Minnesota/Texas Adoption Research Project

Introduction and Purpose

Since the mid-1970's, adoption practices in the U.S. have changed dramatically, and the confidentiality maintained in the past is no longer the norm. The trend is toward "openness" in adoption, in which contact occurs between the adoptive family and birthparent(s), either directly or mediated (e.g., through an adoption agency). Some adoption professionals argue that fully open adoption should be standard practice, that the secrecy of confidential adoptions has been harmful to all parties involved. Others argue that openness is harmful and experimental. Their view is that confidential adoption worked well, so why change it?

Although such professionals hold strong feelings about adoption, almost no research on this topic has been available to guide adoption policy and to answer basic questions about the dynamics of adoptive families and other types of families "yoked" by a common person or a common relationship. In order to address these questions, we designed a research project that focuses on the consequences of variations in openness in adoption for all members of the adoption triad: birthmothers, adoptive parents, and adopted children, and for the relationships within these family systems.

Participants, Methods, and Procedures

Adoptive families and birthmothers were recruited for the study through 35 adoption agencies located across the United States. We sought families in which there was at least one adopted child (the "target child") between ages 4 - 12 at the time of the interview, who was adopted through an agency before his or her first birthday; in which the adoption was not transracial, international, or "special needs"; and in which both adoptive parents were married to the partner they had at the time of the adoption. We simultaneously sought birthmothers who made adoption plans for children placed with these families. Participants in the study were located in 23 different states from all regions of the U.S., making this study the only nationwide one of its kind.

Adoption agencies which were invited to participate in the study had had placement experience with different levels of openness. We asked each agency to select all children who met the criteria outlined above, then to sample randomly among them within levels of openness until they located a set number of families and birthmothers willing to be interviewed. A few families (12 of 190, or 6.3%) were recruited through advertisements in newspapers and periodicals. Although this sample is not a fully random one, families were specifically not recruited on the basis of their success with adoption or their having an interesting story to tell, which is often a problem in volunteer samples.

The final sample in the original study includes 720 individuals: both parents in 190 adoptive families, at least one adopted child in 171 of the families, and 169 birthmothers. The vast majority of adoptive parents were white, Protestant and middle to upper-middle class. Virtually all had adopted because of infertility. The average level of education was 16.2 years for adoptive fathers, and 15.1 for adoptive mothers. Adoptive fathers ranged in age from 32 to 48 (mean=40.7) and adoptive mothers from 31 to 46 (mean=39.1). The average number of adopted children in each home was 1.8. Ninety of the target adopted children were male and 81 were female. Their ages ranged from 4 - 12 (mean=7.8 years). At the time of the birth of their child, the birthmothers ranged in age from 14 to 36 years (mean=19.1). Almost 2/3 of the birthmothers delivered when they were teenagers. At the time of the study, the average age of the birthmothers was 27.1, and the average number of years of education attained was 13.5. The majority of birthmothers indicated a current annual income in the range of $20,000 - 29,000. Almost all were white, half were currently married, and they had from 1 - 5 children.

We attempted to interview as many fully corresponding sets of adoptive parents and birthmothers as possible (i.e., cases in which the birthmother, her child by birth, and the family who adopted her child were all interviewed). The sample includes 77 fully corresponding sets: 11 confidential, 36 mediated, and 30 fully disclosed. Families were sampled across the full range of openness in the adoption. Five major categories were used to differentiate among levels of openness:

confidential adoptions, in which no information was shared between birth- and adoptive parents after 6 months post-placement; (N=62)

time-limited mediated adoptions, in which information was relayed between adoptive parents and birthmothers by a caseworker at the adoption agency, but the information sharing had stopped by the time we interviewed the participants and the parties did not intend to continue communication; (N=17)

ongoing mediated adoptions, in which information exchange mediated by the agency was continuing; (N=52)

time-limited fully disclosed adoptions, in which adoptive parents and the birthmother had had direct fully identified contact which had stopped by the time we interviewed the participants and the parties did not intend to continue communication; (N=2)

and ongoing fully disclosed adoptions, in which direct sharing of information was continuing, usually accompanied by face-to-face meetings (N=57).

Adoptive families were interviewed in their homes in one session that lasted 3 - 4 hours. The session included separate interviews with each parent and with the target adopted child; administration of several questionnaires; and a joint couples interview with the adoptive parents. Birthmothers were interviewed in their home or by telephone. They also completed several questionnaires.


Adoptive Parent Measures: The following measures were administered separately to the adoptive father and the adoptive mother:
ø Demographic Questionnaire: basic information about age, education, occupation, income, ethnicity, religion, and family composition.
ø Adoptive Parent Interview: extensive questions concerning motivation for adoption, experience with adoptive placement, and experiences and feelings about their level of adoptive openness.
ø Acknowledgment of Differences Scale: assesses the amount of acknowledgment of difference, empathy, and communication in the adoptive family. Questions are answered on Likert scales, once pertaining to the period around 6 months after the child's adoption, and once pertaining to the present (Kirk, 1981, as modified by McRoy, Grotevant, & Zurcher, 1988).
ø Child Adaptive Behavior Inventory: 91-item questionnaire about the target child (Miller, 1987); scored in terms of 4 factors: intellectual engagement, poor emotional control, social isolation, and symptoms.
ø Parenting Stress Index: a self-report questionnaire (Abidin, 1986); scales focus on aspects of the child, the parent, and their context that might contribute to parenting stress. The measure has been normed on both clinical and non-clinical samples of parents.
ø Twenty Statements Test: (Kuhn & McPartland, 1954) provides an open-ended and unstructured way for participants to describe themselves. The instrument allows researchers to observe in a systematic fashion aspects of the self based on social relationships (such as those involving adoption).

Following the individual adoptive parent interviews, the couple was brought together to respond to a final series of questions in a Conjoint Couple's Interview concerning their experiences and attitudes as adoptive parents.

Adopted Child Measures

ø Child Interview: designed to elicit open discussion of the child's experiences and feelings about his or her adoptive family situation and knowledge of and attitudes about his or her birthparents; covers general adoption issues as well as issues specific to the level of openness of the child's adoption. Special training was provided for interviewers who worked with the children, and interviewers spent extra time to develop rapport with the child before the interview.
ø Understanding of Adoption Scale: administered as part of the child interview to examine children's understanding of adoption, nature of adoptive family relationships, motives underlying adoption, and adoptive placement; scored in terms of 6 levels of social-cognitive understanding of the adoptive family relationship, ranging from level 0 (no understanding of adoption) to level 5 (adoptive parent-child relationship is seen as permanent and is based on the legal transfer of rights from birth to adoptive parents) (Brodzinsky, Singer & Braff, 1984).
ø Self-Perception Scale for Children: assesses self-concept in children who were age 8 and older; 36 items scored for 6 subscales: cognitive competence, athletic competence, social acceptance, physical appearance, behavioral conduct, and general self-worth (Harter, 1985).
ø Twenty Statements Test: see above

Birthmother Measures

ø Demographic Questionnaire: basic information about age, education, occupation, income, ethnicity, religion, and family composition.
ø Ego Identity Interview: examines exploration and commitment in the domain of career identity (Grotevant & Cooper, 1981);
ø Intimacy Interview: assesses levels of relationship maturity, from self-focused, to role- focused, to individuated-connected through questions addressing caring, commitment, and communication in close relationships (White, Speisman, Costos, Kelly, & Bartis, 1984);
ø Birthmother Interview: extensive set of questions dealing with experience in making the adoption plan and the current adoption situation, including relationships with her birthchild, the adoptive family, and the placing agency.
ø Self-Perception Profile for Adults: questionnaire including 12 subscales in which individuals make self-evaluations regarding competence and adequacy (Messer & Harter, 1986).
ø Health Checklist: checklist for common stress-related physical symptoms (Pennebaker, 1986). Twenty Statements Test: see above

Findings to Date

We have learned a number of important things from the first phase of our study. For example, when compared to parents in confidential adoptions, those in fully disclosed open adoptions generally reported higher levels of acknowledgment of the adoption, more empathy toward the birthparents and child, stronger sense of permanence in the relationship with their child as projected into the future, and less fear that the birthmother might try to reclaim her child. Despite these mean differences, variations within levels of openness were present (Grotevant, McRoy, Elde, & Fravel, Family Process, 1994, 33, 125-146).

We have also examined consequences of variations in openness for the children (Wrobel, Ayers-Lopez, Grotevant, McRoy, & Friedrick, in press). Comparison of parents' and children's reports of openness revealed important gaps between parents' participation in open arrangements and their inclusion of the adopted child in the communication. Although the children, as a group, scored within the normal range on Harter's self-esteem scale and indicated satisfaction with their current adoption arrangement, they also evinced curiosity about their birthfamilies. Almost half the children in mediated adoptions were excluded from contact their adoptive parents were having with their birthmother, but most of these children are not aware of their being excluded. Most of the children in fully disclosed adoptions were included in meetings with birthparents and were aware of the arrangements. Greater information about the adoption was related to higher cognitive levels of understanding about adoption in general. Results of the study were not compatible with speculations by critics of openness, who stated that such arrangements would damage children's self-esteem and cause them confusion. Self-esteem, satisfaction with openness, and curiosity about birthparents did not differ by level of openness. On the other hand, results did not support the alternative position that more openness would enhance these outcomes. Since many of the children were young when visited, a follow- up at adolescence will yield important information about longer-term openness effects.

Another study (Ross, 1995) focused on familial predictors of outcomes (externalizing problems, internalizing problems, self-esteem) for the adopted children in the study. In this non-clinical sample, incompatibility of the child with the family (as perceived by parents) emerged as a strong predictor of both perceived externalizing and internalizing difficulties.

In another study (Mendenhall, Grotevant, & McRoy, Family Relations, 1996) we contrasted two groups: families with mediated adoptions who had had an opportunity for greater openness but chose to retain mediated openness, and families in which the mediated adoptions became fully disclosed. Adoptive couples' interviews were coded for communication facility. Couples increasing openness showed significantly higher levels of self-disclosure, listening skill, empathy, continuity/tracking, respect / regard, and global communication facility than did couples in the other group. We cannot infer that the better communication skills caused them to increase levels of openness, but it may be that their attempts to increase openness were facilitated by their communication skill.

Relationships between the adoptive family and birthmother are also being examined through the lens of boundary ambiguity, which is said to exacerbate family stress because of family members' inability to determine who is inside and who is outside the family system (Boss, 1988). Boundary ambiguity occurs when a family member is physically absent but psychologically present, or vice versa. Deborah Fravel (1995) has examined the psychological presence of the birthmother in the adoptive family system. Tentative conclusions indicate that boundary ambiguity is almost inevitable in adoptive families but that it manifests itself differently by level of openness. Management of the psychological presence of the birthmother may also vary according to both level of openness and some personality characteristics and relationship tendencies of the adoptive parents such as tolerance for ambiguity (Fravel, Grotevant, Boss, & McRoy, 1993; Fravel, 1995).

Karen Schmid (1994) examined relationships between adoptive grandparents and their adult children, as perceived by the adoptive parents. This study focused on the way adoptive parents interpret their parents' reactions to and feelings about adoption in general, and open adoption in particular; how their relationships affect those in the nuclear family; and the consequences of such variations for initiating or maintaining contact between the families by birth and adoption. Most grandparents were little involved in couples' decisions to adopt but were generally supportive. However, many of them were guarded about couples' choice of openness, and adoptive parents had to negotiate these relationships with their parents at the same time they were establishing relational ties with their new child and the child's birthparents.

Finally, Harold Grotevant has also been involved with the Family Story Collaborative Project, an interdisciplinary multi-site collaboration. The team's goal has been to understand how family narratives provide a window into families' construction of meaning about their relationships and how such constructions are related to developmental outcomes for children and qualities of the marital relationship. In the lives of adults, narratives reflect the meaning constructed as they become parents, whether by giving birth or by adopting a child. In adoption, one takes on a new and usually unanticipated role that involves constructing a narrative that somehow explains, accounts for, or justifies this new status. If things are going well, narratives lend a sense of coherence and meaning to the experience of individuals constructing them. Such outcomes are presumably associated with psychological well-being and resilience. The adoption component of the collaboration has examined links between narrative coherence and relationship style of adoptive couples and then examines their relation to the couple's marital satisfaction and the social competence of their adopted child.

The FSCP rating scales were applied to 81 interviews from families randomly selected from the larger Openness in Adoption Research Project sample. Participating families had children between the ages of 4 - 8 and were equally distributed in number across the 3 levels of openness in their adoption: confidential, mediated, and fully disclosed open. The analyses involved 3 interviews per each of 27 families: an individual interview with the adoptive mother, an individual interview with the adoptive father, and a joint interview with the adoptive couple.
The interviews conducted with adoptive parents of more open adoptions demonstrated greater narrative coherence in both individual and couple contexts and for both husbands and wives than the interviews of adoptive parents in confidential or mediated adoptions. For both wives and husbands, coherence ratings made from their individual and couple interviews were moderately and positively correlated. However, spousal correlations on the coherence scales were generally higher for the couple interviews than for the individual interviews. It appears that the "jointness" of the couple interview contributed to greater spousal similarity in that setting. This was expected, given the volunteer nature of this non-clinical sample. Coherence ratings made from the conjoint couple interviews were highly correlated with ratings of the interaction itself. More coherence displayed by each partner in the conjoint narrative was related positively and significantly to the marital partners' greater confirmation of one another in the interview, greater collaborative style in their account, and more warm and inviting interaction with the interviewer. On the other hand, coherence ratings made from the individual interviews were moderately related to ratings of the interaction for wives, but not for husbands. This points to the stronger role that wives might play in setting the tone for the dyadic conversation or at least to the stronger continuity that might exist between wives' behavior in individual and couple interviews. Differences between an individual's narrative coherence in the individual and couple interview contexts were examined as an aspect of the interaction. We assumed that greater coherence rated in the individual than in the couple interview indicated that the individual was being "pulled down" by the spouse when they were together and that such a situation would predict greater dissatisfaction with the marriage. This hypothesis was upheld for wives, but not for husbands. Finally, we predicted that spousal asymmetries in coherence within the couple interview would indicate that one partner was carrying the other partner conversationally or scaffolding his or her participation. Larger asymmetries were related behaviorally to lower couple co-construction, less confirmation of one spouse by the other, and greater marital dissatisfaction of the partner doing the carrying (Grotevant, Fravel, Gorall, & Piper, 1995).

Significance of the Study

The study is significant because it is national in scope; involves a sample (720 individuals) much larger than other adoption studies that employ interview methods and home visits; includes the full range of adoptive openness, including cases in which contact has stopped, some in which contact continues, and others in which it has increased or decreased over time, allowing for tracking of trajectories of openness over time; includes data from all triad members; includes data about adoption agency practices and policies to contextualize the work; links with the investigators' earlier retrospective investigation of emotional disturbance in adopted adolescents, now providing prospective data which can be examined longitudinally to test tentative conclusions drawn earlier; and has significant implications for legal and policy issues concerning adoption in the United States today. Progress toward such understanding will contribute to better theories about human development within the family, better methodology for studying complex living arrangements, a better knowledge base for prevention and intervention programs, and guidance for the establishment of policies that address the "best interests of the child" in cases of adoption.

Funding Sources

We acknowledge with appreciation grant funding from the William T. Grant Foundation, the Office of Population Affairs (U. S. Department of Health and Human Services), the Hogg Foundation for Mental Health, the National Institute of Child Health and Human Development, the John D. and Catherine T. MacArthur Foundation, the Minnesota Agricultural Experiment Station, and the University Research Institute of the University of Texas at Austin.

Follow-Up Study

With funding from the William T. Grant Foundation, the Hogg Foundation for Mental Health, the Federal Office of Population Affairs, the University Research Institute of the University of Texas, the Minnesota Agricultural Experiment Station, and the Silberman Fund, we have now interviewed our families and birthmothers once again. At time 2 (interviews conducted between 1996 and 2000), we re-visited the adoptive families in their homes and conducted telephone interviews with birthmothers. As with Time 1, we conducted individual interviews with each parent, target adolescent, and birthmother, administered a number of standard questionnaires, and engaged the adoptive parents and adolescent in an audiotaped decision-making task.
The focus of the follow-up study continues to be on individual and relationship development in adoptive family systems. The research team at the University of Minnesota continues to focus on the adoptive families, and the team at the University of Texas focuses on the birthmothers, although it should be noted that the data will be examined at the system level as well.

The children in the study who were first seen at age 5-12 are adolescents (ages 12 - 18) at the time of follow-up. In this age group, important developmental issues emerge. Because of their adoptive status, we are particularly interested in adolescents' emerging sense of identity (e.g., Erikson, 1968). Much of the identity literature (see Marcia, 1980, for review) has focused on domains of identity that involve choices, such as selection of a career or a political ideology. Being an adopted person, however, was not the choice of the adolescent. We are very interested in how one's understanding of his or her adoptive status contextualizes other aspects of identity and self-development (Grotevant, 1992, 1993, 1994).

Adolescent adjustment is also being examined. Prior research suggests that adopted children are referred for psychological treatment two to five times more frequently than their non-adopted peers. This finding has been replicated in a number of western countries (see McRoy, Grotevant, & Zurcher, 1988, for review). Many of these young people, like the children in our study, were placed as infants with childless couples. Despite the good prognosis, the emerging problem behaviors typically included aggressiveness, acting-out behavior, and impulsiveness, especially in boys (e.g., Simon & Senturia, 1966). The follow-up study provides a unique opportunity to test hypotheses about the development of problem behavior in adopted adolescents with prospectively collected data. The data collected at time 1, including the Child Adaptive Behavior Inventory, Harter self-esteem measure, and parental assessments of child adjustment from the parent interview provide a strong foundation from which to assess adolescent functioning.

The follow-up study also provides an opportunity to examine parent-adolescent relationships. This will be the first study to examine the interplay between adolescent identity issues around adoption and parent-adolescent relationships. The author's model of individuality and connectedness in parent-adolescent relationships (Grotevant & Cooper, 1985, 1986) will provide a framework for investigation of this topic. Since individuality and connectedness in the marital relationships of the adoptive parents were assessed at time 1 (Bengtson & Grotevant, 1996), these data will permit examination of longitudinal links between family relationships during the childhood years and subsequent parent- adolescent relationships.

All of these individual and relationship variables will be considered within the broader context of the lives of these families. What changes have occurred in the level of openness in the family's adoption, if any? Has family composition changed since time 1 through parental divorce, death, or birth of new children? In what ways have the child and the family been involved with the extended family of the birthparents, if any? The data in this study will provide a unique opportunity to answer these questions with a large national sample.

Outreach and Dissemination Efforts

The world wide web now provides the means for rapid dissemination of research based information to a global audience. For the past several years, Hal Grotevant has been involved in outreach activities through the Children, Youth, and Family Consortium, a major interdisciplinary, collaborative effort linking the University of Minnesota with communities beyond it. With the assistance of our research team, we have helped develop AdoptINFO, internet-based resources on adoption. One of the sections on this web site offers the latest research on openness in adoption, and our most current research findings are posted there for anyone in the world with internet access to review.


Harold D. Grotevant, Ph.D.
Department of Family Social Science
University of Minnesota
1985 Buford Avenue
St. Paul, MN 55108
phone: (612) 624-3756
fax: (612) 625-4227
e-mail: hgrotevant@che.umn.edu


Ruth G. McRoy, Ph.D.
Piester Centennial Professor
School of Social Work
University of Texas at Austin
1925 San Jacinto Blvd.
Austin, TX 78712
phone: (512) 471-0551
fax: (512) 471-9514
e-mail: r.mcroy@mail.utexas.edu

- back -