The Effect of Family Communication Patterns on Adopted Adolescent Adjustment

Author(s): Martha A. Rueter and Ascan F. Koerner

Source: Journal of Marriage and Family , Aug., 2008, Vol. 70, No. 3 (Aug., 2008), pp.
715-727
Published by: National Council on Family Relations

Stable URL: https://www.jstor.org/stable/40056362

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

National Council on Family Relations is collaborating with JSTOR to digitize, preserve and
extend access to Journal of Marriage and Family

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms
 MARTHA A. RUETER University of Minnesota

ASCAN F. KOERNER University of Minnesota*

The Effect of Family Communication Patterns on

Adopted Adolescent Adjustment

Adoption and family communication both affect complexity of family structures, such as step-
adolescent adjustment. We proposed that adop- families, families formed through assisted
tion status and family communication interact reproduction, and adoptive families. Recent re-
such that adopted adolescents in families with views attest to particular interest in adoptive fam-
certain communication patterns are at greater ilies and in adopted child adjustment (cf. Bimmel,
risk for adjustment problems. We tested this Juffer, van IJzendoorn, & Bakermans-Kranen-
hypothesis using a community-based sample of burg, 2003; Juffer & van IJzendoorn, 2005;
384 adoptive and 208 nonadoptive families. Lee, 2003; O'Brien & Zamostny, 2003; van IJ-
Adolescents in these families were, on average, zendoorn, Juffer, & Klein Poelhuis, 2005). These
16 years of age. The results supported our reviews compared adopted, nonadopted, domesti-
hypothesis. Adopted adolescents were at signifi- cally adopted, and internationally adopted youth
cantly greater risk for adjustment problems on several adjustment dimensions, including inter-
compared to nonadopted adolescents in fami- nalizing and externalizing problems, attachment to
lies that emphasized conformity orientation parents, and academic achievement. Overall, these
without conversation orientation and in families reviews reported that most adopted children and
that emphasized neither conformity nor conver- adolescents were well adjusted. A small but nota-
sation orientation. Adolescents in families ble group, however, experienced significant
emphasizing conversation orientation were at behavioral or mental health problems. It is this
lower risk for adjustment problems, regardless group that may account for mean differences in
of adoption status. adjustment that often are observed in studies com-
paring adopted to biological children (Bimmel
et al.; Brand & Brinich, 1999).
Recent changes in the modern family have led re-
Differences in adjustment for this small group
searchers to pay closer attention to the growing
have generally been attributed to a number of
factors unique to adopted children. For example,
relative to nonadoptees, adopted children have
Department of Family Social Science, 290 McNeal Hall, more likely experienced early childhood adver-
1985 Buford Avenue, University of Minnesota, Saint Paul, sity that can result in developmental delays and
MN 55108 ([email protected]). negatively affect early childhood attachment to
*Department of Communication Studies, 244 Ford Hall, parents (Haugaard & Hazan, 2003). Also, the
224 Church St. S.E., University of Minnesota, Minneapolis identity development process can be particularly
MN 55455.
challenging for adopted youth, who may look and
This article was edited by Cheryl Buehler. act differently from their parents and siblings and
who may be
Key Words: adjustment, adolescents, adoption, family com- trying to come to terms with limited
munication patterns. information about their birth parents and cultural

Journal of Marriage and Family 70 (August 2008): 7 1 5-727 7 1 5

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms
 7 1 6 Journal of Marriage and Family

origins (Brodzinsky, Schechter, adjustment are, compared & toHenig, 1992;

family interaction,
Lee, 2003). In regard to more distal factors health
mental whose impact on adjustment
outcomes,
there also might be differences is moderated by infamily interaction. That thresh-
parental is, adop-
olds for making treatment tion and its correlates definewith
referrals, a particular context
adoptive
parents more likely than that interacts with family interaction
nonadoptive processes
parents to
refer children for mental health or behavioral to determine child adjustment.
problems (Juffer & IJzendoorn, 2005).
These factors, however, do not fully explain
Family Communication Patterns Theory
the adjustment difficulties observed in some
adopted children. First, they do not apply uni- A theoretical framework that expands upon exist-
formly to all adoptive families nor to all adopted ing theories (e.g., Baumrind, 1971; Burleson
children in the small group with adjustment prob- et al., 1995; Reiss, 1981) to provide a stronger
lems. Second, the external factors described explanation of the association between family in-
above suggest fairly direct cause-effect relation- teractions and child adjustment in complex fami-
ships. Such simplistic associations are unlikely lies like adoptive families is Koerner and
to represent the complex causal processes that Fitzpatrick's (2002a, 2002b, 2004, 2006) Family
underlie adopted children's adjustment prob- Communication Patterns Theory (FCPT). FCPT
lems. To better understand adjustment among is based on the fundamental insight that creating
adopted children, we need a more thorough a shared social reality is central to family func-
understanding of the complex underlying pro- tioning. Shared reality exists when family mem-
cesses as they occur in most, if not all, adoptive bers' cognitions about an object are accurate,
families. congruent, and in agreement. Sharing social real-
ity with others makes understanding and being
understood easier, leading to more efficiency
Adolescent Adjustment and
and coordination and fewer misunderstandings
Family Communication
and conflict. Consequently, families that share
In general population studies, more than three social reality should communicate with one
decades of research has established a strong asso- another more accurately and with less conflict,
ciation between parent-child interactions and supporting child adjustment.
adolescent adjustment (Reiss, 2000; Steinberg, According to FCP Theory, families create
2001). Research on parent-child communication a shared reality through two processes, conversa-
has consistently demonstrated that parent-child tion orientation and conformity orientation.
interactions characterized by open communi- Conversation orientation is characterized by
cation, warm and supportive behavior, and firm, frequent, spontaneous, unconstrained interac-
consistent enforcement of developmentally appro- tions that allow family members to codiscover
priate expectations positively influence child the meaning of symbols and objects. This orien-
adjustment. Hostile, angry, and conflictual interac- tation encourages all family members to partici-
tions, on the other hand, are associated with poorer pate in defining social reality. Conformity
adjustment. Various labels have been employed to orientation is characterized by uniformity of be-
describe these different types of parenting, includ- liefs and attitudes. Family interactions focus on
ing Baumrind's (1971) authoritarian, authorita- maintaining harmonious relationships that
tive, permissive, and neglecting parenting, reflect obedience to parents, often manifest in
Burleson, Delia, and Applegate's (1995) person- pressure to agree and maintain the family hier-
versus position-centered parenting, and Koemer archy. This orientation allows family members
and Fitzpatrick's (2002b) conversation orientation in authority roles (i.e., parents) to define social
and conformity orientation. reality.
Drawing from this overwhelming evidence, Theoretically orthogonal, these two orienta-
we expect that parent-child interaction plays tions define four family types: consensual, plural-
a similarly relevant role in adopted children's istic, protective, and laissez-faire. Consensual
adjustment. We argue that family interaction is families are high in both conformity and con-
a proximate influence on child and adolescent versation orientation. Communication in con-
adjustment, regardless of adoption status. Fur- sensual families reflects a tension between
ther, family structure and the factors already exploring ideas through open communicative
identified as associated with adopted children's exchanges and a pressure to agree in support of

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms
 Family Communication Patterns 717

the existing family hierarchy. from taste for sweets,Pluralistic

preferences for leisure fami-
lies are low in conformity orientation
activities, endorsement and
of moral and ethical posi- high
in conversation orientation. tions, and political attitudes (Alford,
Family Funk, &
communica-
tion is characterized Hibbing, as open and
2005; Olson, unrestrained,
Vernon, Harris, & Jang,
focusing on producing 2001; Tesser, 1993). Abrahamson,ideas
independent Baker, andand
fostering communication Caspi (2002) have shown that these
competence ineffects are
children.
Protective families are nothigh
limited toin conformity
adults. orien-
They reported significant
tation and low in conversation orientation. genetic effects on political attitudes in children
Communication in these families functions to as young as 12 years old.
maintain obedience and enforce family norms; This research suggests that although geneti-
little value is placed on the exchange of ideas
cally related family members can sometimes
or the development of communication skills.rely on similar cognitive processes to achieve
Laissez-faire families are low in both conversa-
a shared reality, genetically unrelated family
tion orientation and conformity orientation.members must rely on other processes. We and
Family members do not often engage each other others (Brodzinsky et al., 1995; Grotevant
in conversation, and they place little value et al., 2001; Stein & Hoopes, 1985) suggest that
on communication or the maintenance of a how family members communicate with one
family unit. another is particularly important to creating
a shared social reality among adoptive family
members.
Sharing Reality in Complex Families

The concept of a shared reality among family

members is not new. Others describe similar con-
Adoption, Family Communication Patterns,
and Child Adjustment
cepts using similar terms. Reiss (1981) described
shared reality as a family paradigm guiding how To date, studies of adoptive family communica-
members respond to challenges from the exter- tion mostly have examined adoption-specific com-
nal world and Eccles et al. (1993) used stage- munication (e.g., parents talking with an adopted
environment fit theory to explain the importance child about his or her adoption; Brodzinsky,
2006; Wrobel, Kohler, Grotevant, & McRoy,
of compatibility between parental control at-
tempts and adolescents' growing desire for 2003) or examined the direct effect of adoption
autonomy. Deater-Deckard and Petrill (2004) status on family communication. Direct-effect
used dyadic mutuality to describe synchronized, studies have compared various aspects of commu-
mutually warm, and responsive parent-child in- nication (e.g., levels of conflict, amount of verbal
teractions and Grotevant, Wrobel, van Dulmen, interaction) across adoptive and nonadoptive fam-
and McRoy (2001), referring specifically to ilies (Lansford, Ceballo, Abby, & Stewart, 2001;
adoptive families, used parent-child compatibil- Lanz, Ifrate, Rosnati, & Scabini, 1999; Rosnati,
ity or goodness of fit to refer to the similarity & Marta; 1997). For the most part, these studies
between parental expectations and actual or per- reported few differences in communication on
ceived child behavior. The connection between the basis of adoption status.
these conceptualizations of shared reality and In contrast to direct-effect studies, the FCPT
FCP Theory is that in each case, increased sharedsuggests that adoption status and communication
reality is expected to relate to improved family pattern interact to influence child adjustment. On
functioning or child adjustment or both. the basis of research of parent-child communica-
Several sources suggest that, compared to
tion in the general population (Baumrind, 1971;
Burleson et al., 1995; Koerner & Fitzpatrick,
genetically related families, sharing social reality
is likely to be more challenging in adoptive fam- 2004; Steinberg, 2001), we expect that family
ilies (Brodzinsky, Lang, & Smith, 1995; Deater-communication patterns directly affect child
Deckard & Petrill, 2004; Grotevant et al., 2001).adjustment. On the basis of the greater chal-
Among the possible reasons for the added chal- lenges to creating a shared reality among adop-
lenge is that the cognitive processes involved intive family members, we also expect that in
adoptive families the effects of family communi-
perceiving the social world are at least partially
a function of genetic predispositions. Research cation on adjustment will be amplified in specific
supporting this contention has shown medium ways. The purpose of the current study, then, was
to large effects of genetics on attitudes rangingto test the application of the FCP Theory to

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms
 718 Journal of Marriage and Family

explain adolescent adjustment among

been placed for adoption prior adopted
to 2 years of age
(M = 4.7 months, SD =this,
adolescents. To accomplish 3.4 months). we tested a
series of hypotheses: Participating were 63% of the eligible adoptive
families and 57% of the eligible biological fami-
HI: Adopted children lies.
will have
To determine more adjust-
the representativeness of par-
ment problems than nonadopted
ticipating families, a children.
brief phone interview
H2: Adoption status is not associated with a
assessing parents' education, occupational status,
family's FCP.
H3: FCP is associated with child adjustment. marital status, and the number of parent-reported
H3a: Consensual families will experience the behavioral disorders in the participating children
fewest, Laissez-Faire families the most, and was administered to 73% of nonparticipating but
Protective and Pluralistic families a moder-
eligible families. Results showed that the study
ate level of child adjustment problems.
H4: Adoption status and FCP interact to influ- sample is generally representative of the popula-
ence child adjustment. tion of eligible families from which it was drawn
H4a: FCPs that favor conversation orientation and is not markedly different from families with
(Consensual and Pluralistic) will have sim- parents living with two or more children in the
ilar levels of adjustment problems across
metropolitan region where the university is
adoptive and nonadoptive families.
H4b: FCPs that favor control over conversa- located (McGue et al., 2007).
tion (Protective) or use neither orientation
(Laissez-Faire) will show more child ad-
Procedures
justment problems in adoptive families
relative to nonadoptive families.
Participating family members visited the research
lab to complete informed consent forms, self-
Method report surveys, two 5-minute videotaped family
interactions, and the revised Diagnostic Inter-
view for Children and Adolescents (DICA-R)
Sample
(Welner, Reich, Herjanic, Jung, & Amado, 1987).
Participants were 592 families recruited to a lon- Self-report surveys were independently com-
gitudinal research project designed to investigate pleted by each family member. Among other
sibling influences on adolescent drug and alcohol things, these surveys assessed adolescent exter-
use (McGue et al., 2007). All study families nalizing behavior and family and individual
included two parents, the target child (referred demographic characteristics. The videotaped
to as the adolescent; M age = 16.01 years, family interactions were designed to elicit family
SD = 1 .44), and a younger sibling (referred to interactions, including conversation and control
as the sibling; M age = 13.69 years, SD = 1.57) behaviors. Videotaping took place in a room dec-
who was within 5 years of the adolescent's age. orated to look like a living room or dining room,
In 284 families, both children were adopted, in with family members seated around a dining
100 families, the adolescent was adopted and table. Although the video camera was inconspic-
the sibling was biologically related to the pa- uously placed in a bookcase, family members
rents, and in 208 families both children were were aware that they were being videotaped.
biologically related to the parents. A trained interviewer explained the tasks to the
Adoptive families were identified through re- family members, but left the room for videotaping.
cords from three large adoption agencies (600 For the first task, families were presented with
and 700 placements each year). Biological fami- a novel object, a Rorschach inkblot, and asked to
lies were identified using state birth records. come to a consensus about what the inkblot resem-
Researchers located 90% of the identified adop- bled. For the second task, families were presented
tive families and 85% of the identified biological with a moral dilemma (Kohlberg, 1981). In the
families. Once located, a parent in each family story, a man whose wife has been diagnosed with
was interviewed to establish study eligibility. In a fatal disease but cannot afford to buy the only
addition to the children's age requirement, study drug that can save her life. Families were asked
eligibility was limited to families living within to decide (a) whether the man should steal the drug
driving distance of the research lab and to chil- for his wife and (b) whether he should also steal the
dren with no physical or mental handicap that drug for a stranger in need.
would preclude completing the day-long intake Trained interviewers administered the DICA-
assessment, and all adopted children had to have R (Welner et al., 1987) to the adolescents and

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms
 Family Communication Patterns 719

their mothers. The DICA-R had been modified to ond observer, and then comparing the primary
include additional questions and probes neces-and secondary ratings using intraclass correla-
sary for complete coverage of DSM-IV child- tions (Shrout & Fleiss, 1979; Suen & Ary,
hood disorders. Adolescents' symptoms were1989). Intraclass correlations for scales used in
reported by themselves and by their mothers.this study ranged from .5 to .8, a level of reli-
All interview data were reviewed by at least ability considered acceptable for these types of
two individuals with advanced clinical trainingdata (Kenny, 1991; Mitchell, 1979).
who were blind to other family members' symp- The present study used three observational
toms and diagnoses. These reviewers codedscales to assess conversation orientation, Com-
every symptom and diagnostic criterion. Amunication, Listening, and Warmth. Because
symptom was considered present if either the observers rated a family member's behavior
adolescent or the mother reported it. Kappa co-toward each of the other three family members,
efficients for disorders are as follows: Atten- every family member received three scores for
tion deficit hyperactivity disorder (ADHD, .77),each scale. For example, using the Communica-
Conduct Disorder (CD, .80), and Oppositional tion scale, observers rated the mother's ability
Defiant Disorder (ODD, .73). to clearly and appropriately express her own
During their visit, the adolescents also nomi-
point of view, needs, and desires when speaking
nated teachers to provide information about the to the father, to the adolescent, and to the sib-
child's behavior at school. Nominated teachers ling. Family members who expressed their
views in a manner that encouraged conversation
were mailed a rating form, and teacher reports
were received for 69% of the adolescents. Partici- with other family members received higher
pants were compensated for their travel expenses scores than those who did not. The Listening
and given a modest honorarium as compensation scale assessed the extent to which a family
for their time. member verbally or nonverbally or both ver-
bally and nonverbally attended to each of the
other family members when the other member
Measures
was speaking. Here again, each family member
received three Listening scores. The Warmth
FCP. A family's communication pattern is deter-
scale assessed each family member's verbal
mined by observing the extent to which the fam-
ily relies on conversation orientation and
and nonverbal expressions of caring, concern,
and support toward each of the other family
conformity orientation to create a shared reality.
members, for a total of three Warmth ratings
We used Latent Class Analysis to estimate each
family's most likely communication pattern per family member. The Control scale was used
(see Analysis Plan, below). The measures usedto assess conformity orientation. This scale
measured the extent to which a family member
to assess conversation orientation and conformity
orientation, which are described below, were attempted or succeeded in controlling or influ-
used as indicators of a FCP latent factor. encing the attitudes, behavior, and interactions
Trained observers viewed the two family inter- of other family members.
action tasks and globally rated 12 family interac-
tion characteristics using the Sibling Interaction Adolescent externalizing behavior. Adolescent
and Behavior Study Rating Scales, adapted adjustment problems were operationalized as
from the Iowa Family Interaction Rating Scales externalizing behavior in a variety of contexts
(Melby et al., 1998). Each family member's including general delinquency, symptoms of
behavior toward each of the other family mem- behavioral disorders, conflictual relations with
bers was rated using a scale ranging from 1 {not parents, and trouble at school. To obtain this
at all characteristic of the person) to 9 (mainly broad assessment of adolescent externalizing
characteristic of the person). Before viewing behavior, we used five measures derived from
study videotapes, observers received 100 hours multiple information that were combined as
of training and were required to pass written a latent factor with five indicators. Because we
and observation examinations. Trained observ- were primarily interested in the small subset of
ers attended biweekly coder meetings for ongo- adolescents who experience significant adjust-
ing training and to prevent "rater drift." ment problems as compared to adolescents who
Observer reliability was assessed by randomly experience relatively few problems, we used
assigning 25% of all tapes to be rated by a sec- Latent Class Analysis (see Analysis Plan) to

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms
 720 Journal of Marriage and Family

identify two groups differing in Mplus

using the statistical program externalizing
4.21 (Muthen
characteristics. &Muthen, 1998-2006).
For the first indicator, we used the Delinquent The FCP LCA model was created using a sec-
Behavior Inventory (DBI; Gibson, 1967). This ond-order latent factor structure. A set of 16
self-report questionnaire contains a list of 36 be- first-order latent factors, each with three indica-
haviors. For each behavior, adolescents reported tors, served as indicators of the second order
if they had never (1), once (2), or more than once FCP latent factor. The 16 first-order latent
(3) engaged in the behavior. Example DBI items factors assessed each family members' interac-
included "smashing, slashing, or damaging tion with the three other family members for
things," "cutting classes at school," "stealing the three conversation orientation measures
things, " and "using any kind of weapon in a fight. " and the one control orientation measure (4 fam-
DBI responses were summed to create a self-report ily members X 4 measures). For example, the
externalizing behavior measure (a = .89). mother's Communication factor assessed her
Symptom counts obtained from the ADHD, communication to the other family members
CD, and ODD sections of the DICA-R (Welner and was indicated by the observer ratings of
et al., 1987) were used to create an externalizing her communication to the father, to the adoles-
symptoms measure (range = 0-28 symptoms). cent, and to the sibling. The adolescent's and
As described above, adolescents and mothers the sibling's gender were entered as covariates
completed the DICA-R. A symptom was con- of the FCP latent factor.
sidered present if either the adolescent or the The adolescent's Externalizing Behavior
mother reported it. latent factor had five observed variables as indica-
Trained observers rated adolescent behavior tors: (1) self-reported delinquency, (2) externaliz-
toward each parent, as described above, to create ing disorder symptoms, (3) observed hostility to
the third and fourth externalizing behavior meas- the mother, (4) observed hostility to the father,
ures. Using the hostility scale, observers assessedand (5) teacher ratings. The adolescent's age
the extent to which the adolescent's behavior and gender were entered as covariates of the
toward the mother and toward the father was Externalizing Behavior latent factor.
characterized by conflict, anger, defiance, and We had hypothesized the presence of four
contempt. FCP and two Externalizing Behavior classes.
Teacher ratings of adolescent in-class behavior To be confident that these were the most likely
were used to create the final externalizing mea- number of classes, we tested LCA models that
sure. Using a 67-item behavior checklist adapted had fewer and more classes than the hypothesized
from the Conners' Teacher Rating Scale number. Because no single criterion is yet
(Conners, 1969) and the Rutter Child Scale B accepted for deciding the most likely number of
(Rutter, 1967), teachers compared the adolescent classes within a population, we used a combina-
to the average student and rated how characteris- tion of theoretical and statistical criteria. First,
tic a behavior was of the adolescent (1 = not we relied upon theory to provide the starting point
at all characteristic to 4 = very much charac- for our model tests. Thus, to create the FCP vari-
teristic). Example checklist items included "is able, we tested models specifying one, two, three,
four, and five classes. For the Externalizing
defiant," "has difficulty concentrating on school-
work," "is often truant," "initiates physical Behavior variable, we tested one, two, and three
fights," and "obeys the rules" (reverse coded). classes. Statistical criteria included the Bayesian
Responses were summed (a = .97, Spearman- information criterion (BIC; Hagenaars &
Brown interteacher reliability = .82). McCutcheon, 2002) and the Lo-Mendell-Rubin
adjusted LRT (LMR\ Lo, Mendell, & Rubin,
2001). The BIC is a measure of model fit based
Analysis Plan
on the -2 log likelihood statistic with a penalty
Testing our study hypotheses required that we for small samples and increasing parameters. A
develop two categorical latent variables, the large decrease in the BIC value when the num-
FCP variable and the Adolescent Externalizing ber of classes is increased indicates an improved
Behavior variable, and examine associations fit for the model specifying the additional class.
between these two variables and adoption status. The LMR tests the null hypothesis that re-
Both categorical latent variables were created verting to a model with one less class than spec-
through Latent Class Analysis (LCA) performed ified would improve model fit. A statistically

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms
 Family Communication Patterns 721

significant LMR suggests To test H4, we estimated the

that models used
this to
hypothesis
can be rejected and that test H3 twothe model
more times, being
once using the sample tested
produced a significantof improvement adoptive families and a second time in
using the
model fit
relative to a model with one less class. We also sample of nonadoptive families. Thus, we ob-
considered class sizes and model convergence. tained the proportion of adopted and nonadopted
Models that produced classes with few oradolescents
no estimated to be in the high externaliz-
members or that did not converge were rejected.
ing subgroup for each family communication pat-
tern. Proportions were statistically compared
For each of our study hypotheses, we estimated
the probability that an adolescent wouldusing be Fisher's exact test. H4 would be supported
placed in the high externalizing subgroup on if adoptive Laissez Faire and Protective families
the basis of family communication patternshad or significantly higher proportions than nona-
adoption status or both. All probabilities were doptive Laissez Faire and Protective families
and adoptive and nonadoptive Consensual and
calculated as posterior probabilities, and all ana-
lytical models were run as mixture models using Pluralistic families had similar proportions.
Mplus 4.21 (Muthen & Muthen, 1998 - 2006).
The following provides a description of how we
Missing Values Analyses
tested each of the study hypotheses.
Testing HI required that we regress the two- Data from 592 families were available for these
class Externalizing Behavior latent variable analyses,
on 3 1 8 of which had complete data on all
adoption status and two covariates, adolescent's study variables. Almost all missing data were
age and sex using logistic regression. HI would due to missing teacher reports or fathers who
be supported if adopted adolescents had signifi- did not participate in the observation tasks. As
cantly greater odds of being placed in the high noted above, 31% of the teacher externalizing
externalizing subgroup relative to nonadopted behavior ratings were missing. Also, in 23% of
adolescents. the families, fathers did not participate in the
To test H2, we estimated the proportion ofobservational tasks. All other study variables
adoptive and nonadoptive families for eachhad no more than 3% missing data.
FCP class and statistically compared the adop- Current research indicates that when missing
tive and nonadoptive pairs of proportions usingdata are unrelated to the study outcome (i.e.,
Fisher's exact tests. H2 would be supported ifmissing at random), recovering missing data
the tests showed that adoptive and nonadoptive using a reliable estimation procedure is preferable
families were distributed similarly across familyto case deletion (Schafer & Graham, 2002). For
communication patterns. Additionally, we re-each externalizing behavior measure, we com-
pared mean values for adolescents whose father
gressed the FCP latent variable on adoption status
and two covariates, adolescent and sibling gen-did and did not participate in the observational
der, using multinomial logistic regression with tasks or who did and did not have teacher report
the Laissez Faire family communication patterndata. T test results showed no statistically signifi-
as the reference group. H2 would be supportedcant differences on the basis of father participa-
by this test if adoptive and nonadoptive familiestion. Adolescents without teacher report data,
had even odds of placement within each family however, did report significantly higher exter-
communication pattern. nalizing behavior (t = 3.14, p - .002) and
Testing H3 required that we estimate the pro- externalizing symptoms (t = 4.15, p < .00). To
portion of adolescents in the high versus theexamine the possibility that our results could be
low externalizing subgroups for each family biased by missing data, we tested each study
communication pattern. Proportions were com- hypothesis with and without listwise deletion of
pared statistically using Chi Square and Fisher'smissing data. For every hypothesis, the pattern
exact test. H3 would be supported if Consensual of findings was similar, although the smaller
families had the smallest proportion of ado- sample produced fewer statistically significant
lescents placed within the high externalizing results.
subgroup, the Protective and Pluralistic fami- Mplus handles missing data by adjusting
lies had similar, midlevel proportions, and the model parameter estimates using full-information
Laissez Faire families had the largest proportionmaximum-likelihood estimation (FIML; Muthen
of adolescents placed in the high externalizing & Shedden, 1999; Schafer & Graham, 2002). To
subgroup. obtain reliable estimates, Mplus requires that the

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms
 722 Journal of Marriage and Family

pattern.
proportion of available data The second
for bar in every setstudy
each depicts the vari-
able and between each father's
pair mean.
of Thevariables
third bar is the adolescent's
be at least
.10. These proportionsmean,
were and final bar in
all every set is
above the sibling's
.53 and the
mean factor
majority were above .97. score. (Standard errors
Therefore, we and used
t values the
FIML option to deal with missing
for the scores data.
presented in Figure 1 are available
upon request from the first author.)
RESULTS
As shown in Figure 1 , Consensual families
had two parents who were relatively high on
control behavior and all family members tended
Estimating FCP Classes and Externalizing
Behavior Classes to engage in high levels of communication, lis-
tening, and warmth. No one in the typical Plural-
LCA results produced the strongest support for istic family showed high control, and members
the four-class FCP model. The pattern of decline engaged in moderate levels of communication
in the BIC statistic supported the four-class and listening and relatively little warmth. Pro-
model over either the three- or five-class model. tective families had one controlling parent and
The class sizes estimated by the four-class engaged in relatively little communication
model (Consensual = 6.7%, Pluralistic = and moderate levels of listening and warmth.
31.8%, Protective = 21.9%, Laissez Faire = Finally, Laissez-Faire families consistently
39.6%) were the most evenly distributed of allengaged in the lowest levels of all measured
models tested, and most importantly, the pat-behaviors.
terns of family behavior estimated by the four- LCA estimation of adolescent externalizing
class model varied in theoretically expectedbehavior subgroups showed that a two-class
ways. We rejected the five-class LCA modelmodel fit the data best (one-class BIC =
because it estimated a class containing just 1%15709.66, two-class BIC = 12610.97, three-class
of the families and produced a relatively smallBIC = 12454.54; two-class LMR = 529.37, p <
drop in the BIC (four- to five-class BIC change =.00, three-class LMR = 240.15, p = .17). The
68.02) and a statistically insignificant LMRtwo-class model placed 79.9% of the adolescents
(LMR = 182.85, p = .14). The two-class modelin the low externalizing behavior subgroup and
was also rejected because the relative decrease 20.1% in the high externalizing subgroup.
in the BIC statistic from the one- to the two-
class model (BIC change = 1441.21) and the
Hypothesis Testing
LMR statistic (LMR = 952.55, p = .008) sup-
ported the presence of more than two classes.Logistic regression results showed that adopted
The three-class model produced a good fit (two-adolescents were more likely to be placed in the
to three-class BIC change = 258.94, LMR =high externalizing subgroup relative to nonadop-
376.76, p = .002). But three problems with thisted adolescents (odds ratio (OR) = 3.21, 95%
model led us to reject it. First, the mean familyConfidence Interval (CI) = 1.75 - 5.90), sup-
behaviors produced by this model showed fewporting HI. Adolescents' gender and age also
interpretable patterns. Second, the model pro-predicted externalizing subgroup placement.
duced an uneven class distribution of two quiteBoys (OR = 5.68, CI = 3.07 - 10.51) and older
large classes and one small class. Finally, the
adolescents (/? = 0.320, CI = 0.15 - 0.49) were
BIC declines substantially from the three- to themost likely to be placed in the high externaliz-
four-class model (three- to four-class BIC ing subgroup.
change = 122.69), suggesting the possibility of Proportions of adoptive and nonadoptive fam-
a fourth class. ilies within each FCP were quite similar (Consen-
Evidence of the extent to which the four-class sual: adoptive = 6.7%, nonadoptive = 3.8%;
model estimated the expected family communi- Pluralistic: adoptive = 31.3%, nonadoptive =
cation patterns is presented in Figure 1 . Each 30.0%; Protective: adoptive = 20.5%, nonadop-
bar in Figure 1 represents one family member's tive = 26.0%; Laissez Faire: adoptive = 41.1%,
mean factor score. The first bar in every set de- nonadoptive = 40.4%), supporting H2. Statisti-
picts the mother's mean factor score. Thus, the cal comparisons using Fisher's Exact tests found
left-most white bar represents the Control factor no statistically significant differences between
score mean of .38 estimated for mothers placed the proportions of adoptive and nonadoptive fam-
within the Protective family communication ilies within each FCP. Also, multinomial logistic

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms
 Family Communication Patterns 723

Figure l . Graphical Presentation of Mean Factor Scores for the First-Order Factor indicators of the
Family Communication Patterns Latent Variable.

Note: First bar in every set: mother's mean factor score. Second bar: father's mean factor score. Third
mean factor score. Fourth bar: sibling's mean factor score. Bars rising above 0 represent behavior levels
mean. Bars falling below 0 represent behavior levels below the overall mean.

regression results using Laissez-Faire as the of adopted and nonadopted ado-

Proportions
comparison showed that adoptive andlescents
nonadop-within each FCP in the high externaliz-
tive families had even odds of placement
ing subgroup
in followed the expected pattern
each family communication pattern (Consensual: (Consen- adoptive = 2.6%, nonadoptive =
sual OR = 1.64, 95% CI = 0.70 - 3.84; Plural- 0.0%; Pluralistic: adoptive = 16.7%, nonadop-
istic OR = 1.07, 95% CI = 0.66 - 1.72: tive = 12.3%; Protective: adoptive = 18.5%,
Protective OR = 0.76, 95% CI = 0.46 - 1 .25). nonadoptive = 4.1%; Laissez Faire: adoptive =
Across family communication patterns, 26.9%,
pro- nonadoptive = 7.8%), supporting H4.
portions of adolescents in the high externalizing
The nearly 5 : 1 difference in proportions for adop-
subgroup supported H3 (Consensual = 1.1,
tive and nonadoptive adolescents in Protective
Pluralistic = 16.0, Protective = 13.4, Laissez families was statistically significant (p = .047),
Faire = 21.3; / = 150.76, p < .01). Fisher as was the 3:1 ratio for Laissez Faire (p =
Exact tests showed that the proportion of high .005). Proportions of adoptive and nonadoptive
externalizing adolescents estimated for Laissez- adolescents in Pluralistic families were similar
Faire families was significantly larger than the (p - .36). Small cell size precluded comparing
proportion for Pluralistic families (p = .04). proportions in Consensual families.
The was no difference in proportions for Plural-
istic and Protective families (p = .11). Small DISCUSSION
cell size (only one Consensual family adoles-
cent was placed in the high externalizing sub- On the basis of what is known about associations
group) precluded comparing Protective and between family communication and adolescent
Consensual families. adjustment from existing studies (Steinberg,

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms
 724 Journal of Marriage and Family

2001), much of what we adjustment

report problemshere
were the isProtective
not andunex-
Laissez-Faire types.
pected. Our goal, however, was It is
tono surprise
apply that these
the FCP
Theory, which suggests that
communication creating
patterns shared
are associated with ado-
lescent adjustment
social reality among family problems. What we
members report that
plays a cen-
tral role in adolescent isadjustment,
new is that adoption statustoand family
furthering
commu-
our understanding of nication
adopted adolescent
patterns interact adjust-
such that adopted chil-
ment. Our results support dren in these thefamilies
FCPwere atTheory
substantially and
indicate that existinggreater theories based
risk for adjustment problemslargely
relative to on
families with genetically related
nonadopted children. parents
In fact, more than a quarter and
children may not completely apply
of adopted adolescents to families
in Laissez-Faire complex
families, like adoptive fellfamilies.
into the high externalizing subgroup com-
As others have reported, pared to onlywe8% offound
the nonadopted adolescents.
that adop-
tion status is associated with adolescent ad- This suggests that adopted children may be much
justment (Bimmel et al., 2003; Juffer &more vansensitive to the parental indifference and
IJzendoorn, 2005; Keyes, Sharma, Elkins, neglect typical of Laissez-Faire families than
Iacono, McGue, 2007; Lee, 2003; O'Brien & nonadopted children. We also found that control-
Zamostny, 2003; van IJzendoorn et al., 2005). ling parenting without communication is much
We also replicated early research showing thatmore detrimental to adopted children than to non-
family communication patterns directly relate toadopted children. Adopted children in Protective
adolescent adjustment such that children in fam-families were at almost five times the risk of being
ilies that emphasized a combination of conversa-placed in the high externalizing group compared
tion and conformity were least likely to haveto nonadopted children in Protective families.
adjustment problems (Steinberg, 2001). Our find-
ings go beyond previous work to show that adop-
Theory-Based Explanation of Results
tion status and family communication patterns
interact in important ways and better explainWe proposed that the interaction between adop-
adopted adolescent adjustment. Specifically,tion status and family communication pattern
adoptive families that emphasized conformityoccurs because adoptive families face more chal-
over conversation orientation (i.e., protective lenges to creating a shared reality than nonadop-
families) or that used neither conformity nor con-tive families. According to FCP Theory, the
versation orientation (i.e., laissez-faire families)existence of a shared reality means more accurate
either failed to mitigate the risks of adoption asso- communication and fewer misunderstandings
ciated with adolescent adjustment or even ampli- and conflict, reducing the risk of child adjustment
fied them. Adoptive families high in conversationproblems (Koerner & Fitzpatrick, 2004, 2006).
orientation (i.e., consensual and pluralistic fami-Genetically related family members likely share
lies) appeared to mitigate those risks to the extenta sense of belonging based on physical appear-
that their risk for child adjustment problems wasance, blood ties, and shared social attitudes
statistically undifferentiated from nonadoptiveor cognitions based in genetic inheritance
families. (Abrahamson et al., 2002; Alford et al., 2005;
This does not mean that conversation orienta- Olson et al., 2001 ; Tesser, 1993). All these shared
tion is universally positive for adolescent characteristics facilitate their ability to create
outcomes. Our results demonstrate that communi- a shared reality, even in the absence of conversa-
cation without control from parents leads to poor tion. Adoptive families typically do not share
child adjustment, regardless of adoption status. these advantages.
We estimated that 16.7% of adopted adolescent In Protective families, where the parent(s) dic-
and 12.3% of nonadopted adolescents stemming tate the social reality, we speculate that nonadop-
from Pluralistic families were in the externalizing ted adolescents likely share at least some of their
group, which for nonadopted adolescents was the parents' cognitions. Therefore, they might accept
highest proportion. Only when conversation ori- their parents' regulatory messages, even if they
entation was paired with parental control in the are offered without much opportunity for discus-
form of conformity orientation was conversation sion. Adopted adolescents probably have cogni-
orientation associated with superior outcomes. tive processes that differ from their parents.
Family communication patterns that placed Therefore, adopted adolescents in Protective
adoptive families at particular risk for adolescent families may find their parents' regulatory

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms
 Family Communication Patterns 725

messages more difficult to adolescent

scores to assess accept. Aswea
adjustment, tookresult
they are either less compliant serious the often repeated or more
claim that only alikely
small to
experience negative group psychological
of adopted children experiences consequenceadjust-
from their interactions ment with
problems and focused on
their predicting mem- which
parents,
are expressed in externalizing bership in that subgroup. behaviors.
In Laissez-Faire families, where
There are limits social of
to the generalizability reality
this
is neither dictated nor study's findings. For example,
discussed, rebellion we focused on against
parental authority might families with
play adolescent children. Family
a lesser role com- in put
ting adopted adolescents municationat patterns may operate differently
increased risk. The
salient factor in Laissez-Faire families is the among families with younger or older children.
absence of shared reality. We propose that chal- As noted above, longitudinal investigations are
lenges to developing a sense of identity faced needed. Also, as is characteristic of adoptive fam-
by adopted adolescents (Bimmel et al., 2003; ilies, the families in our sample were more edu-
Grotevant et al., 2001) are exacerbated in the cated and had higher incomes than the general
absence of a shared reality. For adopted adoles- population. They also were from the Midwestern
cents, questions about "who am I" can be com- United States and the parents were predominantly
plicated by limited information about birth Caucasian with European ancestry. Future stud-
parents and differences between themselves ies that include, for example, stepfamilies will
and adoptive family members. In nonadoptive need to test the generalizability of our findings
Laissez-Faire families, genetically based similar-to families with more varied socioeconomic,
ities afford at least a minimal sense of shared real-
regional, ethnic, and racial backgrounds.
ity, providing a foundation from which to answer To our knowledge, this is the first study to
questions about one's identity. demonstrate an interaction between adoption sta-
tus and family communication patterns. Accord-
Limitations and Future Directions
ing to our theory, this interaction occurs as
a function of parent-child genetic relatedness.
Although they are based on theory, these argu-This study is just a first step in fully testing this
ments are yet to be fully tested. For example,theory. If replicated through future studies, how-
we theorize that similarities among family mem-ever, our theoretical model could also apply to
bers based on genetic relatedness is the most other complex families in which parents and chil-
likely explanation for the interaction betweendren are genetically unrelated such as step- or
adoption status and family communication pat-blended families and families formed through as-
tern. We did not, however, measure cognitivesisted reproduction. Thus, this study represents
processes. Conducting research that directly an initial step in what could potentially be a much
assesses how family members perceive their en-wider field of study.
vironments, and in particular, how children per-
ceive their parents' regulatory messages will be
NOTE
an important next step in our research program.
Also, this study used cross-sectional data. There-
This research was supported by grants from the National Insti-
fore, it is possible that the observed family com-tute on Alcohol Abuse and Alcoholism (AA1 1886) and the
National Institute on Mental Health (MH066140).
munication patterns developed in response to or
in coincidence with child adjustment problems.
Future, longitudinal tests of this theory are REFERENCES
needed to understand better the complex pro-
cesses proposed here. Abrahamson, A. C, Baker, L. A., & Caspi, A. (2002).
Methodological strengths include using inno- Rebellious teens? Genetic and environmental influ-
vative methods for studying adoptive families ences on the social attitudes of adolescents. Jour-
and their communication. For example, this is nal of Personality and Social Psychology, 83,
the first study we know of that used observational 1392-1408.

data and latent cluster analysis to determiningAlford, J. R., Funk, C. L., & Hibbing, J. R. (2005).
family communication patterns, as identified by Are political orientations genetically transmitted?
the FCP Theory. All previous studies have used American Political Science Review, 99, 153 - 167.
self-reports only (Koerner & Fitzpatrick, 2004,Baumrind, D. (1971). Current patterns of parental
2006). Also, rather than using mean adjustment authority. Developmental Psychology, 4, 1 - 103.

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms
 726 Journal of Marriage and Family

Bimmel, N., Juffer, F., van J.Uzendoorn,

Haugaard, M. H.,
J., & Hazan, C. (2003). Adoption as a &
nat-
Bakermans-Kranenburg, M.
ural J. (2003).
experiment. Development Problem be-
and Psychopathology,
75,909-926.
havior of internationally adopted adolescents:
Juffer, F., & van
A review and meta-analysis. Uzendoorn, M. H.
Harvard (2005). Behaviorof
Review
Psychiatry, 11,64-77. problems and mental health referrals of interna-
Brand, A. E., & Brinich, P. M. (1999). Behavior
tional adoptees: A meta-analysis. Journal of the
problems and mental health contacts in adopted, American Medical Association, 293, 2501 - 2515.
foster, and nonadopted children. Journal of Kenny, D. A. (1991). A general model of consensus
Child Psychology and Psychiatry, 40, 1221 - and accuracy in interpersonal perception. Psycho-
1229. logical Review, 98, 155 - 163.
Brodzinsky, D., Lang, R., & Smith, D. (1995). Par- Keyes, M. A., Sharma, A., Elkins, I. J., Iacono,
enting adopted children. In M. Bornstein (Ed.), W. G., McGue, M. (2007). The mental health of
Handbook of parenting: Vol. 3. Status and social US adolescents adopted in infancy. Archives of
conditions of parenting (pp. 209 - 232). Hillsdale, Pediatrics & Adolescent Medicine, 161, 419 - 425.
NJ: Erlbaum. Koerner, A. F., & Fitzpatrick, M. A. (2002a). Toward
Brodzinsky, D. M. (2006). Family structural open- a theory of family communication. Communication
ness and communication openness as predictors in Theory, 72,70-91.
the adjustment of adopted children. Adoption Koerner, A. F., & Fitzpatrick, M. A. (2002b). Under-
Quarterly, 9, 1-18. standing family communication patterns and
Brodzinsky, D. M, Schechter, M. D., & Henig, family functioning: The roles of conversation ori-
R. M. (1992). Being adopted: The lifelong search entation and conformity orientation. Communica-
for self New York: Doubleday. tion Yearbook, 26, 37 - 69.
Burleson, B. R., Delia, J. G., & Applegate, J. L. Koerner, A. F., & Fitzpatrick, M. A. (2004). Commu-
(1995). The socialization of person-centered com- nication in intact families. In A. Vangelisti (Ed.),
munication: Parents' contributions to their child- Handbook of family communication (pp. 177 - 195).
ren's social-cognitive and communication skills. In Mahwah, NJ: Erlbaum.
M. A. Fitzpatrick & A. Vangelisti (Eds.), Explain-Koerner, A. F., & Fitzpatrick, M. A. (2006). Family
ing family interactions (pp. 34 - 76). Thousand communication patterns theory: A social cognitive
Oaks, CA: Sage. approach. In D. O. Braithwaite & L. A. Baxter
Conners, C. K. (1969). Conner's Teacher Rating (Eds.), Engaging theories in family communica-
Scale. Iowa City, I A: University of Iowa. tion: Multiple perspectives (pp. 50 - 65). Thou-
Deater-Deckard, K., & Petrill, S. A. (2004). Parent - sand Oaks, CA: Sage.
child dyadic mutuality and child behavior pro- Kohlberg, L. (1981). Essays on moral development,
blems: An investigation of gene - environment Vol. 1: The philosophy of moral development.
processes. Journal of Child Psychology and Psy- New York: Harper & Row.
chiatry, 45, 1171-1179. Lansford, J. E., Ceballo, R., Abbey, A., & Stewart,
Eccles, J. S., Midgley, C, Wigfield, A., Buchanan, A. J. (2001). Does family structure matter? A com-
C. M., Rueman, D., Flanagan, C, et al. (1993). parison of adoptive, two-parent biological, single-
Development during adolescence: The impact of mother, stepfather, and stepmother households.
stage-environment fit on young adolescents' expe- Journal of Marriage and family, 63, 840 - 85 1 .
riences in schools and in families. American Lanz, M, Ifrate, R., Rosnati, R., & Scabini, E.
Psychologist, 48, 90- 101. (1999). Parent-child communication and adoles-
Gibson, H. B. (1967). Self-reported delinquency
cent self-esteem in separated, intercountry adopted,
among school boys, and their attitudes to and
theintact non-adoptive families. Journal of Ado-
police. British Journal of Social and Clinical
lescence, 22, 785 - 794.
Psychology, 6, 168-173. Lee, R. M. (2003). The transracial adoption para-
Grotevant, H. D., Wrobel, G. M., van Dulmen, M.
dox: History, research, and counseling implica-
H., & McRoy, R. G. (2001). The emergence of of cultural socialization. The Counseling
tions
psychosocial engagement in adopted adolescents:
Psychologist, 31, 1 '11 -744.
The family as context over time. Journal ofLo, Y., Mendell, N. R., & Rubin, D. B. (2001). Test-
Ado-
lescent Research, 16, 469 - 490. ing the number of components in normal mixture.
Hagenaars, J., & McCutcheon, A. (Eds.). (2002).
Biometrika, 88, 767 - 778.
McGue, M., Keyes, M, Sharma, A., Elkins, I.,
Applied latent class analysis models. New York:
Cambridge University Press. Legrand, L., & Johnson, W. (2007). The environments

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms
 Family Communication Patterns 727

of adopted and non-adopted youth:

Rutter, M. A. (1967). Children'sEvidence
behavior question-on
range restriction from naire
the for Sibling Interaction
completion by teachers: and
Preliminary find-
Behavior Study (SIBS).
ings.Behavior Genetics,
Journal of Child Psychology 37,
and Psychiatry,
449-462. 8, 1-11.
Melby, J. N., Conger, R. D., Book, R., Rueter,
Schafer J. M.,
L., & Graham, J. W. (2002). Missing data:
Lucy, L., & Repinski, D. (1998). The Iowa Family
Our view of the state of the art. Psychological
Methods, 7, HI -111.
Interaction Rating Scales (5th ed.). Unpublished
manuscript. Ames, IA: Institute for Shrout,
SocialP.and
E., & Fleiss, J. L. (1979). Intraclass corre-
Behavioral Research, Iowa State University.lations: Uses in assessing rater reliability. Psycho-
logical reli-
Mitchell, S. K. (1979). Interobserver agreement, Bulletin, 86, 420 - 428.
ability, and generalizability of data collected
Steinberg, L.in(2001). We know these things: Parent-
observational studies. Psychological Bulletin, 86,
adolescent relationships in retrospect and prospect.
376-390. Journal of Research on Adolescence, II, 1 - 19.
Muthen, B., & Muthen, L. (1998-2006). Mplus
Stein, L. ver-
M., & Hoopes, J. L. (1985). Identity formation
sion 4.2. Los Angeles, CA: Muthen & Muthen.
in the adopted adolescent: The Delaware family
Muthen, B., & Shedden, K. (1999). Finite mixture
study. New York: Child Welfare League of America.
modeling with mixture outcomes using the
Suen, EM& Ary, D. (1989). Analyzing quantita-
K. K.,
algorithm. Biometrics, 55, 463 - 469. tive behavioral observation data. Hillsdale, NJ:
O'Brien, K. M., & Zamastny, K. P. (2003). Under-
Erlbaum.

standing adoptive families: An integrative review

Tesser, A. (1993). The importance of heritability in
psychological
of empirical research and future directions for coun- research: The case of attitudes. Psy-
chological Review, 100, 129- 142.
seling psychology. The Counseling Psychologist,
31,619-110. van IJzendoorn, M. H., Juffer, F., & Klein Poelhuis,
C. W. (2005). Adoption and cognitive develop-
Olson, J. M., Vernon, P. A., Harris, A., & Jang, K. L.
(2001). The heritability of attitudes: A study of ment: A meta-analytic comparison of adopted and
twins. Journal of Personality and Social Psychol- nonadopted children's IQ and school performance.
ogy, 80, 845 - 860. Psychological Bulletin, 131, 301 - 316.
Welner, Z., Reich, W., Herjanic, B., Jung, K., &
Reiss, D. (1981). The family's construction of reality.
Cambridge, MA: Harvard University Press. Amado, H. (1987). Reliability, validity, and
Reiss, D. (2000). The relationship code: Deciphering parent-child agreement studies of the Diagnostic
genetic and social influences on adolescent devel- Interview for Children and Adolescents (DICA).
opment. Cambridge, MA: Harvard University Journal of the Academy of Child and Adolescent
Press Psychiatry, 26, 649 - 653.
Rosnati, R., & Marta E. (1997). Parent - child rela-
Wrobel, G. M., Kohler, J. K., Grotevant, H. D., &
McRoy, R. G. (2003). The family adoption com-
tionships as a protective factor in preventing ado-
munication model (FAC): Identifying pathways
lescents' psychosocial risk in inter-racial adoptive
of adoption-related communication. Adoption
and non-adoptive families. Journal of Adoles-
cence, 20, 617 -631. Quarterly, 7, 53 - 84.

This content downloaded from

157.193.5.24 on Mon, 31 Oct 2022 22:56:03 UTC
All use subject to https://about.jstor.org/terms