lesbian mothers get support from the
lesbian community." About four in ten women agreed (38.9%) or strongly agreed (0.6%),
about one-third neither agreed nor disagreed (34.6%), and about one-quarter either strongly
disagreed (3.1%) or disagreed (22.8%) with this statement. The majority of respondents
thought that their lesbian friends would support their decision to have or to not have children:
about three-quarters of respondents either strongly agreed (37.3%) or agreed (37.9%), 21.1%
neither agreed nor disagreed, and 3.7% either disagreed or strongly disagreed that they would
have the support of their lesbian friends in their decisions. Thus, respondents expected support
from their lesbian friends for the respondents' decisions, but were slightly more reserved in
their assessment of support of the community for lesbian mothers in general.

Respondents were asked how much support they would expect from various people if
they became a parent. A summary of these results are shown on Table 15. There is a lot of

missing data in this set of questions due to respondents choice of "not applicable". The

111

response categories were 1 = very supportive, 2 = somewhat supportive, 3 = neither
supportive nor unsupportive, 4 = somewhat unsupportive, 5 = very unsupportivc. As you can
see from the table, the respondents were rather hopeful in their expectations from some
people. The median response was "very supportive" for lesbian friends with children, sisters,
and partners. Other people who were expected to be somewhat supportive if the respondent
became a parent were heterosexual friends with children, mothers, the lesbian community, and
heterosexual friends without children. Heterosexual friends without children, brothers, fathers,
and partner's family all averaged between somewhat supportive and neutral. Respondents
expected the least amount of support from co-workers and "other relatives" .

Thus, the overwhelming majority said the support of their partner was important in
their decision to parent or to not parent, about half of the respondents said the support of
friends and family was important, and about one-third said the same about the lesbian
community. Respondents also expected a high level of support from lesbian friends with
children, sisters, and partners should they become parents. Nearly 40% of the women said the

lesbian community supports lesbian mothers, but about one-quarter said the opposite.

112
Table 15.

Subjective Norms: Expected Support From Others

 
 

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Person
lesbian friends with children 151
Sisters 110
Partner 115 1.0 1.7 1.3
Straight friends with children 149 2.0 2.0 0.9
Mother 137 2.0 2.0 1.2
' Lesbian community 154 2.0 2.2 0.9
lesbian friends without children 158 2.0 2.2 1.0
Straight friends without children 152 2.0 2.4 1.0
Brothers 1 12 2.0 2.4 1.1
Father 116 2.0 2.4 1.4
Partner's family
Other relatives
_ Co—workers

 

113
Norm : ' ti n wi b' M r
Respondents were asked three questions regarding their affiliation with lesbian women
who had or desired children. As shown in Table 16, the majority of respondents said they had
close lesbian friends with children (69.8%) and had close lesbian friends who were thinking
about having children (65.6%). More than one-quarter (27.3%) of respondents said that they
had close lesbian friends who were actively trying to have children. Thus, affiliation with

lesbian women who had children or who desired children was common for the women in this

sample.

Table 16.

Norms: Affiliation with Lesbian Mothers and Desirers

Affiliation Questions N who said Percent
"Yes" not: Cases

Do you have any close lesbian friends who 113 69. 8
' have children?

 

Do you have any close lesbian friends who 107

‘ are thinking about having children?

 

 

 

1
Do you have any close lesbian friends 44 27.3 i
actively trying to have children? j

1 14
Helpful [neerventiens

Respondents were asked to look at a list of activities that could be helpful to people in
drinking about whether they want to have children, or to help people feel comfortable with
their choices. Respondents were asked to indicate how helpful each of these activities would
be for them, on a scale of "helpful", "not sure", and "not helpful." Table 17 shows the
proportion of people who said an activity would be "helpful" to them. As seen in the table,
getting support was deemed helpful by the majority of respondents. Most respondents reported
that support groups or discussion groups for lesbian mothers (73%), ideas on how to get
support from the community for lesbian mothers (72.4%), and practical information on "how
to" become a parent (71.8%) would be helpful. In addition, 62.6% said it would be helpful to
get ideas on dealing with one's new lover's children.

Thus, the need for support and information related to parenting was clear. In addition,
women want help for an area that did not relate to choosing parenthood per se, but to being
involved with a parent, and the possibility for becoming a "stepmom. " In addition, about six
in ten women said that a discussion group for lesbians who are undecided about becoming
parents would be helpful. And about half said it would be helpful to know how to talk to one's

partner about parenting desires and to clarify values about parenting.

115
Table 17.

Helpful Interventions in Thinking About Parenthood

Interventions that would be "Helpful" N who said Percent
Total N=163 _"Helpful" of Cases

   
 
 

  

Support or discussion group for lesbian
mothers

73.0

 
   

 

 

 

 

 

 

 

 

 

 

 

 

Ideas about getting community support for 118 72.4
lesbian mothers
Practical information about "How to" 117 71.8
become a parent
Ideas to deal with your new lover's children 102 62.6
I Discussion group for lesbians who aren't 97 59.5
sure whether they want to become mothers
Ideas about how to talk with one 's partner 80 49.1
about parenting desires
|
' Help with clarifying values about parenting 77 47.2
l Discussion group for lesbians who 30 18.4
( definitely do not want to become mothers
' Ideas about getting community support for 22 13.5

lesbian women who do not want children

116

Structural Equation Model Results
WM

Recall that the theory of reasoned action (TORA) was used as a framework for
exploring the relationships among variables in the decision to parent. In this framework,
subjective norms and attitude are expected to predict intention, and intention is expected to
predict behavior. Because this research is an exploratory step, confirmation of the TORA
model is not goal here. The TORA model was used instead as a tool to examine the constructs
believed to be relevant to the decision to parent in a comprehensive yet parsimonious manner.
The TORA has been used successfully in previous fertility decision research.

The set of relations described here is displayed in Figures 1 and 2. Figure 1 shows the
original conceptual model and Figure 2 shows the revised conceptual model. Figure 1 was not
a testable model due to the limitations of the scales, as discussed in the methodology chapter.
Figure 2 was the testable model most closely approximating the original conceptual model
depicted in Figure l. The difference between these two models is that in Figure 1, there are a
series of attitude toward parenting subscales that are hypothesized to predict intention to
parent, whereas in Figure 2, there is a single attitude toward parenting scale (which consists of
items that made up 4 attitude subscales in Figure 1) hypothesized to predict intention to
parent.

Figure 2 depicts the series of relationships among concepts that are operationalized in
this empirical study. As Figure 2 shows, attitude toward parenting is expected to have a
relationship to intention to parent; it predicts the intention to parent. Similarly, the normative
variable (expected support) is expected to predict intention. The intention to parent is
hypothesized to predict behavior. All these relationships are linear. Notice also that the arrows

go only in one direction; the variable with the arrow pointed toward it is predicted by the

117

variable from which the arrow originates.

It is further hypothesized in this model that attitude toward parenting does not have a
direct relationship to behavior, however, there is an indirect effect of attitude on behavior
through intention. Further, no relationship is specified from behavior to intention, nor from
intention to norms, nor from intention to attitude; it is hypothesized that these relationships do
not exist.

The strength of the relationships cannot be specified in the conceptual model. It is. the
hypothesized relationships of the conceptual model, along with the observed data, that allow
the strength of the relationships to be estimated by the structural equation. Further, the
direction of the relationships cannot be established empirically, these are conceptual issues
which should be reflected by the theoretical constructs in the model. That is, which variables
are predictors, which variables are dependent, or whether the relationship is causal cannot be
determined by the data itself. There are no causal relationships specified in this model. The
design of the study and the level of knowledge of the phenomena under study prohibit any

suggestion of causality.

118

Ewan. 2 mco_m_ooo m

cmE0>> Cosmo;

 

119

E99”. op mco_m_omn_ m.cmE0>> 55mm

 

 

120

 

Error

 

Lesbian Women’s Decisions to Parent

N=163

12 1
The Statistical Model

In Figure 3, the statistical model is shown. This is a direct extension of the revised
conceptual model in Figure 2, but it includes the results of the analysis. This figure is a path
diagram to communicate the SEM at the statistical level. All the relationships hypothesized to
exist in the conceptual model were supported by the analysis with the exception of the norm
variable of expected support. Expected support is not shown in Figure 3 because the T-values
in the analysis indicated that this variable did not fit the data, and did not belong in the model.
As discussed previously, the expected support variable is a scale of the average amount of
support expected (for becoming a parent) from various people in the community, friends, and
family. The attitude toward parenting variable is a scale made up of the mean of 13 attitudinal
items from the survey. The intention to parent variable is a scale made up of the intentions
and desires to parent items from the survey. The behavior variable is a count of all the
activities engaged in that could lead to parenthood.

The statistical model shown in Figure 3 is made up of several components. There are
several free parameters, ie. parameters to be estimated by the structural equation model. There
are 5 free parameters: the regression weights for behavior and for intention, the error variance
terms for behavior and for intention, and the variance of attitude. There are also several fixed
parameters in this model, ie. , those parameters which specify that there is not a relationship
between variables. In this model, the fixed parameters are attitude to behavior (indirect
relationship only), behavior to intention, and intention to attitude.

The numbers on the far right side of the diagram, with arrows going to behavior and
to intention are the error variance terms. The error estimates are derived from the parameter
estimates and the discrepancy function. The number next to the arrow pointing to intention is

the regression weight of attitude on intention. The number next to the arrow pointing to

I22

behavior is the regression weight of intention on behavior. These regression weights are
standardized weights, they indicate the amount of change in the dependent (predicted) variable
for a given change in the predictor. These are both in standard deviation metric.

In a SEM such as this, the squared multiple correlations (R2 ) for the structural
equations are provided. The R2 is a regression statistic that indicates the amount of variance in
the dependent (predicted) variable that is accounted for by the components in the equation, ie. ,
the predictors.

f ' el

Table 18 shows the data used to create the structural equation model. The table
includes the correlation matrix as well as univariate information: means, standard deviations,
kurtosis, and skewness. As shown on the table, skewness and kurtosis values indicate that the
data do not substantially depart from normality. Thus, no corrective procedures were
necessary, and the Maximum Likelihood (ML) estimator was appropriate for use with this
data.

The results of the analysis (parameter estimates, squared multiple correlations, and
overall fit indices) are summarized in Table 19, the implied correlation matrix is shown in
Table 20, and the standardized parameter estimates are diagrammed in Figure 3. The
standardized regression coefficient for intention is .87, meaning that for every one standard
deviation increase in attitude, there is a .87 standard deviation increase in intention. The
standardized regression coefficient for behavior is .62, meaning that for every one standard
deviation increase in intention, there is a .62 standard deviation increase in behavior.

As shown on Table 19, the R2 for behavior is .39. Thus, the (predictor) intention
accounts for 39% of the variance in behavior. The R2 for intention is .76. Thus, the

(predictor) attitude accounts for 76% of the variance in the dependent variable intention. The

123

error variance for behavior is .61 and the error term for intention is .24.

The model fit indices shown in Table 19 reveal that the model is a close fit to the
data. The chi-square value is not statistically significant, indicating the model is a good fit;
(7L2 (1, N = 163) = 0.29,p = 0.59). The GFI = 1.00, the NNFI = 1.01, and the CF] =
1.00, which all indicate a close fit of the model to the data.

As shown in Table 20, the correlation between attitude toward parenting and behaviors
leading to parenthood is .55. This is the indirect effect of attitude on behavior. The direct
relationship between attitude and behavior was tested in the model by re-running the analysis
with a path specifying prediction of behavior from attitude toward parenting. The T-value in
the output showed that this path did not fit the model. Thus, the SEM indicated that attitude
toward parenting was not a sufficient predictor of behaviors leading to parenthood, as would

be expected from the TORA.

124
Table 18.

Observed Data for Structural Equation Model # 1

Correlation Behavior Attitude Expected
wsuppolt"

 

 

 

 

l Behavior
Intention .625 1 .0
Attitude .561 .873 1 .0
Expected .171 .291 .341 1.0
support
Behavior Attitude mExpected
Wimp—port

 

Range
SD
‘ Skew

 

 

 

 

 

 

 

 

Kurtosis

125
Table 19.

Standardized Results of Model # 1

PARAMETER ESTIMATES:

Behavior = .62 * intention, error variance = .61
Intention = .87 * attitude, error variance = .24
Variance of attitude = 1.00

 

SQUARED MULTIPLE CORRELATIONS
FOR STRUCTURAL EQUATIONS:

Behavior R 2 = .
Intention R 2 = .76

 

OVERALL MODEL FIT INDICES:

Chi-square (1, N = 163) = 0.29, p = 0.59

GFI =1.00
NNFI = 1.01
CFI = 1.00

Table 20.

Correlation Matrix Implied by the Structural Equation Model # 1

 

 

 

 

l

126
An le iv M I

As discussed in the methodology chapter, efforts were made to modify the SEM.
Modifications were made in an attempt to tap into theoretical constructs which may have been
inadequately measured in the first model. The normative component of the model - expected
support from others - did not emerge as a predictor in the first model. Thus, other normative
variables were considered, including several support measures, and perceived societal norms
regarding the appropriateness of lesbian parenting.

Other modifications were made because the initially hypothesized attitude components
of the model (ie., creating a family, motherhood identity, nurturance toward children, and
material reasons) turned out to be subscales of a more general attitude toward parenting
construct. Thus, length of time with one's monogamous parma was an attempt to tap into the
”creating a family" construct. Finally, demographic variables considered to be salient vis-a-vis
parenting - age, and number of siblings - were also used in attempts to modify the first
model.

The variables considered as modifications to the first model are shown in Table 21.
The variables used to modify the model, the reasons for use of the variables, and whether the
modifications were validated by the output are all shown in the table. Modifications to the
model were at first tested one variable at a time. Variables were empirically shown to belong
or not belong in the model, based on the modification indices and the T-values for each
parameter. After all the variables had been entered individually, the ones that appeared to fit
the data were entered into the equation simultaneously, and these variables together still
appeared to fit the data.

Modifications to the model were undertaken with the assumption that the new

variables entered into the equation would be predictors of intention, just as attitude was a

127

predictor of intention. Modification indices for length of time in a monogamous relationship,
and affiliation with lesbian mothers, however, revealed that these variables were better
predictors of behavior than intentions. As Table 21 also shows, age was found to be a
predictor of intention to parent.

The correlations between the variables used in the attempt to modify the first model
and the dependent variables (intention and behavior) are shown on Table 22. As the table
shows, there are several scales that are moderately correlated with both behavior and
intention. A high correlation alone is not an indication that a variable may be an adequate
predictor of a dependent variable, because the LISREL path models are analyses of
covariances rather than individual cases (ie., both the correlation and the standard deviation
are considered in the computations).

In examining the correlations shown in Table 22, one can see that expected support
from friends is correlated with both behavior (r = .23) and intention (r = .34); expected
support from lesbian sources is correlated with both behavior (r = .20) and intention
(r = .23); and affiliation with lesbian mothers and desirers is correlated with both behavior
(r = .34) and intention (r = .23). Age is correlated with intention (r = -.34), and time in a

relationship is correlated with behavior (r = .23).

128
Table 21.

Variables Used in Attempted Modifications to the First Model

Variable Reason for Inclusion T-value and Modification
Index Information

Length of time in
monogamous relationship

Attitudinal indicator of sense
of family

T-value & MI indicated this
was a predictor of
behavior, but not intention

 

 

Age

Demographic thought to
relate to parenting choices

T-value & MI indicated this
was a predictor of
intention

 

Number of siblings

Demographic thought to
relate to parenting choices

T-value indicated this did
not predict intention

 

Affiliation with lesbian
mothers & desirers

Normative indicator of
support from others

T-value & MI indicated this
was a predictor of ,
behavior, but not intention

 

Expectations of support
from lesbian friends and
community

Normative indicatOr of
support from lesbian others

T—value indicated this did
not predict intention

 

Expected support from
family

Normative indicator of
support from family of
origin

T-value indicated this did
not predict intention

 

Expected support from
friends

Normative indicator of
support from friends

T-value indicated this did

 

Importance of support

Normative indicator related
to support from others

T-value indicated this did
not predict intention

 

Most people assume that
lesbians don't become
mothers

Normative indicator of
appropriateness of lesbian
parenting

T-value indicated this did
not predict intention

 

Most lesbians assume that
lesbians don't become
mothers

Normative indicator of
appropriateness of lesbian
parenting

T-value indicated this did

1
not predict intention
not predict intention ‘

 

Legal treatment affects
decisions

Normative indicator of
appropriateness of lesbian

 

parenting

 

T-value indicated this did
not predict intention

 

129
Table 22.

Correlations Between the Variables Used in Attempted Modifications to the Model

and the Dependent Variables

_...b i ___+__ 9' 'm ______

Length of time in monogamous relationship

 

Age

 

Number of siblings

 

Affiliation with lesbian mothers & desirers

 

Expectations of support from lesbian friends
and community

 

 

 

 

Most people assume that lesbians don't
become mothers

 

Most lesbians assume that lesbians don't
become mothers

 

 

 

 

Legal treamient affects decisions

130
le ivMe

The final alternative model included the three variables from the first model (attitude,
intention, and behavior), as well as three new variables: length of time with one's paMer,
affiliation with lesbians who are or who desire to be mothers, and age. Table 23 shows the
data used to create the alternative structural equation model. The table includes the correlation
matrix as well as univariate information: means, standard deviations, kurtosis, and skewness.

The results of the alternative model are shown in Tables 24 and 25, and the path
diagram is shown in Figure 4. The standardized regression coefficient for intention based on
attitude is .84, and the standardized regression coefficient for intention based on age is —. l 1.
The standardized regression coefficient for behavior based on intentions is .59. The
standardized regression coefficient for behavior based on length of time in a relationship is
.22, and the standardized regression coefficient for behavior based on affiliation with lesbian
mothers is .19.

As shown on Table 24, the R 2 for behavior is .48. Thus, the (predictor) intention
accounts for 48% of the variance in behavior. The R2 for intention is .77. Thus, the
(predictor) attitude accounts for 77% of the variance in the dependent variable intention. The
error variance for behavior is .52 and the error term for intention is .23.

The model fit indices shown in the table reveal that the model is a close fit to the data.
The chi-square value is not statistically significant, indicating the model is a good fit; (X2 (4,

n = 163) = 4.32, p = 0.36. The GFI = .99, the NNFI = 1.00, and the CE] = 1.00, which
all indicate a close fit of the model to the data.

As shown in Table 25, the correlation between attitude toward parenting and behaviors
leading to parenthood is .56. As with the first model, the direct relationship between attitude

and behavior was tested in the second model by re-running the analysis with a path specifying

I31

prediction of behavior from attitude toward parenting. T-values in the output showed that this
path did not fit the model. Thus, the SEM indicated that attitude toward parenting was not a
sufficient predictor of behaviors leading to parenthood, as expected.
ngparisgn 9f Mgflgls

In comparing alternative models, comparison of the overall goodness of fit indices is
not appropriate (Hoyle & Panter, 1995). A chi-square difference test, however, is appropriate.
With 3 df difference, the 4.03 difference in the chi-square values between the models was not
significant (it is smaller than the critical value at 3 df,'7.81). This indicated that Model ll 1
and Model # 2 were not significantly different. In short, empirically, these models were

equivalent.

 

Lesbian Women’s Decisions to Parent

N=1

133
Table 23.

Observed Data for Structural Equation Model # 2

 

 

 

 

 

 

 

 

 

 

 

1mm
Behavior 1.0
Intention .625 1 .0
Attitude .561 .873 1 .0
Time .236 -.018 .058 1.0 l
Affiliation .340 .204 .190 .151 1.0
Age - .144 - .369 - 306— .2
1 Behavior MI
Mean 1.95 3. 26 35. 69 1 .62 34.15
Range 0-10 1-5 1-5 0-192 0—3 1944
SD 2.236 1.415 .974 43.984 1.020 5.568 ll
Skew 1.193 .373 -.131 1.399 -.135 -.196 ll
"‘ Kurtosis 1.309

 

 

 

 

 

 

 

1 34
Table 24.

Standardized Results of Model # 2

_ _.-_____.__ .._—..._~_ ___ _____ _ ._—____

 

PARAMETER ESTIMATES:

Behavior = .59 * intention + .22 * time + .19 * affiliation,
error variance = .52

Intention = .84 * attitude - .11 * age,
error variance = .23

Variance of attitude = 1.00

 

 

CE] = .99
NNFI =1.00
1 CF] = 1.00

L

SQUARED MULTIPLE CORRELATIONS
FOR STRUCTURAL EQUATIONS:

Behavior R 2 = .48
Intention R 2 = .77

OVERALL MODEL FIT INDICES: .

Chi-square (4, N = 163) = 4.32, p = 0.36

 

Table 25.

Correlation Matrix Implied by the Structural Equation Model # 2

1

Behavior

 

Intention

 

Attitude

 

Time

 

 

 

 

DISCUSSION
Descriptive Analyses

The survey provided a great variety of information on the sample, the parenting
desires of the respondents, and the factors predicting intention to parent and behavior leading
to parenthood. A substantial proportion of the women in the sample had disclosed their lesbian
identity to their friends and family members, and participated in lesbian/gay social events and
political events. The sample consisted of women who had resources - jobs, high incomes, and
extensive education.

There was variety in parenting desires. About one-quarter of the women in the sample
reported that they intended to have children. One striking feature was the amount of people
who said they were not sure of their desires and intentions about becoming parents - about
one in five. Of the women over 44 years of age who were not included in the sample, more
than half were parents. Of the parents under 45 years of age who were not included in the
sample, about one-quarter wanted to have additional children. Of all the women under age 45
responding to the survey (those included and those not included in the sample), about one-
quarter were mothers. About one-quarter of the children of the women not included in the
sample were born/adopted after the women came out as lesbians. Thus, the findings in the
current study would appear to suggest that approximately 25% of lesbian women are mothers
or decide to be mothers. These proportions are more likely to be representative of white,
educated, lesbian women in their 30s in Michigan than they are of lesbian women in general.

These findings are not dissimilar to previous literature on the percentages of current
lesbian mothers: 29% in Ryan and Bradshaw, 20% in Zeidenstein, as well as the percentages
of lesbian women deciding to become parents: 15% in Pies' (1987b) groups, 23% in Ryan and

Bradshaw. Recall that 34% of the women in Bybee's (1990) study were interested in

135

136

parenting, and 60% of the women in Johnson et al.'s study (1987) said they "considered"
parenting. In terms of acting on one's desires, about 10% of the women in the current study
were trying to become pregnant. This is the same proportion as in Johnson's study, but is
higher than Bybee's study, which was 2%.

Comparisons between the current study and the previous studies are limited, however,
by considerations of the samples and the purposes of the studies. In the current study, only
non-mothers under age 45 were included in the sample. Other studies cited did not have such
restrictions. Furthermore, there could be differences due to the focus of studies - gathering
data on parenting desires was not central to the purposes of the previous studies. In addition, a
caveat regarding the sample in the current study must be given. The sample is neither random,
nor assumed to be representative of lesbian women in general. Further discussion of the
limitations of the sample follows in the section labeled "methodological limitations".

About four in ten women in the sample said they thought about their intentions
regarding parenthood frequently. Those who desired children tended to say they wanted 1-2
children (m = 1.9). The number of children desired is somewhat less than the commonly cited
norm of 2-3 children for people in the United States. Thus, the normative number of children
desired by lesbian women appears to be smaller than that for the general population.

Another point of interest was the proportion of women who had done some activity
that could lead to parenthood. Talking with others was one of the most common of activities
pursued. There was also some evidence that even for women who do not want to have
children, the possibility and/or the undesirability of parenting is discussed with others. And
nearly half of the women reported having seriously considered methods to becoming a parent.
Women in this sample were 3 times as likely to pursue artificial insemination under the

supervision of a physician than to pursue this method at home. And there were nearly equal

137

numbers of women considering artificial insemination with a known donor as with an
anonymous donor. For some time there has been discussion about the prevalence of these
activities but very little data. This descriptive data is a starting point for understanding the
choices in methods that lesbian women are making to become parents.

The picture that emerges from this data is that the possibility of parenting is a salient
topic to lesbian women. Many women in the sample were thinking about parenting, thinking
about how they could become a parent, talking about parenting, and taking steps to become
parents. This is in sharp contrast to the notion that lesbian women are not mothers and are not
interested in becoming mothers.

Parenting is an important topic to lesbian women beyond the ”decision to parent"
focus taken in this study. Lesbian women are affected by parenthood even if they are not
parents or do not desire to become parents. The data revealed that women want help talking
with pmers regarding parenting desires, they desire to clarify values about parenting, and
they want help dealing with the children of a new lover. This leads to other questions not
specifically addressed in the survey. These questions include to what degree motherhood
impacts one's dating, one's options in establishing a relationship with a new lover, and the
circle of friends with whom one associates.

The women tended to expect support from certain people in the event that they became
parents but they did not expect everyone in their lives to be supportive. Support from a
paMer, family, and friends was important, and to a lesser degree, from the lesbian
community as well. Less than half of respondents said that the lesbian community is
supportive of lesbian mothers and only one-third of respondents said that support from the
lesbian community is important to them. But ”supportive” was the median response in the

expected amount of support from the lesbian community. Thus, there is variability in the

138

importance of and expectations of support from the community. There is more to be learned
about the place of the lesbian community in lesbian women‘s lives and families.

There was an explicit assumption in the current research that lesbian women's
decisions to parent could be influenced by lesbian-specific concerns and by normative
expectations. This was based on literature suggesting that women had lesbian-specific
concerns related to donor rights, recognition of the nonbiological coparent, fears of
harassment of the child (for having lesbian parents), fears of being "outed" by a child,
changes in relationships with others, and overcoming the normative expectations of lesbian
women as non-mothers (See Corea, 1985; Frank & Brackley, 1989).

The findings of the current research indicate that lesbian-specific concerns are indeed
very real, as evidenced by the information provided by the women regarding the legal
treatment of lesbian mothers, donor rights, recognition of the nonbiological coparent, and
fears of harassment of the child (for having lesbian parents). In addition, there is feedback
about perceived normative messages. The majority of respondents reported that there is a
widespread belief that lesbian women are not mothers, and nearly one in five respondents said
that lesbians also believed that lesbians are not mothers.

Lesbian-specific concerns for parenting should not be dismissed: about one in five
respondents who wanted children and about one in five respondents who did not want to have
children agreed that the way the law treats lesbian women was a factor in their decision to
parent or to not parent. But this legal treatment was more salient for the lesbian women in the
sample who were not sure of their parenting desires and intentions. Nearly half of women who
were not sure of their desires and intentions agreed that legal treatment of lesbian parents was
influential in their parenting decisions.

Although concerns are clearly shown in the data, there is not strong evidence that the

139

parenting decisions of the sample are dictated by lesbian-specific concerns. The emergence of
an attitude toward parenting scale that was not lesbian-specific in content or tone suggests that
the fertility decisions of the sample are not necessarily driven by lesbian-specific concerns. In
conclusion, the current research suggests that there are lesbian-specific mama in parenting
which are noted and contemplated by lesbian women. Whether lesbian women's decisions to
parent are definitively influenced by these concerns is as yet unknown and cannot be

ascertained with the data gathered in this study.

Measurement Conclusions

The development of the attitude toward parenting scale was a central concern in this
research. As noted in earlier chapters, the initial research questions suggested that there would
be a series of attitude toward parenting subscales, each reflecting different aspects of
parenting. These aspects included material concerns, concerns with motherhood as an identity,
nurturance toward children, and creating a family. These subscales were so highly
intercorrelated, however, that one general attitude toward parenting scale was developed for
use in the predictive model. In addition, the highest intercorrelation in the attitude subscales
was between the motherhood identity subscale and the nurturance subscale. It appears that
feelings of nurturance toward children and perceiving oneself as a potential parent are not
separate, but go hand-in-hand. Thus, it was clear that there were not several independent
attitude toward parenting constructs being measured. Rather, there was one attitude toward
parenting construct being measured (albeit with items which included various aspects of
parenting). Thus, it appears that different aspects of parenting are not separate, but are
integrated into the way one conceptualizes parenting.

It is expected that the attitude toward parenting scale developed in this research may

140

be a worthwhile measure for future fertility decision research because of the internal
consistency and because it addresses various aspects of parenting. For example, the scale
includes items which may reflect the most salient reasons to parent (nurturance) and those
most salient for not parenting (material). In addition, the attitude toward parenting scale
includes items which may be salient for all women (identity as a mother, and creating a
family) but which have not received much attention in previous literature, and have been
brought to light by the considerations of how lesbian women may approach the idea of
parenting.

The attitude toward parenthood scale is striking in another sense -— it is not oriented
toward lesbian parenting per se, but toward parenting. The scale lacks lesbian-specific
features. Recall the items making up this scale (see Table 4): there are no items that
specifically relate to difficulties or concerns of lesbian parenting. There are also no items
relating to any lesbian-specific joys or positive challenges in parenting. Whether such lesbian-
positive items would have been a part of the attitude scale is unknown - due to the oversight
of the principal investigator in not including such items in the survey. This suggests that
motherhood may be perceived as plain motherhood, rather than being seen as "lesbian
motherhood. " This would be consistent with Lewin's (1993) findings that lesbian women were
focused on establishing claims to motherhood and affirming motherhood as a central identity,
rather than being focused on lesbian motherhood.

The development of a scale to measure perceived normative information was another
main focus of the study. The measure of norms initially developed - expected support from
various specific others if one became a parent -— was not an adequate predictor of intention to
parent. Therefore, post-hoe efforts were instituted to make use of other normative data from

the survey. Importance of support, affiliation with lesbian mothers, and several expected

141

support measures were all scales developed from the survey data as measures of normative
information. In addition, several individual items were examined as measures of normative
messages of the appropriateness of lesbian parenting - items regarding whether most people
assume that lesbians do not become mothers, whether most lesbian women assume that
lesbians do not become mothers, and whether the way the law treats lesbian mothers is a
factor in the decision to parent or to not parent.

One possible shortcoming of the measure of expectations of support as a normative
measure is that there is room for different interpretations of the responses. Expectations of
support may be an ambiguous way to measure norms regarding lesbian parenting -
expectations could be based on knowledge of the respondents' parenting desires. For example,
low levels of support for parenting from important people in one's life could be due to the
belief that the respondent would not choose to parent. In this case a low level of support could
be a reflection of one's desires, rather than a normative message regarding the appropriateness
of the respondent (or any lesbian woman) becoming a parent.

All these normative scales appeared to be adequate in terms of internal consistency.
There is a strong possibility, (see discussion on the model analyses below) however, that none
of the normative scales developed were really adequate for tapping the normative construct.
The norm scales developed in this study may be related to norms, but are not precisely
measuring norms. Measurement would likely be improved if it more directly measured
specific messages regarding the appropriateness of parenting. In addition, to get a fuller
picture of the normative component of the decision-making process, it may be necessary to
measure relevent features of norms - knowledge of referent groups, perceived messages of

norms, and motivation to comply with norms.

142
Model Analyses

In addition to describing the characteristics of the sample, and the parenting desires,
two (nested) structural equation models were created. While both models adequately fit the
data, the first SEM was deemed to better explain the phenomena of interest -- prediction of
intention and behavior leading to parenthood. In this first model, behavior was predicted from
intention, and intention was predicted from attitude toward parenting. The squared multiple
correlations were substantial for both dependent variables, indicating that a large proportion of
the variance was accounted for by the independent variables in the equation, ie., that these
factors explain the intention and behavior to a large extent.

In the second model, these same relationships held, but additional variables appeared
to be predictors of the dependent variables. In the second (post-hoc) model, behavior was
predicted from intention, length of time in a monogamous relationship, and affiliation with
lesbian mothers (or those who want to be mothers). Intention was predicted by age as well as
attitude toward parenting.

As noted, modification indices for length of time in a monogamous relationship and
affiliation with lesbian mothers revealed that these variables were better predictors of behavior
rather than intention. Upon reflection, there is some conceptual soundness to this empirical
finding. Affiliating with lesbian mothers and having a long-term paMership should not
increase intention to parent per se, but if intention is there, they may aid one to act upon
desires. When one affiliates with lesbian mothers or has a pmer, one can see the possibilities
and anticipate assistance in one's endeavors. Unlike the attitude variable, the time and
affiliation variables represent more tangible experiences with people. If these variables are
indeed indicators of a sense of family and support, respectively, then the findings make sense.

For example, lesbian mothers and lesbians who want to be parents can act as sources of

143

support, and the length of a partnership may be part of the perception of the couple as a
family.

Further, as discussed previously, young lesbian mothers not included in the sample
were more likely to have a parmw than the women in the sample. This suggests the
importance of parmership in fertility decisions and lends credence to the viability of length of
time with a parmer as a predictor of behaviors leading to parenthood.

The second model cannot be relied upon to better describe the data. Post-hoc
modifications allow the potential for spurious and idiosyncratic aspects of the sample to
influence the model fit, and may not be replicable in other samples. Further, there is some
room for debate as to what constructs these three additional variables truly represent. While it
was put forth that length of time represented a family variable, and that affiliation with lesbian
mothers represented a support variable, other possibilities exist. For example, others may act
as role models, support figures, pillars of stability, sources of information, and as family
members. In any of these roles, there is the opportunity to facilitate action.

In conclusion, the first model is better in explaining the data. The post-hoc model was
not statistically better than the first and post-hoe models are not preferred. In addition, the
first model is more parsimonious, and the utility of the TORA has been previously
documented in decision-making research and is derived from theory.

Recall that the TORA expects that an attitude component and a normative component
are the predictors of intention. None of the (1 initial nor the 7 post-hoc) normative measures,
however, were predictors of intention to parent. One possibility is that the TORA does not
adequately describe the relationship between norms and intentions, or the relationship between
norms and intentions for lesbian women or for the women in this sample. Other possibilities to

explain this outcome in terms of the limitations of the measurement were previously discussed.

144

Another possibility is that the weight of the normative and attitudinal components differ for
various subgroups or under different conditions, and that this component, if measured, would
have shed light on the current findings. Further questions arise —- are there conditions under
which importance of support would be weighty enough that it would figure into the decision to
parent? Would the importance of pmer support alone make a difference?

In summary, most of the main assumptions of the TORA appear to fit the data from
the current study -— behavior was predicted from intention, intention was predicted from
attitude, and intention was a mediating step between attitude and behavior. Subjective norm
variables, however, did not appear to be a predictor of intention, contrary to the expectations
of the TORA. Previous studies have shown that TORA was useful in predicting fertility
decisions. Because the current study was exploratory, however, confirming or proving a
model was not a main goal. It was a goal to discover whether TORA was a useful framework
for exploring multivariate relationships. And the results of the current study do indeed suggest
that the TORA framework is useful and informative for examining and describing
relationships between variables in lesbian women's decisions to become parents. Caution,
however, must be taken in interpreting these results because of the unknown representativeness
of the sample.

Finally, although the second model cannot be promoted as a better framework for
examining relationships among variables, it suggests areas for ,further exploration, such as the
role of pmership and affiliation with other people in influencing lesbian's fertility behaviors.

These may represent normative components of the decision to parent.

145

Methodological Limitations

As with any study in which respondents are self-selected, there are potential
methodological limitations and concerns which must be addressed to aid interpretation of the
data. First, over half of the surveys returned were from respondents on the LC mailing list.
To what degree the women who subscribe to LC are different from other lesbian women is
unknown. And the degree to which the women in the sample are representative of the lesbian
women of Michigan, or of the women on the LC mailing list is not known. In addition, recall
that lesbian mothers and desirers were specifically sought near the end of data collection, to
improve the distribution of the intention to parent variable. Nine women of the women in the
Lesbian Mothers Group responded to the survey, and may make up as much as 5.5% of the
sample data.

One question that stands out is how representative is the sample in terms of resources
- education, employment, income? The sample is very unlike the general population of
Michigan in these regards. But is the sample more representative of the women who subscribe
to LC, of lesbian women in Michigan, of the women who go to women's music concerts, or
of the lesbian women who want to be parents? It is likely that there is a tendency for lesbian
women to have higher education than their heterosexual counterparts, and that those with
higher education tended to be interested in filling out the survey. Indeed, there may have been
a sympathy factor: people sometimes asked ”what is this research for? " and were very
responsive when the answer came, ”this is my dissertation research. "

Clearly we have reason to think the sample is got representative of lesbian women -
because there is no data available to describe ”typical" lesbian women, and because of the
high amount of resources of the respondents. But neither can we assume the sample is more

representative of lesbian women who want children, nor of lesbian women who are not

146

interested in parenting. The sample is more likely, however, to be representative of women
for whom the topic of parenthood is salient. The women invested their time and energy into
responding to a survey that took approximately 15 to 20 minutes to complete. It would seem
likely that the women completing this survey think that parenthood is an important topic.

In summary, we cannot say with certainty how representative these data are for all
lesbian women in Michigan. These data are more likely to be representative of educated
lesbian women in Michigan for whom the topic of parenting is salient. The most likely
shortcoming here in terms of gathering participants is that poor lesbian women are not
represented. And unfortunately, their experiences may be very different than the women who
participated in the research. Becoming pregnant through ”artificial means" and adoption are
quite expensive. Do poorer lesbian women have similar desires, but less resources with which
to accomplish them? Do they tend to give up on the idea before it is fully formed, knowing
the financial impossibility of the endeavor? Do they undertake very different strategies to
achieve parenthood?

Another factor to consider is the degree to which social desirability, acquiescence, or
response bias is present in the dataset. Lesbian women agreed that there is a widespread
notion that lesbian women are not mothers. Similar myths exist regarding a lack of nurturance
and an anti-family stance of lesbian women. Some respondents may have consciously or
unconsciously responded to the survey in ways that would negate these (negative) images. The
bias could be subtle or blatant. The bias could be manifested in the attitude or the intention
variables.

The degree of response bias is difficult to gauge. There is no research comparable to
the current study with which to compare the distribution of attitude scores and intentions

regarding parenthood. When additional research is undertaken the representativeness of the

147

current data will be easier to evaluate. There are reasons, however, to think that social
desirability and response bias may be somewhat minimized in the current study. The questions
in the survey were carefully worded to not imply any judgement or preference of response
choices. And about half of the women in the survey provided unpopular ideas - they did not
want to have children, and these women tended to have lower scores on the attitude toward
parenting scale. Also, many women approached the principal investigator with great
enthusiasm. They were eager for more data on lesbian women to become available.

Again, a caution - the representative of the sample compared to lesbian women in
general cannot be ascertained. The findings of the current study tend to suggest that up to one-
quarter of lesbian women are interested in parenting. The literature on lesbian parenting
suggests that there is a notable proportion of lesbian women who are having children or desire
children, and that there has been a widespread "gayby boom" in the lesbian community. And

there has been an explosion of new books about lesbian parenting in the early 1990's.

Implications for Future Research

It appears that lesbian parenting is a topic of interest to many lesbian women, even
those not intending to parent. More research on lesbian women is needed, to validate interest
in becoming a parent, the proportion of women who are interested in parenting, the behaviors
engaged in related to parenting, and the impact of having children on lesbian families and
friendship networks.

We also need to know more about the normative influences in decision-making. This
could include whether lesbian women feel pressured to become mothers, the degree to which
they feel conflict with being a lesbian and a mother, or if there are lesbian-specific concerns

that are barriers to becoming a parent. More in keeping with the TORA model, it would be

148

useful to gather more normative information such as who are the referent groups, whether
these people think respondents "should” become a parent, and what is the level of motivation
to comply with referent groups or individuals. In addition, there are personality attributes,
such as self-esteem, that are correlated with a lack of motivation to comply with others (See
Loken & Fishbein, 1980). Such variables could be useful in furthering our understanding of
the normative component of fertility decisions. Future research could focus more on the
normative component of the TORA model.

Further research is needed to assess the degree to which the current study is
representative of lesbian women demographically. Additional research is needed that taps into
lesbians with different socio-economic backgrounds specifically. This may mean going into
environments, such as bars, that are less conducive to data collection.

Future research specifically about the decision to parent is needed to replicate the
current findings, and to assess the validity of the relationships between attitude, intention, and
behavior described by the model. And more research is needed to further investigate and
assess the utility of the post-hoe modifications to the decision model.

Investigators who make attempts to replicate the current research or pursue related
topics would need to be cognizant of the variability of parenting desires in lesbian women.
Research predicting fertility decisions would need variability in a sample in terms of parenting
desires. In this case, it is recommended that data should be monitored as it is collected to
ascertain the variance in the interest in parenting. There are many other areas of interest that
could be pursued from this point forward and an investigator would need to be clear as to
what population should be tapped. For example, there is a need for research specifically
focused on the pro's and con's and considerations involved in the methods of parenthood. In

this case, an investigator would not be advised to seek lesbian women in general for the study,

149

but to recruit only lesbian mothers and lesbian women interested in parenting.

Conducting research on parenting decisions of lesbian women will be different from
that of heterosexual persons because of the variance in parenting desires. In previous research
investigating voluntary childlessness, recruitment of the sample was necessary. In a lesbian
population, one may more easily find childless lesbian women who intend to remain childless,
but may need to more actively recruit those who desire to parent. The variability of parenting
desires in lesbian women, as found in this sample, make lesbian women a potentially great

resource for understanding fertility decisions.

Future Directions

The data gathered in this study has been a first step in building a picture of lesbian
women's attitudes toward parenting and decisions to parent. Further research is needed,
however, both qualitative and quantitative. By combining quantitative and qualitative methods,
the information can provide a richer, fuller understanding of the concepts studied. This
research was an opportunity to develop and test a new instrument. Further testing of this
instrument, and modifications to it, are in order. Future research should also aid in our
understanding of fertility decision theories and their applicability to lesbian families, and
suggest whether additional theoretical concepts may be important in determining lesbian
women's attitudes toward parenting.

The data from the current research can be disseminated to the lesbian community,
other researchers, and the general public, to aid understanding of lesbian women, lesbian
families, and lesbian parents, and to help dispel some of the myths and fears regarding
lesbianism and lesbian motherhood. Forums for this dissemination could be in the context of

workshops, conferences or brochures.

150

The research data suggests the need for interventions with lesbian women in the areas
of family and parenting. Moreover, information dissemination and interventions can be done
collaboratively with the lesbian community. Interventions can be created from an advocacy
approach rather than a problem approach, as tools to help lesbian women help themselves.
The information gathered can be used to create: aids for decision-making; clarification of

values exercises; discussion groups; lesbian parenting support groups, and so forth.

BIBLIOGRAPHY

BIBLIOGRAPHY

Achtenberg, R. (1990). Pr servin e ' e f ili f lesbi m .
(Available from [National Center for Lesbian Rights, 1663 Mission Street, 5th floor,

San Francisco, CA. 94103]).

Ainslie, I ., & Feltey, K. M. (1991). Definition and dynamics of motherhood and family in
lesbian communities. Magiagg ang Family Raviaw, _1_7(1-2), 63-85.

Ajzen, I. & Fishbein. M. (1980). WWW
Englewood Cliffs, NJ: Prentice-Hall.

Beach, L. R., Townes, B. D., Campbell, F. L., & Keating, G. W. (1976). Developing and
testing a decision aid for birth planning decisions. W
W, 1.5, ”'116'

Beach, L. R. Campbell, F. L, & Townes, B. D. (1979). Subjective expected utility and the

prediction of birth-planning decisions WWW
Bedtime: 24 18-28

Beckman, L. J. (1987). Changes in motivation for parenthood among young married couples.
WW 2(2) 96-110

Bozett, F. (Ed.). (1987). W. New York: Praeger.

Bright, A. D., Manfredo, M. I ., Fishbein, M., & Bath, A. (1993). Application of the theory
of reasoned action to the National Park Service's controlled burn policy. maLQf

W11. 22(3). 263-280.

Bronfenbrenner, U. (1977). Toward an experimental ecology of human development.

AW 32(7), 513-531.

Bronfenbrenner, U. (1986). Ecology of the family as a context for human development:
Research perspectives. WM, 22(6), 723-742.

Brown, K. (1992). Family values. Dam, 2(6), 24-29,55-57.

Brown, K. (1993). Family values part 3. m, 3(2), 20-23.

151

152

Bryant, E. (1990). In the interest of R.C., minor child: The Colorado artificial insemination
by donor statute and the non- traditional family. v r niv r i
61(1), 79-89.

Burgess, R. G. (1982). Fielg researgh: A gogrgebggk and 11ng manual. Boston: Allen &

Unwin.

Bybee, D. (1990, March). Mi 1 bi ° R l v I .
Report submitted to Michigan Organization for Human Rights and Michigan
Department of Public Health. (Available from [Michigan Organization for Human

Rights, PO. Box 27383, Lansing, MI. 48909]).

Chan, D. K., & Fishbein, M. (1993). Determinants of college women's intentions to tell their

pMers to use condoms. rn l l' i P h l , 23(18), 1455-1470.

Chodorow, N. (1974). Family structure and feminine personality. In M. Z. Rosaldo & L.
Lamphere (Eds), WW (pp. 43-66). Stanford: Stanford

University.

Cohen. J. (1988). W (2nd ed). Hillsdale

NJ : Lawrence Erlbaum.

Corea, G. (1985). Waging. New York: Harper & Row.

Currie, D. (1988). Re-thinking what we do and how we do it: A study of reproductive
decisions. W21. 25(2). 231-253.

Davidson, A., & Jaccard, I. (1976). Social-psychological determinants of fertility intentions.

In S H Newman & V D Thompson (Eds ) Wand
W5. (DHEW Publ. No. (NIH) 76-574). Washington D. C.: U. S.

Government Printing Office.

DiLapi, E. M. (1989). Lesbian mothers and the motherhood hierarchy. LQIIInBLQf
mammalian 1.8(1-2). 101-121.

Donovan, C. A. (1983). The Uniform Parentage Act and nonmarital motherhood-by-choice.
W29. 11. 193-253

Dunn, S. K. (1988). A model of fertility decision-making styles among young mothers.
Hummflmanizatinn. M2). 167-175.

Evans. B. (1990)- Mothering as a lesbian issue. Wm. 2(1).
43-52.

Falk, P. J. (1989). Lesbian mothers: Psychosocial assumptions in family law. Am
13152119192153. M6). 941-947.

153

Fishbein, M. (1972). Toward an understanding of family planning behaviors. mm
Appligd Esygholou. 2(3), 214-227.

Fishbein, M. (1979). A theory of reasoned action: Some applications and implications. In H.

Howe & M. Page (Eds.), Ne rask m i m tiv i 7 (pp. 66-116).
Lincoln: University of Nebraska.

Fishbein. M.. & Allen. I. (1975). WWW
meog agd reaeargh. Reading, MA: Addison-Wesley.

Fishbein, M., Chan, D. K., O'Reilly, K., Schnell, D., Wood, R., Becker, C., & Cohn, D.
(1992). Attitudinal and normative factors as determinants of gay men's intentions to
perform AIDS-related sexual behaviors: A multisite analysis. W114

Watchmen. 2_2(l3). 999-1011.

Fishbein, M., Chan, D. K., O'Reilly, K., Schnell, D., Wood, R., Becker, C., & Cohn, D.
(1993). Factors influencing gay men's attitudes, subjective norms, and intentions with

reSpect to performing sexual behaviors JanmaLdLAnnlndadcnlisxcmldu. 23(6).
417-438.

Frank, D. I., & Brackley, M. H. (1989). The health experience of single women who have
children through artificial insemination. WW5}, 3(3), 156-160.

Fried, E. S., & Udry, J. R. (1980). Normative pressures on fertility planning. mum
Enximment. 3(3-4). 199-209.

Gartrell, N. (1984). Issues in psychotherapy with lesbian women. W.
Wellesley, MA: Stone Center Working Paper Series.

G11 de Lamadrid, M. (Ed..) (1991).

Iaslzianuhnosmamnthathmdmumnlicamm
W. (Available from [National Center for Lesbian
Rights, 1663 Mission Street, 5th floor, San Francisco, CA. 94103]).

Gordon. L. (1976). MW
Alum. New York: Penguin.

Gore, S. A., & Frank, D. W. (1993, March). WWW. Paper

presented at the Association for Women in Psychology Conference, Atlanta, Georgia.

Hanscombe. G E & Forster. J (1981) Wm
M. London: Peter Owen.

Heller, K., Price, R. H., Reinharz, S. Riger, S., & Wandersman, A. (1984). W
W Homewood. IL: Dorsey Press

154

Hoyle, R., & Panter. A. (1995). Writing about structural equation models. In R. Hoyle (Ed.),

Smgmga] agaation mglaling; Qongepts, isaaga, and applicatigm (pp.158-176).
Thousand Oaks, CA: Sage.

In re R.C., Minor child, 775 P.2d 27 (C010. 1989).

Jaccard, J., & Davidson, A. (1975). A comparison of two models of social behavior: Results
of a survey sample. Sociomeg, 18, 491-517.

Jaccard, J ., Hand, D., Ku, L., Richardson, K., & Abella, R. (1981). Attitudes. toward male
oral contraceptives: Implications for models of the relationship between beliefs and

attitudes. u f i i IP 01 . QB). 181-191.

Jaccard, J., Helbig, D. W., Wan, C. K., Gutrnan, M. A., & Kritz-Silverstein, D. C. (1990).
Individual differences in attitude-behavior consistency: The prediction of contraceptive

behavior. W. 2.00). 575-617.

Janis. l L.. &Mann. L (1977) W
W. New York: The Free Press.

Johnson, S. R., Smith, E. M., & Guenther, S. M. . (1987). Parenting desires among bisexual
women and lesbians. W126). 198-200.

Johnston, M., Shaw, R., & Bird, D. (1987). "Test-tube baby” procedures: Stress &
judgements under uncertainty. With, _1, 25-38.

Kaiser, D. (1988). Artificial insemination: Donor rights in situations involving unmarried

recipients. Wm. 2§(4). 793-811.

Kenney, J. W., & Tash, D. T. (1992). Lesbian childbearing couples' dilemmas and decisions.
WM. 13. 209-219.

Krieger. S. (1983). WWW Philadelphia:

Temple University.
Labine v. Vincent, 401 US. 532 (1971).

Lesbian Mothers' Group. (1987). Our lovers will never be fathers, but our community can be

our family. In S. Pollack & J. Vaughn (Eds.), W
W (pp. 289-295). Ithaca, New York: Firebrand.

Lewin. E. (1993).L&sblan.tndtheILAc_c_an§atxendeL1n_mnencmnm Ithaca: Cornell

University.

Linn, R. (1991). Mature unwed mothers in Israel: Socio-moral and psychological dilemmas.

LifesndeLEaniilundflconnmiussnas.12(2). 145-170.

155

Lockard, D. (1985). The lesbian community: An anthropological approach. mpg:
Houseman. 1104). 83-95.

Loken, B., & Fishbein, M. (1980). An analysis of the effects of occupational variables on

childbearing intentions. lppmal pf Applied Smlal Psyphplpgy, 19(3), 202-223.

McCandlish, B. M. (1987). Against all odds: Lesbian mother family dynamics. In F. Bozett
(Ed.), WW (pp. 23-36). New York: Praeger.

McCartney, C. F. (1985). Decision by single women to conceive by artificial donor

insemination WW. 4(4). 321-328

McCarty, D. (1981). Changing contraceptive usage intentions: A test of the Fishbein model of
intention. l f li ' lP . 11(3), 192-211.

Mencher, J. (1990). Intimacy 1n 2lesbian relationships: A critical re-examination of fusion.
flpr k in EppgppssN 9,4 2.Wellesley, MA: Stone Center Working Paper Series.

Miller, W. B, & Shain, R. N. (1985). Married women and contraceptive sterilization:

Factors that contribute to pre-surgical ambivalence. WM, 11,
471-479.

Miller, W. B. Shain, R. N., & Pasta, D. J. (1990). The nature and dynamics of

post-sterilization regret in married women. Wampum,
2_Q(6), 506-530.

Neal, A. G., & Groat, T. H. (1980). Fertility decision making, unintended births, and the

social drift hypothesis: A longitudinal study. WW, 3(3-4),
221-236.

Oakley, D. (1985). Premarital childbearing decision making. W, 33(4),
561-563.

Oakley, D. (1986). Low-fertility childbearing decision making. W, 31(3-4),
249-258.

Oakley, D., & Schechtrnan, B. (1980, September). Uncertainty in childbearing decision
making. anlznallaranthnndladan pp- 2-3-

O'Brien. M. (1989). W Boulder. CO:

Westview Press.

Ory, M. G. (1978). The decision to parent or not: Normative and structural components.

InmalaLMattiageandthflEamilx. £90). 531-539

156

Office of Technology Assessment U.S. Congress. (1988).A_rt1_fip1_al_1__erLaanL_ELagn§§__
ni . of 1 7 urv -b kr d (.ReportNo.

OTA-BA- BP-48). Washington DC: US. Government Printing Office.

Pearlman, S. F. (1987). The saga of continuing clash in lesbian community, or will an army
of ex-lovers fail? In Boston Lesbian Psychologies Collective (Ed.), mm

h l ' : Ex I r ' lle e (pp. 313-326). Chicago: University of

Illinois.

Pies, C. (1987a). Considering parenthood: Psychosocial issues for gay men and lesbians

choosing alternative fertilization. In F. Bozett (Ed), WW (pp.
165-174). New York: Praeger.

Pies, C. (1987b). Lesbians choosing children: The use of social group work in maintaining
and strengthening the primary relationship. .LnumaLntfinniaLVXndsandHnmm
Santana. 5(2). 79-88.

Pies, C. (1988). W (2nd ed.). Minneapolis: Spinsters Ink.

Pies, C. A. (1989). Lesbians and the choice to parent. WM, 15(3-4),
137-154.

Polikoff, N. D. (1987). Lesbian mothers, lesbian families. Legal obstacles, legal challenges.

In S Pollack & l Vaughn (Eds) WWW
(pp. 325- 332). Ithaca: Firebrand.

Pollack, S. (1987). A lesbian-feminist perspective on research. In S. Pollack & J. Vaughn
(Eds.) WWW (pp 316-324) Ithaca:

Firebrand.

Pollack, S. (1990). Lesbian parents: Claiming our visibility. MW, 19(1-2),
181-194.

Rabin, A. (1965). Motivation for parenthood. W
W. 22. 405-411

Rappaport. 1. (1977). WE New York: Holt.
Rinehart, & Winston.

Reproductive technology and the procreation rights of the unmarried. (1985). W
am. 28. 669-685.

Rich, A. (1980). Compulsory heterosexuality and lesbian existence. Wm
Ween-11631660

157

Richardson. L. W. (1981). W (2nd
ed.). Boston: Houghton Mifflin.

Robinson, 5., & Pizer, H. F. (1985). Having a baby without a man. New York: Simon &

Schuster.

Rohrbaugh, J. B. (1988). Choosing children: Psychological issues in lesbian parenting.
Wamnundlnerany. 8(1). 51-64.

Rothman, B. K. (1989). : ni-
mm. New York: W. W. Norton.

 

Schneider, M., & O' Neill, B. (1993). Eligibility of lesbians and gay men for spousal benefits:
A social policy per8pective WWW. 2(1). 23 -31

Shaw, N. S. (1989). New research issues in lesbian health. meen'§ Smdips, 11, 125-137.

Silka, L., & Kiesler, S. (1977). Couples who choose to remain childless. EanLilLElanning
Primates. 2(1). 1625.

Slater, S., & Mencher, J. (1991). The lesbian family life cycle: A contextual approach.
Americaninin'nainLQttnonsxnniamr. 6.1(3). 372- 382

Spallone. P.. & Steinberg. D L (Eds) (1987) Madnnatdnnlhemythattemndyn
W. New York: Pergamon Press.

Stiglitz, E. (1990). Caught between two worlds: The impact of a child on a lesbian couple's
relationship. W. _1_Q(l-2). 99-116.

Straits, B. C. (1985). Factors influencing college women' s responses to fertility

decision-making vignettes LnnmaLnfMattiagunflhafiainily. 41(3). 585-596

Sunderland, A. (1990). Influence of human immunodeficiency virus infection on reproductive

decisions WWW. 11(3). 585-594

Tennyson, M. A. (1988). Experiences of a woman who intended to relinquish her infant for

adoption. MAW. 3(3). 139-152.

Townes, B. D., Beach, L. R., Campbell, F. L., & Wood, R. L. (1980). Family building: A

social psychological study of fertility decisions. Wm, 3(3-4),
210-220.

Townes, B. D., Campbell, F. L., Beach, L. R., & Martin, D. C. (1976). Birth-planning

values and decisions. Preliminary findings. WW
educadnnalasnss (pp 113- 130)

158

Voss, J. H., Knaub, P. K., & Eversoll, D. B. (1984). Perceptions of 1ssues regarding timing
of parenthood held by college men and women: A comparative study. _apenting
Studies. 1(2). 47-53

Weber v. Aema Casualty & Surety Co., 406 US. 164 (1972).

West, K. K., & Morgan, L. A. (1987). Public perceptions of the ideal number of children for
contemporary families. W. 2(3). 160-172.

West, S., Finch, J, & Curran, P. (1995). Structural equation models with nonnormal
variables: Problems and remedies. In R. Hoyle (Ed. ), W
W (pp 56-75) Thousand Oaks. CA: Sage

Wood, R. J., Campbell, F. L., Townes, B. D., & Beach, L. R. (1977). Birth planning
decisions. hlblip flgaldl Brlgfg, 61(6), 563-565.

Zeidenstein, L. (1990). Gynecological and childbearing needs of lesbians. ,lmaLpf
Eutsedylddmfem. 35(1). 10-18-

APPENDICES

APPENDIX A

Appendix A

Example of an Initial Contact Letter to a Community Organization

(NOTE: The actual letter was sent on department letterhead)

July 21, 1994

Lesbian Connection
PO. Box 811
East Lansing, MI. 48826

Dear Lesbian Connection:

I am a lesbian graduate student at Michigan State University surveying adult lesbians
about parenting. The survey asks about things like the role of the community in our families,
how many lesbians are mothers, how many plan to become mothers, and the decision factors
in becoming a parent. I am interested not only in lesbians who are or plan to become mothers
but in all lesbians' views about motherhood.

I am writing to you because I need your assistance - I hope you will help me get in
touch with lesbians in your organization and allow the survey to be mailed to some of your
members. There are several different ways to do the mailing, and different ways for your
organization to help.

In the later part of August I will call you to discuss the survey and how your
organization can participate. I consider this project a community effort and hope that you will
not see this project as a burden, but as an opportunity. I plan to share the survey information
gathered with the participating organizations and would like to explore ways we can work
together.

The research I am proposing is worthwhile and the information gathered will
ultimately be used to benefit lesbian families and communities in Michigan. Our families are
important and we deserve recognition. I feel confident that you will do all that you feel
comfortable doing to make this a successful survey.

Peace,

Nancy McCrohan

Department of Psychology OR (home) 1028 Parker
Michigan State University Lansing, MI 48912
East Lansing, MI 48824-1117 517-485-5774

159

APPENDIX B

Appendix B

Cover Letter of the Lesbian Connection Mailing

(NOTE: The actual cover letter was sent on LC letterhead)

May 10, 1995
Hi!

We're sending the enclosed survey to about one-fourth of our Michigan mailing list on
behalf of Nancy McCrohan, an MSU graduate student who is looking for lesbians in Michigan
to fill out her survey. The survey is completely voluntary and anonymous, and asks your
Opinions about parenting and whether or not you want to become a parent.

The student wants a variety of lesbians to fill out the survey, including women who do
not want to have children, women who are not sure, women who want children, and women

who are already parents.

The survey takes about 20 to 25 minutes to fill out, and asks about things such as
societal attitudes toward lesbian parenting, suppOrt from others for not being a parent or
support for being a parent, attitudes about motherhood as an identity or role, reasons for
having children and reasons for not having children, and demographic questions.

You can rest assured that Lg's mailing list remains confidential - our mailing list is
never sold or given away. We ourselves did all the addressing of the envelopes.

You can use the enclosed pro-paid business envelope to return you survey. Nancy also
arranged for a discount coupon for Goldenrod as a thank you for filling out the survey.

If you have any questions or comments feel free to drop us a line or give us a call.
Thanks for your help.

For lesbian Connection,
The Ambitious Amazons - Margy & Sandy

160

APPENDIX C

Appendix C

Survey Instrument
(NOTE: This instrument has been photographically reduced)

DON 'T WANT TO HA VE CHILDREN ??... THIS SURVEY IS FOR YOU I!
LOOKING FORWARD TO HAVING CHILDREN ??...THIS SURVEY IS FOR YOU !!
NOT SURE IF 8: WHEN YOU WANT CHILDREN ??...THIS SURVEY IS FOR YOU I!
YOU ALREADY HA VE CHILDREN .72... THIS SURVEY IS FOR YOU I!
SURVEY OF LESBIAN WOMEN'S OPINIONS

ABOUT FAMILIES AND CHILDREN

This is a survey about lesbians' ideas about our families and becoming parents. This survey is for
ANY adult lesbian in Michigan, NOT JUST lesbians who are mothers, or who plan to become
mothers. Lesbians, lesbian couples. and lesbians with children have been mostly invisible in our
society. although this has been changing a little recently. With this survey, I hope to better
understand our families and ideas about parenting. You can help by filling out this survey. Women
are individuals, with all kinds of experiences and lives. I tried to keep this in mind when I wrote the
questions. I want to learn about all kinds of lesbian women, from all races and classes and ages and
families. If you have ideas about families that I don't ask about, please write on the back of the
survey or add another sheet of paper.

This survey is being done by Nancy McCrohan. a lesbian graduate student at Michigan State
University. A report of the findings of this survey will be given to lesbian organizations in Michigan
when it is completed. I hope to use the information to help ourselves and our families.

If you have already filled out a survey, pass on the extra copy of the survey to another lesbian who
has not filled it out. If you know of other lesbians who would like to fill out this survey. make copies
or write me to ask for more copies.

The survey is confidential and anonymous, so please do NOT put your name on it. This survey will
take about 20-25 minutes to complete. You indicate your voluntary agreement to participate by
completing and returning this survey. If there are any questions you don't want to answer, just
leave them blank. If you have any questions or concerns about this survey, please contact Nancy
McCrohan at 517-485-5774.

THANKSI!

Please return this survey as soon as possible or by June 15. 1995.

I61

162

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

P R INF M Tl
A1. To what degree do you identify as a lesbian? Circle the number below that best describes your sexual
identity.
Lesbian Bisexual Heterosexual
Identity 1 2 3 4 5 6 7 Identity
A2. About how many people in the following groups know your sexual orientation (you are ”out” to)?
Circle the answer that best describes the amount of people.
Family None A few About Most Everyone
half
Lesbian I Gay friends None A few About Most Everyone
half
Straight friends None A few About Most Everyone
half
Co-workers None A few About Most Everyone
half
A3. Please circle the number of newsletters or publications you read or to which you subscribe.
Q 1 2m lesbian publications
9 1 2 pr mprg lesbian/gay publications (for both men and women)
Q 1 2m feminist Iwomen's publications (not lesbian only)
A4. How often do you go to lesbian Frequently Occasionally Rarely Never
/gay bars?
A5. How often do you attend lesbian Frequently Occasionally Rarely Never
lgay social events, such as
concerts or dances?
A6. How often do you participate in Frequently Occasionally Rarely Never
lesbian /gay political events.
4 such as marches?
A7. Do you have any close lesbian friends Yes No
who have children?
A8. Do you have any close lesbian friends Yes No
who are thinking about having children?
A9. Do you have any close lesbian friends Yes No

actively trying to have a child?

163

A10. How would you describe your current relationship? Circle one answer.
1. In a monogamous relationship with one woman
Length of this relationship:
In a primary relationship with one woman, but have other sexual partners
Single, seeing more than one woman
Single, but involved with one woman
Single, not seeing anyone romantically

Other (Please describe I

 

 

Power of attorney? Yes No
Patient advocate for medical decisions? Yes No
Yes No
Yes No

Last will 81 testament?
Co-parenting agreement?

2
3
4
5
6
A11. Have you ever completed any legal paperwork with a partner for:
1
2
3
4
5 Other? (Please describe I

 

 

 

 

 

 

 

 

 

 

 

 

A12. Have you ever had a commitment ceremony? Yes No

A13. How often did you take care of children More About More About Rarely
when you were under 18 years old? than weekly than monthly or never
(This includes paid or unpaid, watching weekly monthly
relatives or non-relatives)

A14. How often have you taken care of More About More About Rarely
children in the past two years? (Not than weekly than monthly or never
including your own if you are a parent) weekly monthly

A15. How often do you have contact with More About More About Rarely
children right now ? (Not including your than weekly than monthly or never
own if you are a parent) weekly monthly

A16. How often would you like to have More About More About Rarely
contact with children right now ? (Not than weekly than monthly or never
including your own if you are a parent) weekly monthly

A17. Are you currently a parent or a co-parent?
1. Yes
2. No

A18.

1E YOU ARE A PARENT please give the age 81 sex of each of your children, and whether you identified
as a lesbian at the time you had that child.

 

 

Age and gender of child

 

 

Did you have the child
before or after coming out? ‘

Before

After

 
 

   
 

 

 

 

A19.

A20.

A21 .
A22.
A23.
A24.
A25.
A26.

A27.

A28.

164

Do you have a partner who already had children before you were partners?

P’P'PP’P.‘

No, l have no partner
No. my partner did not have children before we were partners
Yes. but I don't consider myself a co-parent

Yes. and now I feel I am a co-parent

Yes. but I am not sure of my role with the children

Yes. and my role with the children is:

 

With whom do you live? Circle ALL that apply.

$999.“???pr

How many sisters do you have?
How many brothers do you have?
What year were you born?
What is your yearly household income before taxes?
What is your main job or occupation?

What is your zip code?

Lover / partner
Communal or cooperative housing
Female roommatelsl

Male roommatelsl

No one (live alonel

My childlrenl

Partner's childlrenl

Our childlrenl (mine and my partner's)
Other (Please describe)

 

 

 

 

 

 

 

How many years of school have you completed? Circle ONE that fits best.

P’P‘PPPT'E FNPPPP’NT‘

8 years or less (no high school)

Some high school

Graduated high school or GED
Vocational Training

Some college

College degree

Some graduate or professional school
Graduate or professional degree

hat is your race or ethnic background? Circle ONE that fits best.

Asian
Black / African American
Hispanic

Native American (Indian)
White I Caucasian
Other (Please describe)

 

165
I P NT

PLEASE ANSWER THE FOLLOWING QUESTIONS WHETHER OR NOT YOU PLAN TO HAVE CHILDREN, OR
WHETHER OR NOT YOU ARE A PARENT. There are many ideas and concerns people hold about having
children. People often feel mixed about having children. Sometimes people want and don't want children all
at the same time.

81. What statement below best describes your FEELINGS about becoming a parent? Do you WANT to
have a child Ichildren /another child? Circle the number that best describes your feelings.

I do not want to have a child

Most of the time I think that I don't want to have a child

I am not sure if I want to have a child

Most of the time I think that I want to have a child

I definitely want to have a child

9991“."

82. What statement below best describes your PLANS about becoming a parent? Do you INTEND to have
a child Ichildren /another child? Circle the number that best describes your feelings.

I do not intend to have a child > (Go to Cannon 8.?)

Most of the time I don't intend to have a child

I am not sure if I intend to have a child (Go to Question 84)

Most of the time I intend to have a child

I definitely intend to have a child --> (Go to Question 85)

 

5".“pr

83. How often do you think about your intention to NOT have a child Ichildren /another child?
Very frequently

Frequently -

Occasionally (Go to Ouashbn B7)

Rarely

Never

5".“wa

84. How often do you think about WHETHER you will have a child Ichildren /another child?
Very frequently

Frequently

Occasionally (Go to Ouaso'on 86)

Rarely

Never

9‘99”"?

85. How often do you think about your intention TO HAVE a child Ichildren /another child?
1 Very frequently
2. Frequently

3. Occasionally
4. Rarely
5 Never

B6. E you want children (or more children). how many children TOTAL would you like to have?

87. What would NOT APPEAL to you about having a child (or another child)?

 

 

 

166

88. What would APPEAL to you about having a child (or another child)?

 

 

 

BS. What are (or would be) your greatest worries about becoming a parent/being a parent?

 

 

 

BABE—“MAW

What is your opinion of the following statements? Do you strongly agree. agree, neither agree nor disagree.
disagree. or strongly disagree with the following statements? Circle the answer that best describes how you

feel about each statement.

 

 

 

 

 

 

 

 

 

 

 

 

C1. MOST PEOPLE assume that Neither
lesbians do NOT become mothers Strongly Agree agree Disagree Strongly
agree nor disagree

disagree

C2. MOST LESBIANS assume that Neither
lesbians do NOT become mothers Strongly Agree agree Disagree Strongly
agree nor disagree

disagree

C3. The way the law treats lesbian Neither
mothers is a factor in my decision Strongly Agree agree Disagree Strongly
to parent OR to not parent agree nor disagree

disagree

C4. It is important for me to have the Neither
support of my friends in my Strongly Agree agree Disagree Strongly
decision to parent OR to not parent agree nor disagree

disagree

C5. It is important for me to have the Neither
support of my family in my decision I Strongly Agree agree Disagree Strongly
to parent OR to not parent agree nor disagree

disagree

C6. It is important for me to have the Neither
support of the lesbian community in Strongly Agree agree Disagree Strongly
my decision to parent OR to not agree nor disagree

parent disagree

 

 

Again, the next few pages should be answered by EVERYONE, whether you are a parent or not, whether you
even want a child (or another child) or not. People have ideas that may influence if, when. and how many
children they will have. Here are some ideas that people have about IF and WHEN they want a child. To what
degree do you agree or disagree with the following statements? Circle the answer that most closely matches

167

your feelings.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SA = Strongly Agree
A =- Agree
N = Neither Agree nor Disagree
0 = Disagree
SD = Strongly Disagree
np -= no partner Idoes not apply
C7. I want to love a child SA A N D SD
C8. Having a child would make my family complete SA A N 0 SD
C9. I want to nurture a child and help her or him grow SA A N D SD
C10. I don't like being around children SA A N 0 SD
C11. I have never seriously considered having children SA A N D SD
C12. I don't want to add to the population explosion SA A N D so
C13. Caring for a child is a joyful experience SA A N 0 SD
C14. I don't want to experience pregnancy or childbirth SA A N 0 SD
C15. 1 would like be involved with a child as a “close aunt' or SA A N 0 SD
close friend
C16. I don't have enough time to be a parent SA A N D SD
C17. It's important to have insurance before having a child SA A N 0 SD
C18. After you have children you make time for them SA A N D SD
C19. Having a child would be a contribution to society SA A N D so
C20. There is so much to think about and plan for before having SA A N D SD
children
C21. Arranging for legal recognition of a co-parent is important SA A N D SD
if a lesbian couple has children
C22. My lesbian friends will support my decision to have or to SA A N D SD
not have children
C23. It would be a problem having a child 'outing' me SA A N D SD
024. I have never pictured myself as a mother SA A N D SD
C25. I want to leave a part of me behind SA A N 0 SD
C26. I worry that my children could be harassed for having a SA A N D SD
lesbian mother
C27. I want the challenge of raising a child SA A N 0 SD

 

 

168

 

SA a Strongly Agree

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A - Agree
N :- Neither Agree nor Disagree
D a Disagree
SD = Strongly Disagree
np = no partner Idoes not apply
C28. Being a parent does not fit with how I see myself SA A SD
C29. I want to do other things (like travel or finish my SA A N 0 SD
education) before having any children
C30. Having a child would create a connection to the past and SA A N 0 SD
the future
C31. I think motherhood is an oppressive institution SA A 0 SD
C32. I have thought a lot about whether or not I want to have a SA A 0 SD
child
C33. I don't want to interrupt my job / career (or partner's iobl SA A D SD
C34. I don't want to bring children into the world the way it is SA A SD
C35. Having a child is an experience I want to share with a SA A SD
partner
C36. Having a child would increase the bond with my partner SA A 0 SD NP
C37. 1 would not feel comfortable as a mother SA A SD
C38. I would be very concerned about a sperm donor's legal SA A N D so
rights to a child
C39. Having a child would create a sense of family SA A N D SD
C40. I would rather spend money on other things SA A N D SD
C41. I don't want to be tied down with children SA A N 0 SD
C42. 1 want to have an impact on the world SA A N D SD
C43. I don't want the responsibility of having a child SA A N 0 SD
C44. When a lesbian couple has a child. it is difficult for the SA A N D SD
non-biological co-parent to be recognized as a full parent
C45. My partner doesn't want children SA A N 0 SD NP l
C46. I find children too stimulating SA A N D SD
C47. Lesbian mothers get support from the lesbian community SA A N 0 SD
C48. It is important to have the support of my partner in my SA A N 0 SD NP
decision to parent OR to not parent
C49. When I think of the future. I don't see myself having SA A N 0 SD

children

 

 

 

 

 

 

 

169
p , F TH

How do you think other people would feel if you had a child? Do you think you would have the support of
these people if you decided to have a child? If you already have children, how much support do you think you

have from these people right now?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

VS = Very Supportive

SS = Somewhat Supportive

N = Neither Supportive nor Unsupportive

SU a Somewhat Unsupportive

VU = Very Unsupportive

na :- Does Not Apply to Me (don't have

these persons in my life)

D1 . Mother VS SS N SU VU na
DZ. Father VS SS N SU VU na
D3. Brothers VS SS N SU VU na
D4. Sisters VS SS N SU VU na
DE. Other relatives VS SS N SU VU na
06. Partner - VS SS N SU VU na
07. My children 'vs 58 N su vu na
DB. Lesbian community VS SS N SU VU na
09. Partner's family VS SS N SU VU na
01 0. Straight friends without children VS SS N SU VU na
DI 1. Straight friends with children VS SS N SU VU na
012. Lesbian friends without children VS SS N SU VU na
01 3. Lesbian friends with children VS SS N SU VU na
DI 4. Co-workers VS SS N SU VU na

 

 

 

 

 

 

 

 

 

170

EABI E, INTEBMENI IQNS

Here is a list of activities that could be used to help people think about whether they want to have children
or not. or to help people feel comfortable with their choices. How helpful would these activities be for you.

personally? Circle the answer that best describes your opinion of the helpfulness of each activity.

 

 

 

 

 

 

 

 

 

 

 

 

 

E1. Discussion group for lesbians who aren't sure whether Helpful Not sure Not
they want to become parents helpful

E2. Practical information about "How To" become a parent Helpful Not sure Not
helpful

E3. Help with clarifying values about parenting Helpful Not sure Not
helpful

E4. Ideas about how to talk with one's partner about Helpful Not sure Not
parenting desires helpful

E5. Discussion group for lesbians who definitely do NOT Helpful Not sure Not
want to become mothers helpful

E6. Support group or discussion group for lesbian mothers Helpful Not sure Not
helpful

E7. Ideas about getting community support for lesbian Helpful Not sure Not
women who do NOT want children helpful

E8. Ideas about getting community Support for lesbian Helpful Not sure Not
mothers Helpful

E9. Ideas on how to deal with your new lover's children Helpful Not sure Not
Helpful

 

 

171

PART F. W

F1. Have you OR your partner had ANY discussion about parenting OR done ANY activities related to
becoming a parent. in the past two years?
1. Yes ------ > (If Yes, please answer the quesdons on this page I
2. No ----- > (If No, go to last page)

F2. Have you OR your partner done ANY OF THE FOLLOWING ACTIVITIES related to becoming a parent.
in the past two years? For each activity, circle the appropriate answer.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1. Contacted an agency regarding a home study for an Yes No
adopflon

2. Had a home study for an adoption completed Yes No
3. Talked with other lesbians who are mothers Yes No
4. Talked with other lesbians who want to be mothers Yes No
5. Charted ovulation cycle Yes No
6. Contacted sperm bank(s) Yes No
7. Discussed ideas with my partner Yes No
8. Discussed ideas with potential sperm donors Yes No
9. Inseminated under a doctor's supervision Yes No
10. Inseminated at home Yes No
11. Used ovulation regulating drugs Yes No
12. Applied for foster parent status Yes No
13. Tried to adopt Yes No
14. Discussed ideas with potential “surrogate“ mother Yes No
15. Other (Please describe)

 

 

 

F3. Have you or your partner seriously considered any of the following METHODS FOR BECOMING A

 

 

 

 

 

PARENT?

1. Pregnancy from an anonymous sperm donor _Yes No

2. Pregnancy from a known sperm donor Yes No

3. Pregnancy from sperm of a male relative of the Yes No
nonbiological partner

4. Foster parenting Yes No
Adoption Yes No

 

6. Discussed ideas with potential "surrogate" mother Yes No

 

 

 

 

 

172

THANK-YOU very much for your information and your time.

If you would like to add anything else, please do so.

If you would like to be part of a one-time discussion group about this survey to be held
Fall 1995, please send a note with your name. address. and phone number to the
address listed below. Please send your note of interest separate from this survey in
order to protect your anonymity.

To return this survey. use the envelope inside the survey.
(If the envelope is not there. please mail to Nancy McCrohan, Department of
PsychologY. Michigan State University, East Lansing, MI 48824).

"illll'lllllllillllllll