You are on page 1of 7

A Comparison of the Family Systems of Adolescent Sexual Offenders and Nonsexual Offending

Delinquents
Author(s): Gary P. Bischof, Sandra M. Stith and Stephan M. Wilson
Source: Family Relations, Vol. 41, No. 3 (Jul., 1992), pp. 318-323
Published by: National Council on Family Relations
Stable URL: http://www.jstor.org/stable/585197
Accessed: 01-11-2015 19:32 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp

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 support@jstor.org.

National Council on Family Relations and Wiley are collaborating with JSTOR to digitize, preserve and extend access to
Family Relations.

http://www.jstor.org

This content downloaded from 134.184.26.108 on Sun, 01 Nov 2015 19:32:32 UTC
All use subject to JSTOR Terms and Conditions
A Comparison of the FamilySystems of Adolescent Sexual
Offenders and Nonsexual Offending Delinquents*
Gary P. Bischof, Sandra M. Stith, and Stephan M. WIlson**

While literature on adolescent sexual offenders has burgeoned over the last decade and families are often involved in treatment,
little is known about the family systems of these offenders. Adolescents' perception of family adaptability and cohesion are assessed
(using FACES-Ill) in an effort to identify whether and in what ways the family systems of adolescent sex offenders (n = 39) differ from
violent juvenile delinquents (n = 25), nonviolent juvenile delinquents (n = 41), and from nonproblem families (normative data). The data
reveal that families of sex offenders are characterized by greater family cohesion when compared with other delinquents, but sex
offenders perceive their families as less cohesive than do members of nonproblem families. No significant differences between the
groups were found for family adaptability. Implications for practice are offered.

U ntil the 1980s, adolescent sexual their families differ from other juvenile behavior as serving some positive func-
tion or purpose within the family system
offenders received little attention in delinquents. Findings may also promote
the professional literature. Adoles- continued development of theori6s on the (e.g., a son's delinquent behavior may
cent sexual offenders were not taken seri- etiology of sexual offenses by adoles- serve to unite parents who may be expe-
ously; their behavior was often explained cents, and help foster early identification riencing marital problems). Two aspects
as normal experimentation or develop- of at-risk families. of the family system that have been iden-
mental curiosity. A "boys will be boys" tified as indicators of family functioning
attitude prevailed (Knopp, 1985). Literature Review are adaptability and cohesion (Olson et
Although research has burgeoned in the al., 1983). Family adaptability is defined
Ageton (1983), in a general popula-
last decade, it has largely been con- as the ability of a family system to change
tion study, found that approximately 3 to
cerned with individual characteristics of its power structure, role relationships, and
4% of adolescents aged 15-21 years had relationship rules in response to situa-
offenders and offenses. Little is known
committed a sexual offense (i.e., an esti-
about the family systems of these offend- tional and developmental stress. Cohe-
mated 500,000 offenses annually). Crime
ers. It is also unclear how, if at all, these sion assesses the degree to which family
reports and surveys have determined that
offenders and their families differ from members are separated from or connect-
adolescents are responsible for approxi-
other juvenile delinquents or from adoles- ed to their family and is defined as the
mately 20% of rapes and 30 to 50% of
cents from nonproblem families since few emotional bonding that family members
cases of child sexual abuse (Davis & Leit-
studies compare these groups or include have toward one another (Olson, Portner,
enberg, 1987; Deisher, Wenet, Paperny,
a representative sample of nonoffenders & Lavee, 1985).
Clark, & Fehrenbach, 1982; Groth & Lore-
as a comparison group. do, 1981). These figures may underesti- Aside from descriptive studies, which
This study was designed to assess mate the actual number of adolescent have primarily addressed individual
adolescent sexual offenders' perceptions sexual offenders, due to the high number offender and offense characteristics,
of the adaptability and cohesion level of of rapes and incidents of child sexual research on adolescent sexual offenders
their families. Male adolescent sexual abuse which go unreported. In addition, is sparse. Many studies are limited by
offenders were compared with two com- only a small number of complaints ever small sample size and geographical bias.
parison groups: male juvenile delinquents result in an arrest (Groth & Loredo, 1981). Studies including normal control groups
who had committed violent nonsexual Studies of adult sexual offenders (Abel, are sparse (Davis & Leitenberg, 1987).
offenses and male juvenile delinquents Mittleman, & Becker, 1984; Becker & Although there is conjecture on how the
who had committed nonviolent nonsexual Abel, 1985) indicate that about half of family influences the commission of an
offenses. Male juvenile delinquents were adult offenders report that their first sex- offense and clinicians generally agree
selected for a comparison group ual offense occurred as an adolescent that an effective assessment of adoles-
because, while adolescent sexual offend- and often offenses escalated in frequency cent sexual offenders includes an assess-
ers and other juvenile delinquents are fre- and severity over time. These alarming ment of the offender's family (Becker,
quently placed together in residential findings have led to increased efforts in 1990), little scientific research has been
treatment facilities, most experts advo- the identification and treatment of adoles-
cate offense-specific treatment for ado- cents who sexually abuse and to the
lescent sexual offenders (Knopp, 1985). recognition of this group of offenders as a *The data used in this study were collected for the first
Professionals in the sex offender treat- distinct juvenile justice problem and clini- author's Master's Thesis at Virginia Polytechnic Institute and
State University's Department of Family and Child Develop-
ment field claim that adolescent sexual cal population. ment, Northern Virginia Graduate Center.
- Gary Bischof is a doctoral student in Marriage and
offenders are unique and distinct from A thorough examination of the vari- Family Therapy, Purdue University, 523 Russell Street, W.
other nondelinquent and delinquent ado- ous sociocultural contexts of adolescent Lafayette, IN 47906. Sandra Stith is an Assistant Professor of
Marriage and Family Therapy in the Department of Family and
lescents (Knopp, 1985; O'Brien, 1985). sexual offenders is beyond the scope of Child Development, Virginia Polytechnic Institute and State
Families of these offenders are frequently this study. However, one significant con- University, Northern Virginia Graduate Center, Falls Church,
text for developing adolescents is the VA 22042. Stephan Wilson is an Associate Professor of Fami-
involved in treatment and findings from ly Studies and Director of the University of Kentucky Research
this study are likely to enhance services family system. The behavior of each indi- Center for Families and Children, 197 Erikson Hall, Lexington,
to these families. Results from this study vidual in the family is influenced by the KY 40506.

may help clarify whether, and in what reciprocal interactions among family
ways, adolescent sexual offenders and members. Clinicians often view deviant (Family Relations, 1992, 41, 318-323.)

318 FAMILYRELATIONS July 1992

This content downloaded from 134.184.26.108 on Sun, 01 Nov 2015 19:32:32 UTC
All use subject to JSTOR Terms and Conditions
conducted to confirm or disclaim these more rigid than nonproblem families. ings are consistent with correlations
impressesions about the family dynamics Conversely, mothers of substance McGaha and Fournier (1988) found
of adolescent sexual offenders. Review- abusers saw their families as slightly between the level of iigidity and dis-
ing the literature, Monastersky and Smith more chaotic than nonproblem families. engagement in the family system and the
(1985) conclude that studies are virtually In a study of juveniles referred to seriousness of offense within the general
unanimous in identifying the family as a juvenile delinquent population.
juvenile court intake (McGaha & Fournier,
crucial influence in the development or 1988) which assessed both the adoles- Smith and Monastersky (1987) noted
elicitation of the offending behavior, but it cents' (n = 40) and parents' (n = 50) per- a difference between the perceptions of
is not clear how this occurs. ceptions using FACES-Il, the study adolescent sexual offenders in their study
Knopp (1982), reporting on limited sample was found to be significantly less and the perceptions of juvenile delin-
clinical impressions of an unspecified cohesive and much more rigid than the quents from previous studies. The ado-
number of families in one adolescent sex- national norms. McGaha and Fournier lescent sexual offenders' perceptions of
ual offender program, found that families also reported that family adaptability and their families were generally similar to
of adolescent sexual offenders reflect two cohesion were related to the type of their parents' perceptions, while studies
types of family systems. Either the fami- offense. The Circumplex Model classifies of juvenile delinquents have revealed that
lies are very rigid and enmeshed, with families based on scores for adaptability the delinquents frequently viewed their
strict rules and a perfectionist bent to and cohesion. In the original conceptual- family as more harsh and unresponsive
parental expectations, or they are very ization of FACES-Ill, 16 different types of than did other family members. Further-
chaotic with a great deal of role confu- families were identified using this instru- more, the authors suggest that future
sion. Knopp reports that staff in another ment (Olson et al., 1985). These family research compare the family systems of
adolescent sexual offender program serv- types were grouped under three general sexual offending and nonsexual offending
ing very violent and dangerous sexual family types: extreme, midrange and bal- juvenile delinquents in an effort to identify
offenders in a long-term locked facility anced. Extreme families (i.e., considered distinguishing characteristics between the
used the word "chaotic" when describing most dysfunctional), who scored in the groups. Neither of these studies (Bera,
the families of the majority of the adoles- extreme range on both measures, tended 1985; Smith & Monastersky, 1987) has
cents. In these families it was not uncom- to commit more violent crimes (e.g., been published in the professional litera-
mon for one of the parents to have assault) or status offenses (e.g., running ture.
demonstrated deviant behavior very simi- Sefarbi (1990) evaluated adaptability
away, incorrigibility,or truancy) while bal-
lar to the child's. Thus, families of adoles- anced (moderate on both measures) and and cohesion (FACES-Il) for adolescent
cent sexual offenders have been sexual offenders who denied their offense
midrange (moderate in only one mea-
described in the literature as rigid and (n = 5) and those who admitted to their
sure) families were more likely to commit
enmeshed or chaotic. The implications of offense (n = 5). The "deniers" tended to
property crimes.
these terms is that these families are be in enmeshed family systems, while
either inflexible or too flexible and that Only three studies have addressed "admitters" were in disengaged family
they are too close emotionally, the result adaptability and cohesion of families of systems. Deniers' families were also
being a dysfunctional family system that adolescent sexual offenders formally. characterized by overwhelmed mothers
discourages or prevents healthy develop- Each of these studies used FACES-Il. who relied on the parentified adolescent
ment among its members. However, the Considering adolescent sexual offenders sexual offender for physical and emotion-
authors drawing these conclusions typi- (n = 51) who had committed offenses of al support. A history of abandonment, first
cally have been clinicians who have various patterns and levels of severity, by fathers, and later by mothers for cru-
based their ideas solely on clinical Bera (1985) found no significant differ- cial periods, distinguished the families of
impressions. One recent retrospective ence between these families and non- admitters.
study examined male college students' problem adolescent families. Further, no
earlier experience as perpetrators of child differences were found in the family sys- In summary, the family system is a
sexual abuse (Fromuth, Burkhart, & tems of "mild"(i.e., few and less violent significant context for adolescents. A
Jones, 1991). These authors found no dif- offenses) versus "severe" (i.e., many and review of the literature revealed only one
ferences in the family backgrounds of the more violent) adolescent sexual offend- small published study assessing the
16 men who reported that they had ers. adaptability and cohesion of the family
molested a child and the 566 men who systems of adolescent sexual offenders.
While Bera used only the adoles- Other research has not been published,
reported no such history.
cent's perception of his family system, thus limiting knowledge about families of
While only a limited number of stud- Smith and Monastersky (1987) gathered adolescent sexual offenders. Families are
ies have used reliable measures to perspectives from adolescent sexual frequently involved in treatment and are
assess adaptability and cohesion of fami- offenders (n = 66) and their mothers (n = thought to be influential in the etiology of
lies of adolescent sexual offenders, stud- 71) and fathers (n = 51). Their sample sexual offenses by adolescents. No stud-
ies have appeared addressing these included adolescents treated in an out- ies directly compared the family systems
dimensions in families of more general patient program, a majority of whom had of sexual offending and nonsexual offend-
troubled adolescents. For example, committed less aggressive offenses. ing juvenile delinquents, though experts
assessing adolescents in outpatient treat- They found that the adolescent sexual suggest there are differences between
ment and prevention programs, adoles- offender families were more likely than these families. This study was designed
cent substance abusers (n = 148) and the general population to be character- to address these gaps in the literature by
their mothers (n = 135) and fathers (n = ized as rigid in response to changes (i.e., comparing family adaptability and col-e-
67) reported their perceptions of their low adaptability) and emotionally dis- sion of adolescent sexual offenders, vio-
families using FACES-Ill (Volk, Edwards, engaged (i.e., low cohesiveness). In their lent juvenile delinquents, nonviolent
Lewis, & Sprenkle, 1989). These families study, degree of violence of the offense juvenile delinquents, and of a representa-
were found to be far more disengaged was related to the family system; the tive sample of nonproblem adolescents. It
than nonproblemfamilies, while on mea- more rigid and disengaged the family was expected that the family systems of
sures of adaptability,substance-abusing (according to the parents' perceptions) adolescent sexual offenders would differ
adolescents saw their families as slightly the more violent the offense. These find- from violent and nonviolent delinquents
July 1992 FAMILYRELATIONS 319

This content downloaded from 134.184.26.108 on Sun, 01 Nov 2015 19:32:32 UTC
All use subject to JSTOR Terms and Conditions
and from nonproblem adolescents, thus ily adaptability has to do with the extent to lent offenses against persons were
supporting recognition of adolescent sex- which the family system is flexible and included in the violentjuvenile delinquent
ual offenders as a distinct juvenile justice able to change. Sample items for the group and juvenile delinquents who self-
problem and clinical treatment population. adaptability scale include, "Different per- reported committing nonviolent offenses
sons act as leaders in our family" and such as property offenses, nonviolent
Methods "Rules change in our family." crimes against persons, status offenses,
Although FACES-Ill was originally
or substance abuse violations were
Participants conceptualized as measuring adaptability
included in the nonviolent delinquent
and cohesion as curvilinear measures
group.
Questionnaires were distributed to
109 adolescent males in various out- (i.e., families with balanced and midrange Demographic variables. The follow-
patient and residential programs; 105 levels of adaptability and cohesion were ing demographic factors were included:
questionnaires were returned (96.3%). considered more healthy than families age and race of participants, family
Participants were adolescent males, aged with extreme scores on these measures), income, and parental employmentstatus
12-18, who were grouped as follows: (a) more recent research has disputed the and occupation. These were assessed
adolescent sexual offenders (n = 39) who curvilinear hypothesis (Green, Harris, withstandardfact sheet items.
self-reported having committed child sex- Forte, & Robinson, 1991; Olson, 1991a,
ual molestation or who were involved in 1991b). In fact, David Olson (1991a), one Procedure
treatment programs designated for iden- of the authors of the instrument, recently A paper-and-pencilself-reportsurvey
tified adolescent sexual offenders; (b) indicated that, "there is considerable evi- was administered to adolescent partici-
juvenile delinquents (n = 25) who self- dence that FACES-Ill is a linear measure, pants by a research investigatoror treat-
reported committing violent offenses with high scores on cohesion and adapt- ment professional. Programs/facilities
(e.g., homicide, manslaughter, robbery, ability being related to more functional participatingin the study signed a state-
aggravated assault); and (c) juvenile family relationships. High scores on cohe- ment of participation, indicating their
delinquents (n = 41) who self-reported sion and adaptability are reconceptual- understandingof the purpose and proce-
committing nonviolent offenses such as ized as measuring Balanced family types dures of the study. Parents/guardians/
property offenses, nonviolent crimes and low scores on the two dimensions as custodians gave their written permission
against persons, status offenses, or sub- measuring Extreme family types (p. 74). and informed consent for the youth in
stance abuse violations. A nondelinquent The revised linear conceptualization of their charge to participate in the study
control group was not included in this FACES-Ill with corresponding cutoff and anonymitywas assured.
study, but FACES-Ill has been normed to points for Very Cohesive, Cohesive,
the general population of adolescent fam- Somewhat Cohesive, Disengaged, Very Results
ilies and these norms were compared to Flexible, Flexible, Structured, and Rigid
scores by the groups in this study. Partici- was used in this study.
pants were not paid, but were encour-
DemographicFactors
Internal consistency reliability for the
aged to participate in the research as a cohesion subscale was .86 and for the Analyses of Variance (ANOVAs)
way to help other young people and soci- adaptability scale was .64 for the current
were executed among the three groups
ety in general. study sample. The perceived form, cover-
on demographicfactors. There was a sig-
nificantdifference in age among the three
ing perceptions of current family function-
Measures groups [F(2) = 6.41, p = .002]; mean age
ing, was used; adolescents were asked to for the adolescent sexual offender group
Family Adaptability and Cohesion report on their family at the time of the was 15.39, whereas mean age for violent
Evaluation Scale (FACES-Ill) (Olson et offense. and nonviolent juvenile delinquents was
al., 1985). FACES-Ill is a 20-item self- Delinquency and Sexual Offense 16.16 and 16.34, respectively. Results
report scale that is the most recent ver- Self-Report. Participants were provided a from a Chi-squareanalysis demonstrated
sion of the Family Adaptability and list of several categories of delinquent that there were no significantdifferences
Cohesion Evaluation Scale. The instru- acts and sexual offenses and were asked in race among the three groups (73%
ment is designed to assess perceptions to indicate whether they had committed white, 16% black, 3% Latino, 3% Asian,
of the family system. The revised version an offense from each category and 5% other for entire sample).
differs in several important ways from ver- whether they had been held by the police
sions (FACES and FACES-Il) which were Participants came predominantly
or convicted for any offense from a partic- from the Washington, D.C. metropolitan
used in earlier research on families of ular category. The seven categories area. Level of family income differedsig-
adolescent sexual offenders. The instru- included: violent offenses against per-
ment is shorter, it has two empirically
nificantlybetween the three groups [F(2)
sons, property offenses, nonviolent = 4.44, p = .01]. Mean familyincome was
independent (orthogonal) dimensions so offenses against persons, general sex $38,400, $41,000, and $52,000 for ado-
it better achieves the theoretical criteria offenses, child molestation, offenses lescent sexual offenders, violent juvenile
for a Circumplex Model; it is more rele- against public order and drug abuse vio- delinquents, and nonviolent delinquents,
vant for a variety of nonnuclear family lations, and status offenses. This self- respectively. The majority of fathers
forms, and it has specific norms for fami- report instrument was used to classify the (74%)and mothers (71%)were employed
lies across the life cycle. Participants participants into the appropriate group full-time, with no significant differences
respond to each question with one of five (i.e., adolescent sexual offenders, non- between groups. Mothersmost frequently
Likert-type answers from almost never to violent delinquent, or violent delinquent). workedin service occupations (33%)or in
almost always. Family cohesion assesses Participants who indicated that they had technical or clerical work (32%). Fathers
the degree to which family members are committed child molestation and those worked primarily in service or military
separated from or connected to their fam- who were involved in treatment programs occupations (26%) and in administration,
ily. Sample items for the cohesion scale designated for identifiedadolescent sex- engineering, or scientific endeavors
include, "Familymembers feel very close ual abusers were included in the adoles- (19%). There were no significant differ-
to each other"and "Familymembers like cent sexual abuser group. Juvenile ences among the groups on parents'
to spend free time with each other."Fain- delinquentswho reportedcommittingvio- occupation.
320 FAMILYRELATIONS July 1992

This content downloaded from 134.184.26.108 on Sun, 01 Nov 2015 19:32:32 UTC
All use subject to JSTOR Terms and Conditions
Table 1. pared with the perceptions of violent juve-
ANOVAs for Study Samples and Normed Scores for Family Adaptability and Cohesion nile delinquents and nonviolent juvenile
Adolescent Violent Nonviolent delinquents. These data were compared
Sexual Juvenile Juvenile with normative data for nonproblem ado-
Offender Delinquent Delinquent lescents and their families. Family adapt-
(n = 37) (n = 24) (n = 40) Norms
Variable Mean Mean Mean Mean F-ratio ability and cohesion were considered. It
was anticipated that the family systems of
FACES-Ill: adolescent sexual offenders would differ
Adaptability 23.76a 24.54a 23.05a 24.30a .44
Cohesion 30.08a 24.96b 25.05b 37.1 Oc 87.76* from violent and nonviolent delinquents
and from nonproblem adolescents, thus
Note. Norms for FACES-Ill are for parents and adolescents combined (n = 1,315). supporting recognition of adolescent sex-
abcDuncanmultiple range test results: Clusters with different letters differ significantly (p < .05).
*p < .001. ual offenders as a distinct juvenile justice
problem and clinical treatment population.
solving, and many strict rules. Four levels There were no significant differences
Family Adaptability between the groups in this study or
of cohesion have also been described
and Cohesion (i.e., very cohesive, cohesive, somewhat between the study samples and nonprob-
cohesive, and disengaged). Very cohe- lem adolescent families in family adapt-
Because adaptability and cohesion
sive and cohesive are characterized as ability. Mean scores for adaptability for all
have been recently reconceptualized as
most healthy and as most balanced on the samples in this study were in the
linear constructs (Olson, 1991a, 1991b),
ANOVAs were performed for measures of cohesion. Disengaged families (very low moderate range, toward the rigid (i.e., low
cohesion) are typified by low emotional adaptability) level. This confirms findings
adaptability and cohesion (see Table 1).
There were no significant differences bonding, closed internal boundaries, rigid by Bera (1985) that outpatient adolescent
generational boundaries, and a general sexual offenders were very similar to non-
among the four groups (i.e., adolescent
sense of separateness. In contrast, very problem adolescents in their perception of
sexual offenders, violent delinquents,
nonviolent delinquents, and nonproblem cohesive and cohesive families are char- family adaptability. Yet, it is inconsistent
acterized by balanced levels of emotional with results of an analysis of outpatient
adolescents) on family adaptability. There
bonding. See Table 2 for the full range of adolescent sexual offenders (Smith &
were differences [F(3) = 87.76, p < .001]
levels for both adaptability and cohesion. Monastersky, 1987) which found that
for cohesion. Post hoc Duncan tests con-
families of adolescent sexual offenders
firmed that adolescent sexual offenders The mean scores for adaptability were more likely than the general popula-
perceived their families as having higher (i.e., adolescent sexual offenders = 23.76, tion to be characterized as rigid in
levels of emotional bonding among family violent = 24.54, nonviolent = 23.05) all fall response to changes.
members (X = 30.08) than did either vio- within the structured level considered
lent (X = 24.96) or nonviolent (X = 25.05) moderate (toward the rigid level), and Adolescent sexual offenders per-
delinquents. Non-problem families (X = within the range for healthy family func- ceived their family system as more cohe-
37.10) were perceived as being signifi- tioning. While the mean scores of the four sive, that is, viewed their family members
cantly more cohesive than adolescent groups differed significantly on levels of as more bonded emotionally to each
sexual offenders, violent delinquents, and cohesion, all the scores fall within the dis- other (Olson et al., 1985), than did delin-
nonviolent juvenile delinquents according engaged level, indicating that adolescents quents in this study. While mean scores
to post hoc Duncan tests. in all three groups perceived their families for family cohesion were significantly dif-
The revised guide to the FACES-Ill as having low emotional bonding, closed ferent, all fell within the lowest level for
(Olson, 1991b) provides cutoff scores for internal boundaries, rigid generational cohesion (i.e., disengaged), which raises
levels of adaptability and cohesion. Four boundaries, and a general sense of sepa- the question about whether this statistical
levels of adaptability have been delineat- rateness. Adolescent sexual offenders difference is meaningful. In addition, ado-
ed (i.e., very flexible, flexible, structured see their families as more cohesive than lescent sexual offenders perceive their
and rigid). Very flexible and flexible fami- do other delinquents, but still as disen- families as significantly less cohesive
lies are characterized as most healthy gaged, the extreme low level of cohesion. than do nonproblem adolescents. These
and as most balanced on adaptability. findings support those found by Smith
Rigid families (very low adaptability) are Discussion and Monastersky (1987), that families of
characterized by authoritarian leadership, adolescent sexual offenders are more
Adolescent sexual offenders' percep- likely than nonproblem families to be
limited negotiation and poor problem tions of their family system were com- emotionally disengaged. The preponder-
ance of low family cohesiveness for fami-
Table 2.
FACES-Ill Family Types (as percentages) for Study Families with Offenders and Nonproblem Fam- lies of adolescent sexual offenders in this
ilies. study conflicts with clinical impressions
(Knopp, 1982) that many of these families
Adolescent Violent Nonviolent
Sexual Juvenile Juvenile are enmeshed. Only one adolescent sex-
Offender Delinquent Delinquent Nonproblem ual offender in this study viewed his fam-
Family Type (n = 37) (n = 24) (n = 40) (n = 1,315) ily as being very cohesive.
Cohesion: The lack of enmeshed families of
Disengaged 54.1 75.0 80.0 18.6
Somewhat cohesive 24.3
adolescent sexual offenders could be
16.7 10.0 30.3
Cohesive 18.9 8.3 10.0 36.4 attributed to several factors. First, prob-
Very cohesive 2.7 - - 14.7 lems inherent in the FACES-Ill instrument
Adaptability:
may not have allowed the study to identify
Very flexible 18.9 12.5 12.5 13.9 enmeshed families. That iS, the instru-
Flexible 24.3 41.7 27.5 32.9 ment may be more sensitive to identifying
Structured 32.4 25.0 32.5 37.3 disengaged families, rather than en-
Rigid 24.3 20.8 27.5 15.9 meshed families. Second. only the ado-
July 1992 i FAMILY RELATIONS 321

This content downloaded from 134.184.26.108 on Sun, 01 Nov 2015 19:32:32 UTC
All use subject to JSTOR Terms and Conditions
lescent's perception of his family was divided between outpatients and resi- violent juvenile delinquents in terms of
assessed in this study. Adolescents tend dents. Placement in residential care often adaptability. Though there was a signifi-
to view their family with greater nega- comes only after other less restrictive cant difference in mean scores for family
tivism and tend to perceive their families alternatives have been exhausted, and cohesion, the means all fell within the
as more disengaged than do other family frequently family dysfunction is a criterion same disengaged level of cohesion.
members (Olson et al., 1983). Normative for placement out of the home. Future Treatment approaches and family inter-
data for adolescent perceptions using studies should either include offenders ventions proven effective with general
FACES-Ill is not available. Normed from various levels of outpatient and resi- juvenile delinquents are likely to be help-
scores represent combined adolescents' dential treatment, or should control for ful with adolescent sexual offenders, too.
and parents' perceptions. This may placement setting. Clinical literature and experience with
account for the preponderance of low Information in this study was ob- general juvenile delinquents are more
family cohesion scores in the study sam- tained by adolescent retrospective self- advanced and lessons learned with juve-
ple. Interestingly, though, Smith and report. Participants were asked to report nile delinquents appear to be transferable
Monastersky (1987) surveyed both ado- on conditions at the time of their offense. to adolescent sexual offenders and their
lescents and parents using FACES-Ill In some cases, a long period of time had families.
and found that adolescent sexual offend- passed since the offense, inviting the While similar to delinquents' families
ers and their parents perceived their fami- possibility of treatment effect and inaccu- in some ways, the families of adolescent
lies very similarly. Similarity between rate recall of prior family conditions. For a sexual offenders closely approximated
adolescent and parent scores was signifi- more accurate and complete perspective nonproblem families on level of cohesion.
cantly higher for families of adolescent of the family system information ideally These findings support some previous lit-
sexual offenders than it was for families should be obtained early in the assess- erature (Fagan & Wexler, 1988) that ado-
of nonsexual offending delinquents. ment or screening process and should lescent sexual offenders and their
Therefore, for these families, the adoles- include the perceptions of several family families are more similar to nonproblem
cent's perception alone may be an accu- members. It is recommended that family adolescents and their families than are
rate representation of the entire family's instruments be included in routine other delinquents. This would indicate
perception. assessment, as families are frequently that these families have competencies
The second factor that could account involved in treatment. A more compre- that would be helpful for the clinician to
for the discrepancy between clinical hensive understanding of the family con- notice, call attention to, and enhance in
impressions and family members' experi- text would prove valuable in treatment treatment. It also suggests that a dysfunc-
ence of their family is the vantage point, planning and intervention design. tional family system is not necessarily
either from inside the family or as an out- Detailed information about offenses associated with sexual offenses by ado-
sider looking in (Olson, 1977). Clinical and offense history was not obtained in lescents. It may be that for some types of
impressions are outsiders' perspectives; this study. Future studies should attempt offenders, sociocultural contexts other
clinicians see many families of adolescent to differentiate family characteristics than the family (e.g., peer group, mass
sexual offenders as enmeshed, with dif- among various types of adolescent sex- media, societal norms related to mas-
fuse internal boundaries, and closed ual offenders. Indeed, Bera (1985) found culinity and aggression) are more influen-
external boundaries (Knopp, 1982). They little difference between families of ado- tial in the etiology of the offenses.
often report that these families rely on lescent sexual offenders and nonproblem The finding that most adolescent
their family almost exclusively for support adolescent families in general, but signifi- sexual offenders view their families as
and that they are socially isolated. Family cant differences emerged between vari- disengaged and few perceive their fami-
members, though, perceive their family ous types of adolescent sexual offenders, lies as very cohesive can be beneficial to
from the inside, resulting in a different classified according to offenses and clinicians. Understanding the adoles-
view. Hence, they see their family and the offense patterns. cent's perspective can be helpful in join-
relationships therein as less cohesive and ing and creating strategies that fit with the
more separate. This explanation was also A limitation mentioned above was the
lack of normative data for the adoles- adolescent's world view. Utilization of
proposed to account for a similar discrep- interventions designed to enhance emo-
ancy between the clinical literature (i.e., cent's perception for FACES-Ill. Compar-
isons of adolescents' perceptions for tional support and bonding among family
clinicians viewing families as enmeshed members would seem appropriate for
and over-involved) and family perceptions samples in this study and combined ado-
lescent and parent scores for nonproblem most of these families.
(i.e., family members viewing themselves
as disengaged) for adolescent substance families may be questionable. Admittedly, While the adolescent views his family
abuse families (Volk et al., 1989). it is preferable to gather information from as disengaged from an insider perspec-
several family members when assessing tive, the clinician can maintain an outsider
Limitations family variables, yet this is not always perspective that may be helpful in hypoth-
possible, and was beyond the scope of esizing about the family system. The sys-
Participants in this study were volun- this study. The reporting of separate fami- temic notion that delinquent behavior
tary and self-selected, and therefore are ly member's perceptions for normative serves some positive function in the fam-
not necessarily representative of the data is suggested so that accurate com- ily system could be applied here. Adoles-
delinquent populations included here. parisons can be made when access is cent sexual offenders who view their
Parental/guardian permission first had to limited to individual family members. families as disengaged may be attempt-
be obtained, and then adolescents decid- ing through their behavior to unite family
ed if they would participate after learning Implications for Practice members around some common cause.
what would be required. Participants Another systemic hypothesis is that the
came from programs/facilities that agreed The review and findings of this study
have a number of implications for inter- adolescent sexual offender's offense is a
to cooperate with the study. Nonsexual misplaced attempt at "closeness" with
offendingdelinquentscame primarilyfrom vention with families of adolescent sexual
offenders. The familysystems of adoles- other family members from whom he
four residentialprograms,and all but one feels disengaged. Further, the offense
was involved in a residential program. cent sexual offendersappearto be similar
Adolescent sexual offenders were closely to the familysystems of violent and non- could be seen as an effortto seek help

322 FAMILYRELATIONS July 1992

This content downloaded from 134.184.26.108 on Sun, 01 Nov 2015 19:32:32 UTC
All use subject to JSTOR Terms and Conditions
for other unacknowledged problems (e.g., behavior:The role of the physician.Joumalof Adolescent program'sperspectiveon researchdirections.InE. M.Otey
Health Care, 2, 279-286. & G. D. Ryan (Eds.), Adolescentsex offenders:Issues in
incest between other family members, Fagan,J., &Wexler,S. (1988). Explanationsof sexual assault researchand treatment(pp. 147-163).Rockville,MD:USD-
substance abuse) that may be occurring among violent delinquents. Journal of Adolescent HHS.
Research, 3, 363-385. Olson, D. H. (1977). Insiders'and outsiders'views of relation-
in the family. Fromuth,M., Burkhart,B., & Jones, C. (1991). Hiddenchild ships: Research studies. In G. Levinger& H. L. Rausch
molestation:An investigationof adolescent perpetratorsin Perspectiveson the meaningof
(Eds.), Close relationships:
a nonclinical sample. Journal of Interpersonal Violence, 6, intimacy (pp. 115-136). Amherst: University of Mas-
REFERNCES 376-384. sachusetts Press.
Green,R. G., Harris,R. N., Forte,J.A.,&Robinson,M. (1991). Olson, D. H. (1991a). Commentary:Three-dimensional(3-D)
Abel, G., Mittleman, M., & Becker, J. (1984). Sexual offenders: EvaluatingFACES-Illand the circumplexmodel: 2,440 circumplexmodeland revisedscoringof FACES-Ill.Family
Results of assessment and recommendations for treat- families. Family Process, 1, 55-73. Process, 30, 74-79.
ment. Unpublished manuscript, available from New York Groth,A. N., & Loredo,C. M. (1981). Juvenilesexual offend- Olson,D. H. (1991b, October).FamilyCircumplex Model:Clini-
State Psychiatric Institute, Sexual Behavior Clinic. ers: Guidelines for assessment. International Journal of cal assessment and treatmentintervention.Paperpresent-
Ageton, S. S. (1983). Sexual assault among adolescents. Lex- Offender Therapy and Comparative Criminology, 25, 31-39. ed at the AnnualConferenceof the AmericanAssociation
ington, MA: Heath. Knopp, F. H. (1982). Remedial intervention in adolescent sex forMarriageand FamilyTherapy,Dallas,TX.
Becker, J. V., & Abel, G. G. (1985). Methodological and ethical offenses: Nine program descriptions. Orwell, VT: Safer Olson, D. H., McCubbin,H. I., Barnes, H. L., Larsen,A. S.,
issues in evaluating and treating adolescent sex offenders. SocietyPress. Muxen, M. J., & Wilson, M. A. (1983). Families: What
In E. M. Otey & G. D. Ryan (Eds.), Adolescent sex offend- Knopp,F. H. (1985). Recentdevelopmentsin the treatmentof makes themwork.BeverlyHills,CA:Sage Publications.
ers: Issues in research and treatment (pp. 109-129). adolescent sex offenders. In E. M. Otey & G. D. Ryan Olson, D. H., Portner,J., & Lavee, Y. (1985). FACES-Ill.St.
Rockville, MD: USDHHS. (Eds.), Adolescent sex offenders: Issues in research and Paul:FamilySocial Science, Universityof Minnesota.
Becker, J. V. (1990). Treating adolescent sexual offenders. treatment(pp. 1-27).Rockville,MD:USDHHS. Sefarbi, R. (1990). Admittersand deniers among adolescent
Professional Psychology: Research and Practice, 21, 362-365. McGaha,J., & Fournier,D. (1988). Juvenilejustice and the sex offendersand theirfamilies:A preliminarystudy.Amer-
Bera, W. H. (1985). A preliminary investigation of a typology of family:A systems approachto familyassessment. Marriage ican Journalof Orthopsychiatry,60, 460465.
adolescent sex offenders and their family systems. Unpub- and Family Review, 12, 155-172. Smith,W. R., & Monastersky,C. (1987). Strategiesforstudy-
lished masters thesis, University of Minnesota, St. Paul. Monastersky,C., & Smith,W. (1985). Juvenilesexual offend- ing the role of the family in the commission of sexual
Davis, G. E., & Leitenberg, H. (1987). Adolescent sex offend- ers: A familysystems paradigm.InE. M.Otey &G.D. Ryan offenses by adolescents.Unpublishedmanuscript.
ers. Psychological Bulletin, 101, 417-427. (Eds.), Adolescent sex offenders: Issues in research and Volk, R. J., Edwards,D. W., Lewis, R. A., & Sprenkle,D. H.
Deisher, R. W., Wenet, G. A., Paperny, D. M., Clark, T. F., & treatment(pp. 164-175).Rockville,MD:USDHHS. (1989). Familysystems of adolescent substance abusers.
Fehrenbach, P. A. (1982). Adolescent sexual offense O'Brien,M.(1985).Adolescentsexual offenders:An outpatient FamilyRelations,38, 266-272.

July 1992 FAMILYRELATIONS 323

This content downloaded from 134.184.26.108 on Sun, 01 Nov 2015 19:32:32 UTC
All use subject to JSTOR Terms and Conditions

You might also like