The Efficacy of an Intervention to Promote Parental Investment in Hispanic immigrant Families

P1: GRA
Prevention Science [PREV]

pp917-prev-469241

July 8, 2003

14:33

Style file version Nov. 04, 2000

C 2003)
Prevention Science, Vol. 4, No. 3, September 2003 (°

Familias Unidas: The Efficacy of an Intervention to Promote
Parental Investment in Hispanic Immigrant Families
Hilda Pantin,1,4 J. Douglas Coatsworth,2 Daniel J. Feaster,1 Frederick L. Newman,3
Ervin Briones,1 Guillermo Prado,1 Seth J. Schwartz,1 and Jos´e Szapocznik1

This paper reports a test of the efficacy of Familias Unidas, a Hispanic-specific, ecologically
focused, parent-centered preventive intervention, in promoting protection against and reducing risk for adolescent behavior problems. Specifically, the intervention was designed to
foster parental investment, reduce adolescent behavior problems, and promote adolescent
school bonding/academic achievement, all protective factors against drug abuse and delinquency. One-hundred sixty seven Hispanic families of 6th and 7th grade students from three
South Florida public schools were stratified by grade within school and randomly assigned to
intervention and no-intervention control conditions. Results indicated that Familias Unidas
was efficacious in increasing parental investment and decreasing adolescent behavior problems, but that it did not significantly impact adolescent school bonding/academic achievement.
Summer-vacation rates of adolescent behavior problems were six times higher in the control
condition than in the intervention condition. Furthermore, change in parental investment during the intervention was predictive of subsequent levels of adolescent behavior problems. The
findings suggest that Familias Unidas is efficacious in promoting protection and reducing risk
for adolescent problem behaviors in poor immigrant Hispanic families.
KEY WORDS: prevention; family; adolescents; hispanic; parental investment.

INTRODUCTION

reported use during the previous month (NIDA,
2001). Juvenile crime rates are similarly high. In 1998,
616,000 crimes were committed by juveniles, and 22%
of all violent crimes involved at least one offender under the age of 18 years (Snyder & Sickmund, 1999).
Moreover, Hispanic adolescents are overrepresented
among drug abusing and delinquent youth (Snyder &
Sickmund, 1999; Vega & Gil, 1999).
The antecedents of adolescent drug abuse and
antisocial behavior are known to involve family processes such as parental disinterest, disengagement,
and uninvolvement in adolescents’ lives. In nonHispanic White samples, indices of lack of parental
involvement, such as low parental monitoring of adolescent activities and lack of bonding to the adolescent, have been concurrently and prospectively
linked to adolescent drug abuse and antisocial behavior (Palmer & Hollin, 2001; Rosenstein & Horowitz,
1996; Steinberg et al., 1994). In non-Hispanic Whites,
it is generally assumed that poor or inept parenting is
responsible for such uninvolvement (e.g., Pettit et al.,

Adolescent problem behaviors, such as drug
abuse and delinquency, are pervasive in American society. In a recent Monitoring the Future report, 50–
70% of middle and high school students reported
drug or alcohol use during their lifetimes, 40–60%
reported use during the previous year, and 20–40%

1

Center for Family Studies, Department of Psychiatry and Behavioral Sciences, University of Miami School of Medicine, Miami,
Florida.
2
Department of Human Development and Family Studies,
Pennsylvania State University, University Park, Pennsylvania.
3
Department of Health Services Administration, Florida International University, North Miami, Florida.
4
Correspondence should be directed to Hilda Pantin, PhD, Center for Family Studies, Department of Psychiatry and Behavioral Sciences, University of Miami School of Medicine, 1425
N.W. 10th Avenue, 3rd Floor, Miami, Florida 33136; e-mail:
[email protected].

189
C 2003 Society for Prevention Research
1389-4986/03/0900-0189/1 °

P1: GRA
Prevention Science [PREV]

190

pp917-prev-469241

July 8, 2003

14:33

Style file version Nov. 04, 2000

Pantin, Coatsworth, Feaster, Newman, Briones, Prado, Schwartz, and Szapocznik

1997; Tarter et al., 1999). However, in some groups,
such as recent Hispanic immigrants, factors related
to immigration and acculturation, in addition to unskilled parenting, may play a major role in distancing
parents from their adolescents (e.g., Gil et al., 2000).
Hispanics are the largest and fastest growing minority group in the United States, comprising nearly
50% of current immigrants (Lollock, 2001). Therefore, understanding the factors that contribute to the
development of problem behaviors in Hispanic immigrant adolescents, and developing and testing prevention programs to address these factors, is of great
importance.
Problem-behavior prevention programs have received a great deal of recent attention, given that they
have the potential to minimize interpersonally and
socially damaging behaviors and to redirect adolescents onto more positive developmental paths (e.g.,
Conduct Problems Prevention Research Group, 2000;
Dishion & Kavanagh, 2000). As a result, our program of prevention research has focused on intervening during early adolescence to arrest the development of substance abuse and antisocial behavior.
Given the pivotal role of family in these problematic
outcomes, we have focused on designing and implementing family-based prevention programs.
Earlier in our program of prevention research,
we designed and implemented three successive prevention programs, each targeting one of the family
conditions associated with drug abuse and antisocial
behavior in Hispanic immigrant adolescents (i.e., differential acculturation, family distance/conflict, and
parental isolation). First, we focused on educating
parents about American culture and promoting biculturalism (i.e., endorsement of both Hispanic and
American cultural values) to help parents understand and handle their acculturating adolescents and
the contexts that they encounter (Szapocznik et al.,
1984, 1986). Having shown that such an intervention
could promote parent–adolescent closeness and reduce adolescent behavior problems in Hispanic immigrant families, we proceeded to focus specifically
on fostering communication and negotiation skills to
prevent escalations in family conflict and distance
(Szapocznik, Santisteban et al., 1989b). Finally, having shown that carefully designed prevention programs could reduce both cultural and normative conflicts within the family (as well as adolescent behavior
problems), we turned our attention outside the family, to facilitating social support and reducing isolation
among Hispanic immigrant parents. This third intervention used a group-oriented, participatory format

so that parents in the group would become natural
support sources for one another (Mancilla et al., 2002).
It was also successful in reducing adolescent associations with delinquent peers and increasing parental
support.
An Integrative Prevention Program
Familias Unidas (see Coatsworth et al., 2002a,
for a detailed description of the intervention) combines the strengths of these three prior interventions
to promote parental investment (i.e., positive parenting, involvement, and support) and decrease adolescent behavior problems. This integrative program is
multilevel, addressing processes operating at varying systemic levels: increasing parental investment
within the family, fostering proactive connections between the family and other important systems such
as peers and school, and garnering external support
for parents. The multilevel nature of our intervention
draws upon ecodevelopmental theory (Szapocznik &
Coatsworth, 1999), which proposes that risk and protective processes operating at varying systemic levels
compound one another to create an overall profile
of risk and protection. By promoting protective factors within the family and between the family and
other important systems, our integrative prevention
program represents an attempt to address risks at multiple systemic levels and to prevent those risks from
compounding one another.
Familias Unidas was designed to prevent drug
abuse and antisocial behavior in two ways. First, the
intervention format targets the three family conditions that we have found to be associated with these
negative outcomes. The program is intended to (a) familiarize parents with, and involve them in, the major
extrafamilial contexts in which their adolescents participate (i.e., the peers and school), (b) reinvest parents in their adolescents’ lives by facilitating parent–
adolescent bonding and cohesion, and (c) build
supportive relationships among Hispanic immigrant
parents so that parents would feel less isolated
and would be more likely to be invested in the
lives of their adolescents. Second, the intervention
techniques facilitate protective mechanisms known
to inhibit the development of drug abuse and antisocial behavior in adolescents across ethnic groups
(including Hispanics). Specifically, Familias Unidas
targets increases in parental investment, decreases in
adolescent behavior problems, and increases in adolescent school bonding and academic achievement.
All of these changes have been shown to protect

P1: GRA
Prevention Science [PREV]

pp917-prev-469241

July 8, 2003

14:33

Promoting Parental Investment
adolescents in various ethnic groups against risks for
drug abuse and antisocial behavior (e.g., Ellickson &
Morton, 1999; Forehand et al., 1997; Steinberg et al.,
1994). Moreover, increases in parental investment
have been shown to predict increases in school
bonding/academic achievement (Seitz et al., 1985)
and decreases in adolescent behavior problems
(Pabon, 1998)
This study tested five hypotheses. First, it was hypothesized that, relative to a no-intervention control
condition, families participating in Familias Unidas
would evidence greater increases in parental investment. Second, it was hypothesized that families participating in Familias Unidas would evidence greater
decreases in adolescent behavior problems. Third, it
was hypothesized that families participating in Familias Unidas would evidence greater increases in adolescent school bonding/academic achievement. Fourth, it
was hypothesized that changes in parental investment
would mediate the effects of the intervention on adolescent behavior problems. Fifth, it was hypothesized
that changes in parental investment would mediate
the effects of the intervention on adolescent school
bonding/academic achievement.
METHOD
Design
A mixed design was employed, with intervention
condition (Familias Unidas and no-intervention control) as the between groups factor and time as the
within groups factor. Assessment batteries were administered at baseline (immediately prior to intervention), and at four subsequent 3-month intervals
(i.e., 3, 6, 9, and 12 months). Upon completing the
baseline assessment, the families were stratified by
adolescents’ school and grade and were randomly assigned to either the Familias Unidas intervention or to
the no-intervention control condition in a 60/40 ratio
(60% of the sample was assigned to the experimental
condition to provide power for subsequent analyses
of intervention process). The Familias Unidas intervention lasted approximately 9 months, such that the
9-month assessment coincided with the end of the intervention and the 12-month assessment functioned
as a follow-up.
Participants
Inclusion criteria for the current study were: (a)
Hispanic 6th and 7th grade students with no history of

Style file version Nov. 04, 2000

191
psychiatric hospitalization; (b) residing with at least
one Hispanic immigrant parent willing to participate
in the intervention; and (c) the family had no plans to
relocate out of the catchment area. Participants were
recruited from three middle schools in low-income
areas of Miami. Letters were sent to parents of all
students in the three participating schools, inviting
them to learn more about the program. Of the 475
students whose parents returned the letters, 1.2%
indicated that their adolescents were transferring
to another school outside the catchment area,
69.4% indicated interest in participating, and 29.2%
responded that they were not interested. Of the
330 families who indicated interest in participating,
49% were not included in baseline assessments and
randomization to condition. The primary reasons for
noninclusion were parents’ schedule conflicts (15%),
unresponsiveness to letters and phone calls from
project staff (10%), parents deciding not to enroll in
the program (8%), moving out of the catchment area
(6%), invalid phone numbers or addresses (5%), and
current treatment for family crises or psychological
disorders (5%).
A total of 167 participants/families met inclusion criteria, completed a baseline assessment, and
were randomized to the experimental or control conditions. In total, 96 adolescents and their families (57
with male adolescents and 39 with female adolescents) were assigned to the Familias Unidas condition,
and 71 adolescents (45 males, 26 females) and their
families were assigned to the no-intervention control. The adolescent gender distributions in the two
conditions were not significantly different ( p > .20).
The mean age of the adolescents was 12.40 years
(SD = 0.80 years; range from 10.69 to 14.89 years).
In the experimental condition, each adolescent and
at least one primary caregiver participated in the intervention conjointly.
Consistent with the demographics of the area, the
largest percentage of participants were Cuban (39%),
followed by Central and South Americans (29 and
17%, respectively), and a small proportion of Puerto
Ricans/Dominicans (5%). The remaining 10% identified themselves as “Other” Hispanic. The majority of
parents (94%) and half of the adolescents (49%) were
born outside of the United States. The range of years
living in the United States for parents was from less
than 1 year to 42 years, with a median of 11 years. More
than a quarter (26.3%) of the adolescents had been
residing in the United States for less than 5 years, and
the average length of residence for adolescents was
8.5 years (range 0–14 years). Fifty-seven percent of

P1: GRA
Prevention Science [PREV]

pp917-prev-469241

192

July 8, 2003

14:33

Style file version Nov. 04, 2000

Pantin, Coatsworth, Feaster, Newman, Briones, Prado, Schwartz, and Szapocznik

the families reported speaking only Spanish at home,
and 36% of families reported speaking Spanish and
some English at home. The median annual household
income was between $15,000 and $20,000. The modal
level of parent education was 12th grade (35%), with
a substantial percentage obtaining only elementary or
some high school education (28%) and the remaining
(37%) receiving some post secondary school education including college or graduate school.
Intervention
In Familias Unidas, five general techniques
were used to promote these changes. First, drawing on Freire’s participatory learning model (Freire,
1970/1983), problem posing and participatory exercises were used to promote active parental involvement in intervention activities. Second, group discussions were used to increase parents’ understanding of
their role in protecting their adolescents from harm
and to facilitate parental investment in adolescents.
Third, program activities provided parents with opportunities to interact with adolescents’ peers. These
interactions, in turn, served to connect parents to their
adolescents’ peer network and to facilitate supervision and associations with positive peers. Fourth, visits by adolescents’ school counselors were utilized to
connect parents to their adolescents’ school world and
to promote school bonding and achievement. Fifth,
planned parent–adolescent discussions during homebased family sessions were used to facilitate bonding
within the family (which promotes parental investment) and to solidify the gains achieved in parentgroup sessions.
The intervention was delivered through familycentered, multiparent groups that met weekly for approximately 9 months. Parents attended a mean of
24 group sessions (SD = 13.9). Each group consisted
of 10–12 parents and was led by a trained facilitator.
One parent figure from each participating family attended the group sessions. Each participating parent
received $10 for each session attended. Each group
session lasted approximately 1 hr.
The intervention proceeded in three stages (see
Table 1, and see Coatsworth et al., 2002a, for a detailed description of the intervention’s implementation). Each stage was guided by a set of specific objectives, such that as many sessions were dedicated
to each stage as the facilitator deemed necessary to
meet the objectives (cf. Greenberg & Newman, 1996).
It should be noted that parental investment was the
primary direct target of the intervention activities at

Table 1. Intervention Stages in Familias Unidas
Intervention stage
1. Engagement

2. Promoting parental
investment by introducing the three
primary adolescent
worlds (family, peers,
and school) and by
eliciting
parental
concerns in these
worlds
3. Fostering parenting
skills necessary for
decreasing adolescent problem behavior and increasing
adolescent academic
achievement/school
bonding

Objectives
• Outlining objectives of the intervention
• Using commonalties among
group parents to build support
networks among parents in each
group to reduce isolation
• Validating parents’ concerns
• Eliciting parental goals for improving adolescents’ functioning
in each world
• Placing parents in charge of the
intervention’s direction

• Family world: positive parenting,
involvement, support, behavior
management
• School world: communicating
with school personnel, monitoring homework
• Peer world: Monitoring social activities, establishing management
networks with peers’ parents

all three stages. Parental isolation and differential acculturation were “targets” only inasmuch as they are
thought to contribute to low parental investment.
The first stage was devoted to engaging parents
into the intervention and creating cohesion among the
parents in each group. During the second stage, facilitators attempted to promote parental investment by
introducing three primary adolescent “worlds” (family, peers, and school), eliciting parents’ specific concerns within each world, and assuring parents that
the intervention would be tailored to address their
concerns. Invariably, parents’ issues tended to center
on (a) distance, conflict, and disobedience within the
family; (b) problems with school attendance, performance, or interest; and (c) unsupervised associations
with peers. The third stage involved intervening to foster parenting skills necessary to decrease adolescent
problem behavior and to increase adolescent school
bonding/academic achievement
During this third stage, home visits were interspersed between group sessions to provide parents
with opportunities to work with their adolescents to
implement skills related to each of the three worlds
(e.g., discussing behavior management, peer supervision issues, and homework). Families received a
mean of 2.4 home visits during the course of the
intervention.

P1: GRA
Prevention Science [PREV]

pp917-prev-469241

July 8, 2003

14:33

Style file version Nov. 04, 2000

193

Promoting Parental Investment
Control Condition
Families in the control condition did not receive any formal interventions from project staff. Control families were contacted only for the purpose of
scheduling and completing assessments.

tervention integrity meetings. Fidelity to the Familias
Unidas protocol was excellent. Of the 732 facilitator
interventions rated, 89.4% were prescribed, whereas
10.6% were proscribed.

Procedure
Facilitators
Three female Hispanic master’s-level facilitators
with an average of 5 years’ clinical experience conducted the Familias Unidas group sessions and home
visits. The senior author (H.P.), who is the developer of Familias Unidas, trained and certified facilitators for specific competence in and sensitivity
to both Hispanic and American culture. Facilitators
were also trained to offer support, validation, and
guidance to group parents. Training included didactic
lessons, role-plays, and mock groups. Each trainee was
required to conduct six sessions with one six-family pilot group. The training program spanned three eighthour days. Based on the training experiences in this
study, we have begun to develop a second generation
intervention to streamline the training process and to
facilitate transportability to community settings.

Fidelity to Familias Unidas
Two mechanisms were established to ensure fidelity of the intervention. First, the senior author
supervised and reviewed videotapes of each group
session and provided ongoing corrective feedback to
facilitators. Deviations from protocol and problems
of implementation were identified and discussed with
the facilitators in weekly supervision meetings. Facilitators received one hour of individual supervision and
2 hr of group supervision per week.
Second, to assure intervention adherence, 25%
of all group sessions were randomly selected for
videotape rating by independent adherence raters.
Raters used a standard adherence form to record the
presence or absence of 10 prescribed (e.g., placing parents in leadership roles) and 5 proscribed (e.g., excessive lecturing) facilitator behaviors. Adherence raters
were trained to an interrater reliability intraclass correlation “gold standard” of .80 or above with the second author (J.D.C.). Interrater reliability was reevaluated monthly to control for drift. Adherence problems
identified by raters were discussed with the project
supervisor (H.P.) and the facilitators in biweekly in-

This study was approved by the University of
Miami Institutional Review Board for the Protection
of Human Subjects. Adolescents and their parents
signed assent and consent forms, respectively, prior
to initiating participation.
Assessors were trained to criteria and were monitored on a weekly basis throughout the course of the
study. The assessment procedures were manualized,
and assessors were trained in measure administration,
rapport building, confidentiality, and emergency procedures.
The measures reported in this paper were part
of a larger assessment battery administered to participants. The average completion time for the larger
battery ranged from 45–75 min for parents and for
adolescents. All measures were administered in interview form and in the preferred language of the participant. The majority of parents (62.9%) completed
their assessments in Spanish, whereas the majority
of adolescent assessments were completed in English
(92.7%). Assessors recorded participants’ responses
on laptop computers. Eight adolescent-report measures, and four parent-report measures, were analyzed
for this report. The Spanish versions of the measures
used in this study were established through back translation with committee review and resolution of discrepancies (Kurtines & Szapocznik, 1995), to ensure
that the Spanish and English versions were equivalent in tone, style, conceptual meaning, and content.
Although the measures used in this study were developed for use with mainstream American samples,
assessors did not report any conceptual or cultural
problems in administering the measures to the Hispanic immigrant parents or adolescents in this study.
Recruitment and baseline assessments occurred
in the fall and winter of 6th or 7th grade. Assessment
2 occurred in the spring, Assessment 3 in the summer,
Assessment 4 in the fall of the following school year,
and Assessment 5 in the winter of the following school
year. Participants were engaged into the study and
administered baseline assessments on a rolling basis
during the fall and winter of 6th or 7th grade, and
the timing of subsequent assessments was arranged
according to the date of each participant’s baseline

P1: GRA
Prevention Science [PREV]

pp917-prev-469241

July 8, 2003

194

14:33

Style file version Nov. 04, 2000

Pantin, Coatsworth, Feaster, Newman, Briones, Prado, Schwartz, and Szapocznik

assessment (i.e., as close to 3-month intervals as possible). Families were compensated for each assessment
that they completed.

Measures
This section is organized in terms of three outcome composites: parental investment, adolescent
behavior problems, and adolescent school bonding/academic achievement. Each composite was derived by summing adolescent and parent reported
scales. Each scale was standardized, using the baseline
mean and standard deviation, prior to summation.

Parental Investment
As noted in the introduction, parental investment encompasses various positive aspects of parenting, such as encouragement, validation, support, and
involvement. These positive aspects of parenting can
operate within the family, between the parents and
school, and between the parents and prosocial adolescent peers (Coatsworth et al., 2002b).
The parental investment composite was measured as the equally weighted sum of five scales: (a)
the parent and adolescent reported Extent of Involvement and Positive Parenting subscales from the Parenting Practices Scale (Gorman-Smith et al., 1996), a
25-item measure assessing various dimensions of parenting; and (b) the adolescent-reported Family Support subscale from the Social Support Appraisal Scale
(Dubow & Ulman, 1989), a 31-item measure assessing
perceived support from family, peers, and teachers. Internal consistency reliability for the parental investment composite in the current study was α = .90.

Adolescent Behavior Problems
Behavior problems are defined as conduct destructive to oneself or others (Loeber et al., 1998). Our
behavior problems composite includes intrapersonal
(e.g., impulsivity) and interpersonal (e.g., aggression)
aspects of functioning (cf. Cicchetti et al., 1991; Colder
& Chassin, 1993).
The adolescent behavior problems composite
was measured as the equally weighted sum of eight
parent and adolescent reported scales. The adolescent
reports assess both internal states and observable behaviors. Parent reports were obtained only from a be-

havioral perspective (e.g., “problem behavior”), because parent reports of teens’ internal states require
excessive inference on the part of the parent (Kazdin
et al., 1983). Eight scales were used to create the
summed composite. Parent reports were obtained using the Conduct Disorder, Socialized Aggression, Attention Problems, and Motor Excess subscales from
the Revised Behavior Problem Checklist (Quay &
Peterson, 1987), an 89-item measure of child behavior problems. Adolescent reports were obtained from
(a) the Anger Control and Hyperactivity subscales of
the Conners-Wells Self-Report Scale (Conners et al.,
1997), a 27 item measure of self-control and restlessness; (b) the Aggression subscale from the Interpersonal Competence Inventory (Cairns et al., 1994), a
21-item measure assessing aggression, popularity, and
academic competence; and (c) the Behavior Scale
Part I (Resnicow, 1997), a 13-item index assessing the
frequency of deviant and antisocial behaviors. The internal consistency reliability estimate for the adolescent behavior problems composite in this sample was
α = .96.

Adolescent School Bonding/Academic Achievement
School bonding is defined as interest in academic
pursuits and attending school willingly (Newcomb
et al., 2002). School bonding and academic achievement are defined as adjusting positively to school
and completing schoolwork that merits high academic
grades (Seitz et al., 1985). The adolescent school
bonding/academic achievement composite was measured as the equally weighted sum of 10 scales: (a)
the parent and adolescent reported School Bonding, School Achievement, and Disinterest in School
subscales from the School Attitudes/Bonding Scale
(Resnicow, 1996), a 35-item instrument that assesses adolescents’ connections to school, teachers
and learning; (b) the parent and adolescent reported
Academic Achievement subscales from the Adolescent Competence Scale (Coatsworth, 1992), an 18item scale that assesses adolescent and parent reports of the adolescent’s competence in academic
achievement, peer relationships, involvement in activities, and classroom behavior; (c) the adolescentreported Academic Competence subscale from Interpersonal Competence Scale (Cairns et al., 1994),
and (d) the Intellectual/School Self-Concept subscale
from the Piers–Harris Children’s Self-Concept Scale
(Piers, 1984), an 80-item measure that assesses adolescent self-concept in six domains (behavior, academics,

P1: GRA
Prevention Science [PREV]

pp917-prev-469241

July 8, 2003

14:33

Style file version Nov. 04, 2000

195

Promoting Parental Investment
popularity, happiness, physical appearance, and anxiety). The internal consistency reliability estimate for
the adolescent school bonding/academic achievement
composite in this sample was α = .80.
RESULTS
Data Analytic Strategy
Hypotheses 1, 2, and 3 were tested using mixedmodel analyses of variance, focusing on the condition
by time interaction. Mixed model methods (Littell
et al., 1996) contain both a fixed effect portion (i.e.,
the observed predictor(s), represented as a repeated
measures ANOVA) and a random effects component,
which can account for systematic error associated with
the nonindependence of nested observations within a
family (cf. Bryk & Raudenbush, 1992). All analyses
were conducted on an intent-to-treat basis; all participants were included in the analyses regardless of the
number of sessions or assessments they completed.
Student’s t-tests were used to determine the presence or absence of baseline differences in each outcome variable. The Type I error rate for these tests was
set at a conservative value of α = .20. In cases where
baseline differences by condition emerged, baseline
scores were covaried in subsequent analyses, and subsequent analyses were restricted to the remaining assessment points.
An Expectation Maximization (EM) algorithm
(Little & Rubin, 1987) was used to impute any missing
responses for each variable at each assessment time,
provided that the adolescent and/or parent provided
some valid data at that assessment point. Retention
rates were high, with 95% of participants completing
assessments at three or more time points and 98%
of the original baseline sample completing the final
12-month assessment. Families in the control condition were no more likely to miss an assessment point
(13.8%) than were those in the experimental condition, 10.0%, χ 2 (1, N = 835) = 2.87, ns.

Testing Hypotheses 4 involved ascertaining
whether Baron and Kenny’s criteria for mediation
were satisfied (Baron and Kenny, 1986). To evaluate these criteria, the mixed-model ANOVA for adolescent behavior problems was rerun, with change in
parental investment entered as an additional covariate. If, in addition to Hypotheses 1 and 2 being supported, the effect of the covariate on the behavior
problems was significant, and if the direct effect of
the intervention was reduced to nonsignificance, then
all four criteria for mediation would be met. Hypothesis 5 was tested in a similar way, with school bonding/
academic achievement as the outcome variable.

Baseline Differences in the Dependent
Measures by Condition
Table 2 provides descriptive statistics by condition for the parental investment, adolescent behavior
problems, and school bonding/academic achievement
measures at baseline. Only one baseline mean difference was statistically significant: the Familias Unidas
condition evidenced lower mean levels of adolescent
behavior problems than did the control condition,
t(165) = 1.32, p < .19. Thus, baseline scores for this
composite were covaried in the analyses that follow.

Tests of Intervention Effects
Parental Investment
Mixed model analyses of variance revealed a significant Time × Condition interaction on parental investment, F(4, 577) = 2.68, p < .04 (see Fig. 1). As
evidenced by the figure, the control condition trajectory was more positive than was the experimental condition trajectory between baseline and 3 months, but
the control condition trajectory flattened out and began to decrease sharply at 9 months. The experimental

Table 2. Baseline Standardized Score Means by Condition
Experimental condition
(n = 96)

Control condition
(n = 71)

Test for
Differences

Measure

Mean

SD

Mean

SD

t

p

Parental investment
Behavior problems
School bonding/
academic achievement

−1.21
1.03
−0.61

3.52
5.12
6.91

−.99
2.12
−0.54

3.27
5.54
6.86

0.40
1.32
0.07

.69
.19
.95

P1: GRA
Prevention Science [PREV]

pp917-prev-469241

July 8, 2003

196

14:33

Style file version Nov. 04, 2000

Pantin, Coatsworth, Feaster, Newman, Briones, Prado, Schwartz, and Szapocznik
School Bonding/Academic Achievement
A mixed model analysis of covariance did not
reveal a significant Time × Condition interaction on
school bonding/academic achievement, F(4, 576) =
1.11, p < .35.

Attendance Effects

Fig. 1. Parental investment by condition and time.

condition evidenced its greatest increase between 3
and 6 months. Although it too began to decrease between 9 and 12 months, this decrease was milder than
was that in the control condition.

Adolescent Behavior Problems
A mixed-model analysis of covariance (with
baseline scores covaried) revealed a significant Time
× Condition interaction on adolescent behavior problems, F(3, 424) = 4.25, p < .006 (see Fig. 2). The experimental condition evidenced a steady decline in
behavior problems, whereas the control condition evidenced a sharp increase between 3 and 6 months before decreasing sharply.

Fig. 2. Adolescent behavior problems by condition and time.

To determine whether intervention dosage was
a significant predictor of change in outcome variables within the experimental condition, the number
of group sessions that each family attended was interacted with time in growth curve models for the two
outcome composites that differed significantly by condition (i.e., parental investment and adolescent behavior problems). Linear and quadratic models were
estimated for both outcome composites.
A significant linear Time × Attendance interaction emerged for parental investment, F(1, 98) =
4.70, p < .04. Additionally, there was a significant
quadratic time trend, F(1, 91) = 8.43, p < .005, but
no quadratic Time × Attendance interaction. The attendance analysis evidenced a clear dose-response
effect; parents with lower baseline levels of investment tended to attend more intervention sessions
and to display greater increases in investment during the intervention. No linear or quadratic Time ×
Attendance interactions emerged for adolescent behavior problems. For parental investment, Fig. 3 displays the best-fitting quadratic lines for four representative levels of attendance (0–9, 10–19, 20–29, and
30+ sessions). These levels were chosen because they
represented approximately 25, 50, 75, and 100% of
the total possible number of sessions. These dosage
levels also somewhat approximated the distribution
of sessions attended (0–9, 37.5%; 10–19, 17.71%; 20–
29, 29.17%; and 30 or more, 15.63%).

Fig. 3. Parental investment by attendance and time.

P1: GRA
Prevention Science [PREV]

pp917-prev-469241

July 8, 2003

14:33

Style file version Nov. 04, 2000

197

Promoting Parental Investment
Mediating Role of Change in Parental Investment
on Change in Adolescent Behavior Problems
To test the hypothesis that change in parental
investment would mediate the effects of the intervention on adolescent behavior problems, a Time ×
Condition mixed-model analysis of covariance was
conducted on the adolescent behavior problems composite,5 with change in parental investment included
as an additional covariate (along with baseline adolescent behavior problems). For each step in the analysis, changes in parental investment between consecutive assessment intervals (e.g., between baseline and
3 months) were used to predict adolescent behavior problems scores at the end of that interval (e.g.,
3 months). This procedure was repeated for the remaining time intervals (i.e., 3–6 months, 6–9 months,
and 9–12 months). This consecutive-intervals analytic
strategy allows for the incorporation of data from all
assessment points and allows for the nonlinear pattern of change in parental investment over time (cf.
Blaney et al., 1997, regarding the analytic strategy).
Results indicated that change in parental investment
between consecutive assessment intervals was significantly related to subsequent levels of adolescent behavior problems, F(1, 378) = 15.30, p < .0001. However, the Time × Condition interaction on adolescent
behavior problems remained statistically significant,
F(3, 372) = 3.60, p < .02, indicating that the fourth
criterion for mediation was not met.

DISCUSSION
Findings from this study suggest that the Familias Unidas intervention performed largely as hypothesized with regard to parental investment and to
adolescent behavior problems, but not with regard
to adolescent school bonding/academic achievement.
Results using an intent-to-treat design indicated that
parents participating in the Familias Unidas intervention showed modestly but significantly greater improvements in parental investment compared to those
in the control condition. Moreover, families participating in Familias Unidas reported consistent reductions in adolescent behavior problems, whereas control group families reported somewhat inconsistent

5

A similar analysis was not conducted for adolescent school bonding/academic achievement because the effect of the intervention
on this outcome composite was not statistically significant.

changes over time. No condition difference emerged
for adolescent school bonding.

Parental Investment
As noted earlier, parental investment (the combination of positive parenting, involvement, and support) has been found to be protective against adolescent problem behaviors (Barnes et al., 1995; Costa
et al., 1999; Marshal & Chassin, 2000). The present
findings demonstrate the feasibility of experimentally
manipulating parental investment. Moreover, the relationships between change in parental investment
and subsequent levels of adolescent behavior problems suggest that parent-centered interventions such
as Familias Unidas have the potential to decrease risk
factors for later substance abuse and delinquency in
Hispanic immigrant adolescents, and to counter the
lack of parental investment often created as a result of immigration and acculturation (cf. Szapocznik
et al., 1984). This finding supports the pivotal role
of parental investment in adolescent development.
However, despite the relationship between change in
parental investment and subsequent levels of adolescent behavior problems, changes in parental investment did not fully mediate intervention effects on
adolescent behavior problems. It appears that, in addition to influencing adolescent problem behaviors
indirectly through parental investment, the intervention may have exerted a direct influence on adolescent
problem behaviors through aspects of parenting that
were not measured in the assessment battery, such
as establishment of supervisory networks with peers’
parents. Additionally, changes in parental investment
predicted changes in adolescent behavior problems
in both the experimental and control conditions. This
suggests that parental investment may exert a strong
influence on adolescent problem behaviors, regardless of whether parental investment is experimentally
manipulated or is changing naturally over time.
For families participating in Familias Unidas,
changes in parental investment was directly related
to the number of parent-group sessions attended.
The over-time relationship between attendance and
changes in parental investment suggested a doseresponse effect. Parents with the lowest initial reported investment levels attended the greatest numbers of group sessions and appeared to benefit most
from the Familias Unidas intervention.
The increases in parental investment observed
in the control condition were unexpected and may

P1: GRA
Prevention Science [PREV]

pp917-prev-469241

July 8, 2003

198

14:33

Style file version Nov. 04, 2000

Pantin, Coatsworth, Feaster, Newman, Briones, Prado, Schwartz, and Szapocznik

have resulted from repeated assessments every three
months inquiring in detail about parents’ involvement
with their adolescents. These assessments may have
served as an “intervention” to increase parent reports
of investment. We also cannot rule out the possibility
that the increases in parental investment in the control condition may have resulted from contamination,
given that experimental and control families lived in
the same neighborhoods and their children attended
the same schools.
Adolescent Behavior Problems
The Familias Unidas intervention was also successful in decreasing adolescent behavior problems.
This finding lends support to the use of parentcentered interventions to ameliorate adolescent behavior problems and to decrease the likelihood of
more severe conduct problems and substance involvement later in adolescence.
Although the experimental condition showed a
decrease in adolescent behavior problems over the
length of the study, the control condition appeared
to evidence a similar pattern of decline. However, in
the control group, adolescent behavior problems increased sharply between 3 and 6 months before decreasing again at 9 months. Closer examination of
this apparent anomaly in the control condition may
suggest an interaction between intervention participation and enrollment in school. The sharp increase
in adolescent behavior problems in the control condition at 6 months coincided with most adolescents’
summer vacation (at 6 months, 85% of experimental condition assessments, and 82% of control condition assessments, were conducted while the adolescents were not in school). In low-income immigrant
families where parents are unfamiliar with or cannot
afford structured adolescent activities (cf. Black &
Krishnakumar, 1998), the summer vacation period is
likely to translate into unstructured and unsupervised
time for teens. Unsupervised time, in turn, poses risks
for engagement in problematic behaviors (Pettit et al.,
1999). The Familias Unidas parenting skills may have
helped to counter these summer vacation effects on
adolescents’ behavior problems. Although not specifically hypothesized, it is reasonable to speculate that
the intervention effects would be most evident during
this time of greatest challenge. The protective value of
Familias Unidas was evidenced at the summer assessment point, at which the level of adolescent behavior
problems reported in the control condition was almost
five times that reported in the intervention condition.

Adolescent School Bonding/Academic Achievement
Although it was hypothesized that the intervention would promote adolescent school bonding and
academic achievement, results indicated that intervention adolescents did not evidence gains over those
reported by control adolescents. Targeting the adolescent’s school world more directly and extensively,
through activities such as joint parent–adolescent
meetings with school counselors, may be necessary
to facilitate adolescent school bonding. Working indirectly through parent mobilization may be insufficient
to improve teens’ academic performance and interest.

Attendance Effects
Two important findings emerged with regard to
parent-group session attendance. First, the number
of sessions attended was inversely related to baseline parent-reported levels of investment. Parents reporting less investment in their adolescents tended
to attend more sessions than did parents reporting
higher investment levels. Second, baseline reported
investment levels were inversely related to increases
in parent-reported investment during the intervention. Parents lower in reported investment at baseline improved to a greater extent than did parents
reporting higher levels of investment. Moreover, the
attendance effects are clearly not evidence of regression to the mean or of ceiling effects, given that (a)
parents at all levels of baseline investment tended to
report increases in investment over time and (b) mean
parental investment scores did not approach the highest possible value at any time point or for any attendance level. Two important conclusions can be drawn
from the attendance effects. First, the fact that less invested parents attended more sessions and evidenced
greater improvements in investment reflects the ability of Familias Unidas to facilitate investment in the
neediest parents. The degree of increase in investment
was clearly a function of the number of sessions attended, possibly reflecting a tendency for parents to
stop attending sessions once they had reached a specific or desired level of investment. In spite of the
fact that parental investment tended to increase during the intervention regardless of baseline levels, the
different rates of attendance and change trajectories
in parental investment associated with various baseline investment levels suggest that the intervention is
likely to be most effective with low-investment, immigrant Hispanic parents.

P1: GRA
Prevention Science [PREV]

pp917-prev-469241

July 8, 2003

14:33

Promoting Parental Investment
Limitations and Future Directions
Several limitations warrant discussion. First, although a number of parenting dimensions have been
enumerated as risk and protective factors for adolescent drug abuse and problem behavior, only a small
subset of these dimensions were targeted in Familias Unidas and included in the assessment battery.
Further, only increases in positive dimensions of parenting (positive parenting, involvement and support),
and not decreases in negative or coercive parenting,
were measured.
Second, and more generally, the exclusive use
of questionnaire measures may have introduced bias
into the results (cf. Magura & Kang, 1996). However,
the fact that each construct was comprised of measures from multiple reporters (i.e., both parent and
adolescent reports) is likely to correct for some of this
bias (cf. Verhulst & van der Ende, 1991). Still, because
the level of questionnaire bias and cross-informant
agreement in Hispanic immigrant parents and adolescents is not known, the possibility of bias cannot
be discounted.
Third, it would be useful to assess changes in peer
affiliations over time. This would allow for analysis of
the predictive relationship between changes in peer
affiliations/social activities and decreases in adolescent behavior problems.
Fourth, the use of a second control group, in
which parents meet in groups but do not discuss
specific parenting skills, would have provided a control for facilitator, expectancy, group, or attention
effects. There are advantages and disadvantages to
selecting a no-intervention control condition. For
example, although a no-intervention control condition provides maximum ecological validity, the nointervention control may neglect to control for other
potential non-specific confounds such as effects of
participating in a group, level of attention, facilitator
qualities and enthusiasm, or participant expectancies
(Szapocznik et al., 1989a).
Fifth, the intervention length (9 months) and
depth of training required may inhibit transportability to practice-based settings. Future interventions using the parent-centered participatory learning format
should be shorter and further standardized to facilitate ease of training and implementation. To achieve a
more compact intervention, it may be useful to utilize
a combination of didactic and participatory exercises.
In this type of model, within each set of exercises, for
example, parent-group participatory learning discussions would follow didactic presentations of key issues

Style file version Nov. 04, 2000

199
and parenting skills (cf. Pantin et al., 2002). Also, the
number and content of group sessions and home visits would have to be standardized. Such a streamlined and manualized version of Familias Unidas is
currently under development (Pantin et al., in press).
Sixth, the results may have been compromised
by the low enrollment rate. Of the 475 parents who
were sent letters, only 330 responded that they were
interested in participating. Of these families, only 51%
(167 of 330) were assessed and randomized. Hence,
of the initial sample who were sent letters, only 35%
(167 of 475) actually participated. This may raise concerns about the generalizability of the findings. Similar
problems with parent participation in preventive interventions have been previously reported (Spoth &
Redmond, 1996).
Seventh, the stratified randomization procedure
utilized to assign participants to conditions may have
created the possibility of contamination. We randomized participants to condition within each school,
rather than assigning separate schools to the experimental and control conditions, to ensure that both
conditions were drawn from the same population. For
this reason, we cannot discount the possibility that experimental condition adolescents and/or parents may
have discussed the intervention activities with control
condition adolescents and/or parents. In addition, the
positive changes found in the control group may be
partially explained by the fact that this sample may
have been comprised of self-selected families who
were highly interested in the intervention.
Finally, it would be useful to examine the maintenance of intervention effects. Further data collection points would allow for examination of the longterm trajectory of parental investment and adolescent
problem behaviors following intervention termination. Assessing participants again in high school, when
drug use and delinquency are most likely to emerge
(e.g., Moffitt, 1993), would be useful to allow for comparison of drug use and delinquency rates between
the experimental and control conditions, and to ascertain the prospective relationship between increases in
parental investment and adolescents’ subsequent initiation of drug use and delinquency. Measuring the
longer-term maintenance of intervention gains is also
an important aspect of program evaluation; for example, if it was found that parental investment continued
to decrease after the end of the intervention, it might
be necessary to add periodic follow-up booster sessions at regularly scheduled intervals.
Despite these limitations, this study investigated
the effects of a preventive intervention with an

P1: GRA
Prevention Science [PREV]

200

pp917-prev-469241

July 8, 2003

14:33

Style file version Nov. 04, 2000

Pantin, Coatsworth, Feaster, Newman, Briones, Prado, Schwartz, and Szapocznik

understudied but growing population, making these
findings particularly valuable. The results of this study
demonstrate the promise and potential of a parentcentered preventive intervention for Hispanic immigrant families with adolescents at possible risk
for drug abuse and other problem behaviors. The
increases in parental investment and decreases in
adolescent behavior problems provide evidence that
Familias Unidas reduced some undesirable “side effects” of immigration and helped to reinvolve Hispanic immigrant parents in their adolescents’ lives.

ACKNOWLEDGMENTS
This study was funded by Center for Substance Abuse Prevention Grant 1 UR6 SPO7961 and
National Institute of Mental Health Grant 1 R01
MH61143, both to Jose´ Szapocznik, Principal Investigator. We thank the three facilitators in this study:
Cecilia Ferro, Dolores Perdomo, and Monica Zarate.
We also thank Shenandoah, South Miami, and Ponce
de Leon Middle Schools, from which participant families were recruited.

REFERENCES
Barnes, G. M., Farrell, M. P., & Banerjee, S. (1995). Family influences on alcohol abuse and other problem behaviors among
Black and White adolescents in a general population sample.
In G. M. Boyd, J. Howard, & R. A. Zucker (Eds.), Alcohol
problems among adolescents: Current directions in prevention
research (pp. 13–32). Hillsdale, NJ: Erlbaum.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator
variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182.
Black, M. M., & Krishnakumar, A. (1998). Children in low-income,
urban settings: Interventions to promote mental health and
well-being. American Psychologist, 53, 635–646.
Blaney, N. T., Goodkin, K., Feaster, D., Morgan, R., Millon, C.,
Szapocznik, J., & Eisdorfer, C. (1997). A psychosocial model
of distress over time in early HIV: The role of life stressors,
social support and coping. Psychology and Health, 12, 633–
653.
Bryk, A., & Raudenbush, S. W. (1992). Hierarchical linear models:
Methods and data analysis. Newbury Park, CA: Sage.
Cairns, R. B., Leung, M. C., Gest, S. D., & Cairns, B. D. (1994). A
brief method for assessing social development: Structure, reliability, stability, and developmental validity of the Interpersonal Competence Scale. Behaviour Research and Therapy, 33,
725–736.
Cicchetti, D., Ganiban, J., & Barnett, D. (1991). Contributions from
the study of high-risk populations to understanding the development of emotion regulation. In J. Garber & K. A. Dodge
(Eds.), The development of emotion regulation and dysregulation: Cambridge studies in social and emotional development
(pp. 15–48). New York: Cambridge University Press.

Coatsworth, J. D. (1992). Project Competence: A developmental
approach to defining and assessing competence. The Community Psychologist, 26, 19–21.
Coatsworth, J. D., Pantin, H., McBride, C., Briones E., Kurtines,
W., & Szapocznik, J. (2002b). Ecodevelopmental Correlates
of Behavior Problems in Young Hispanic Females. Applied
Developmental Science, 6(3), 126–143.
Coatsworth, J. D., Pantin, H., & Szapocznik, J. (2002a). Familias
Unidas: A family centered ecodevelopmental intervention to
reduce risk for problem behavior among Hispanic adolescents.
Clinical Child and Family Psychology Review, 5(2), 113–132.
Colder, C. R., & Chassin, L. (1993). The stress and negative affect
model of adolescent alcohol use and the moderating effects
of behavioral undercontrol. Journal of Studies on Alcohol, 54,
326–333.
Conduct Problems Prevention Research Group (2000). Merging
universal and indicated prevention programs: The Fast Track
model. Addictive Behaviors, 25, 913–927.
Conners, C. K., Wells, K. C., Parker, J. D., Sitarenios, G., Diamond,
J. M., & Powell, J. W. (1997). A new self report scale for assessment of adolescent psychopathology: Factor structure, reliability, validity, and diagnostic sensitivity. Journal of Abnormal
Psychology, 25, 487–497.
Costa, F. M., Jessor, R., & Turbin, M. S. (1999). Transition into
adolescent problem drinking: The role of psychosocial risk and
protective factors. Journal of Studies on Alcohol, 60, 480–490.
Dishion, T. J., & Kavanagh, K. (2000). A multilevel approach to
family-centered prevention in schools: Process and outcome.
Addictive Behaviors, 25, 899–911.
Dubow, E. F., & Ullman, D. G. (1989). Assessing social support
in elementary school children: The Children’s Social Support
Survey. Journal of Clinical Child Psychology, 18, 52–64.
Forehand, R., Miller, K. S., Dutra, R., & Chance, M. W. (1997).
Role of parenting in adolescent deviant behavior: Replication
across and within two ethnic groups. Journal of Consulting and
Clinical Psychology, 65, 1036–1041.
Freire, P. (1983). Pedagogy of the oppressed (M. B. Ramos, trans.).
New York: Herder & Herder. (Original work published 1970).
Gil, A. G., Wagner, E. F., & Vega, W. A. (2000). Acculturation,
familism, and alcohol use among Latino adolescent males:
Longitudinal relations. Journal of Community Psychology, 28,
443–458.
Gorman-Smith, D., Tolan, P. H., Zelli, A., & Huesmann, L. R.
(1996). The relation of family functioning to violence among
inner-city minority youth. Journal of Family Psychology, 10,
115–129.
Greenberg, L. S., & Newman, F. L. (1996). An approach to psychotherapy change process research: Introduction to the special section. Journal of Consulting and Clinical Psychology, 64,
435–438.
Kazdin, A. E., French, N. H., & Unis, A. S. (1983). Child, mother,
and father evaluations of depression in psychiatric inpatient
children. Journal of Abnormal Child Psychology, 11, 167–180.
Kurtines, W. M., & Szapocznik, J. (1995). Cultural competence
in assessing Hispanic youths and families: Challenges in the
assessment of treatment needs and treatment evaluation for
Hispanic drug abusing adolescents. In E. Rahdert, & D.
Czechowicz (Eds.), Adolescent drug abuse: Clinical assessment and therapeutic interventions (NIDA Monograph no.
156, pp. 172–189). Rockville, MD: National Institute on Drug
Abuse.
Littel, R. C., Milliken, G. A., Stroup, W. W., & Wolfinger, R. (1996).
SAS system for mixed models. Cary, NC: SAS Institute.
Little, R. J. A., & Rubin, D. B. (1987). Statistical analysis with missing
data. New York: Wiley.
Loeber, R., Farrington, D. P., Stouthamer-Loeber, M., Van
Kammen, W. B. (1998). Multiple risk factors for multiproblem boys: Co-occurrence of delinquency, substance abuse, attention deficit, conduct problems, physical aggression, covert

P1: GRA
Prevention Science [PREV]

pp917-prev-469241

July 8, 2003

14:33

Promoting Parental Investment
behavior, depressed mood, and shy/withdrawn behavior. In
R. Jessor (Ed.), New perspectives on adolescent risk behavior
(pp. 90–149). New York: Cambridge University Press.
Lollock, L. (2001). The foreign born population in the United
States, March 2000 (Current Population Report P20–534).
Washington, DC: U.S. Census Bureau.
Magura, S., & Kang, S. Y. (1996). Validity of self-reported drug use
in high risk populations: A meta-analytical review. Substance
Use and Misuse, 31, 1131–1153.
Mancilla Y. E., Newman, F. L., Tejeda, M. J., Zarate, M., &
Szapocznik, J. (2002). Padres Lideres: Efficacy of an intervention to promote parent leadership in Hispanic immigrant families. Manuscript submitted for publication.
Marshal, M. P., & Chassin, L. (2000). Peer influence on adolescent
alcohol use: The moderating role of parental support and discipline. Applied Developmental Science, 4, 80–88.
Moffitt, T. E. (1993). Adolescence-limited and life-coursepersistent antisocial behavior: A developmental taxonomy.
Psychological Review, 100, 674–701.
National Institute on Drug Abuse [NIDA] (2001). Monitoring the
Future 2000, Vol. 1: Secondary school students. Rockville, MD:
Department of Health and Human Services.
Newcomb, M. D., Abbott, R. D., Catalano, R. F., Hawkins, J. D.,
Bettin-Pearson, S., & Hill, K. (2002). Mediational and deviance theories of late high school failure: Process roles of
structural strains, academic competence, and general versus
specific problem behaviors. Journal of Counseling Psychology,
49, 172–186.
Pabon, E. (1998). Hispanic adolescent delinquency and the family:
A discussion of sociocultural influences. Adolescence, 33, 941–
955.
Palmer, E. J., & Hollin, C. R. (2001). Sociomoral reasoning, perceptions of parenting, and self-reported delinquency in adolescents. Applied Cognitive Psychology, 15, 85–100.
Pantin, H., Schwartz, S. J., Coatsworth, J. D., Briones, E., &
Szapocznik, J. (in press). Familias Unidas: A systemic, parentcentered approach to preventing problem behavior in Hispanic adolescents. In P. H. Tolan, J. Szapocznik, & S. Sambrano
(Eds.), Preventing substance abuse. Washington, DC: APA.
Pantin, H., Schwartz, S. J., Sullivan, S., & Szapocznik, J. (2003). An
ecodevelopmental approach to HIV prevention in Hispanic
immigrant adolescents: Implications for designing prevention
programs. Manuscript submitted for publication.
Pettit, G. S., Bates, J. E., & Dodge, K. A. (1997). Supportive parenting, ecological context, and children’s adjustment: A sevenyear longitudinal study. Child Development, 68, 908–923.
Pettit, G. S., Bates, J. E., Dodge, K. A., & Meece, D. W. (1999).
The impact of after-school peer contact on early adolescent
externalizing problems is moderated by parental monitoring,
perceived neighborhood safety, and prior adjustment. Child
Development, 70, 768–778.
Piers, E. V. (1984). Piers-Harris Children’s Self-Concept Scale
(The way I feel about myself): Revised manual. Los Angeles:
Western Psychological Services.
Quay, H. C., & Peterson, D. R. (1987). Manual for the Revised
Behavioral Problem Checklist. Unpublished manuscript, University of Miami, Coral Gables, FL.
Resnicow, K. (1996). School Attitudes/Bonding Scale. Unpublished
measure, Emory University, Atlanta, GA.

Style file version Nov. 04, 2000

201
Resnicow, K. (1997). Behavior scale for inner city minority youth.
Unpublished measure, Emory University, Atlanta, GA.
Rosenstein, D. S., & Horowitz, H. A. (1996). Adolescent attachment and psychopathology. Journal of Consulting and Clinical
Psychology, 64, 244–253.
Seitz, V., Rosenbaum, L. K., & Apfel, N. H. (1985). Effects of a
family support intervention: A ten-year follow-up. Child Development, 56, 376–391.
Snyder, H. N., & Sickmund, M. (1999). Juvenile offenders and victims: 1999 National Report. Pittsburgh, PA: Office of Juvenile
Justice and Delinquency Prevention.
Spoth, R. L., & Redmond, C. (1996). A theory based parent competency model of protective parenting processes incorporating intervention attendance effects. Family Relations, 45, 139–
147.
Steinberg, L., Fletcher, A., & Darling, N. (1994). Parental monitoring and peer influences on adolescent substance use. Pediatrics,
93, 1060–1064.
Steinberg, L., Lamborn, S. D., Darling, N. A., Mounts, N. S., &
Dornbusch, S. M. (1994). Over-time changes in adjustment and
competence among adolescents from authoritative, authoritarian, indulgent, and neglectful families. Child Development, 65,
754–770.
Szapocznik, J., & Coatsworth, J. D. (1999). An ecodevelopmental
framework for organizing the influences on drug abuse: A developmental model of risk and protection. In M. Glantz & C. R.
Hartel (Eds.) Drug abuse: Origins and Interventions (pp. 331–
366). Washington, DC: American Psychological Association.
Szapocznik, J., Rio, A. T., Murray, E., Cohen, R., Scopetta, M. A.,
Rivas-Vazquez, A., Hervis, O., Posada, V., & Kurtines, W.
(1989a). Family versus child therapy for problematic Hispanic
boys. Journal of Consulting and Clinical Psychology, 57, 571–
578.
Szapocznik, J., Rio, A., Perez-Vidal, A., Kurtines, W., Hervis, O.,
& Santisteban, D. (1986). Bicultural Effectiveness Training
(BET): An experimental test of an intervention modality for
families experiencing intergenerational/intercultural conflict.
Hispanic Journal of Behavioral Sciences, 8, 303–330.
Szapocznik, J., Santisteban, D., Kurtines, W., Perez-Vidal, A., &
Hervis, O. (1984). Bicultural Effectiveness Training: A treatment intervention for enhancing intercultural adjustment in
Cuban American families. Hispanic Journal of Behavioral Sciences, 6, 317–344.
Szapocznik, J., Santisteban, D., Rio, A., Perez-Vidal, A.,
Santisteban, D. A., & Kurtines, W. M. (1989b). Family Effectiveness Training: An intervention to prevent drug abuse and
problem behavior in Hispanic adolescents. Hispanic Journal
of Behavioral Sciences, 11, 3–27.
Tarter, R., Vanyukov, M., Giancola, P., Dawes, M., Blackson, T.,
Mezzich, A., & Clark, D. B. (1999). Etiology of early age onset
substance use disorder: A maturational perspective. Development and Psychopathology, 11, 657–683.
Vega, W. A., & Gil, A. G. (1999). A model for explaining drug use
behavior among Hispanic adolescents. Drugs and Society, 14,
57–74.
Verhulst, F. C., & van der Ende, J. (1997). Assessment of child
psychopathology: Relationships between different methods,
different informants and clinical judgment of severity. Acta
Psychiatrica Scandinavica, 84(2), 155–159.