Barabás Katalin Egészségfejlesztés Könyv
Barabás Katalin Egészségfejlesztés Könyv
Barabás Katalin Egészségfejlesztés Könyv
Susan M. Sheridan
Elizabeth Moorman Kim
Editors
Family-School
Partnerships in
Context
Research on Family-School Partnerships
Volume 3
Series Editors
Susan M. Sheridan
Nebraska Center for Research on Children, Youth, Families and Schools
University of Nebraska-Lincoln
Lincoln, Nebraska, USA
Elizabeth Moorman Kim
Nebraska Center for Research on Children, Youth, Families and Schools
University of Nebraska-Lincoln
Lincoln, Nebraska, USA
Family-School Partnerships
in Context
Editors
Susan M. Sheridan Elizabeth Moorman Kim
Nebraska Center for Research on Children, Nebraska Center for Research on Children,
Youth, Families and Schools Youth, Families and Schools
University of Nebraska-Lincoln University of Nebraska-Lincoln
Lincoln, NE, USA Lincoln, NE, USA
The motivation for the writing of this series grew out of a meeting of the
Interdisciplinary Alliance for Partnerships Research in 2010. We acknowledge the
countless contributions these eminent scholars made to the conceptualization,
development, and realization of these volumes. We thank the sponsors of this meet-
ing, the National Science Foundation (#0921266) and the American Educational
Research Association. The opinions expressed herein belong to the grantees and do
not reflect those of the funding agencies. The editors are indebted to Marj McKinty
for her help through all stages of the preparation of these volumes.
vii
Contents
ix
Contributors
xi
xii Contributors
xiii
Chapter 1
Complexities in Field-Based Partnership
Research: Exemplars, Challenges,
and an Agenda for the Field
The benefits of engaging families in children’s education are among the most
convincing and consistent findings in the educational literature. There is over-
whelming research support attesting to the effects of parent involvement and
family–school partnership practices to positively advance children’s educational
trajectories. Trends in the research suggest that family-based educational interven-
tions appear to be most successful when they (a) involve collaborative partnering
between families and schools; (b) promote healthy relationships between families
and schools, and parents and children; and (c) use evidence-based parent and teacher
practices. Two family–school partnership interventions, Getting Ready (Sheridan,
Marvin, Knoche, & Edwards, 2008) and Conjoint Behavioral Consultation (CBC;
Sheridan & Kratochwill, 2008) exemplify these current trends. In this chapter, we
describe our experiences conducting partnership research as we have developed and
tested these models over the last 25 years. We first define our approach to partner-
ship intervention research by distinguishing between parental involvement practices
and family–school partnerships. We discuss a sample of our research activities asso-
ciated with Getting Ready and CBC, present a number of challenges we have
encountered in the conduct of our partnership research, and propose a research
agenda to advance the empirical work on family–school partnerships.
Definitions
Parental Involvement
Family–School Partnerships
Whereas parental involvement models and activities tend to highlight the efforts
displayed by parents in ways that support what schools do to promote learning, family–
school partnership models are defined as child-focused approaches wherein fami-
lies and various professionals cooperate, coordinate, and collaborate to increase
opportunities and accomplishments related to children’s and adolescents’ social,
emotional, behavioral, and academic development (Albright & Weissberg, 2010;
1 Complexities in Field-Based Research 3
Downer & Myers, 2010; Lines, Miller, & Arthur-Stanley, 2011). Parent involve-
ment models often involve what each system (home and school) does in isolation;
family–school partnership models emphasize the bidirectional relationship between
families and schools, and intend to enrich student outcomes through cross-system
supports and continuities across both home and school settings.
Family–school partnerships are grounded in an ecological-systems theory
(Bronfenbrenner, 1979, 1992) positing that children learn and grow within distinct
and overlapping systems, and that learning and development are dependent upon
reciprocal interactions between that which occurs between the child/family and
school/schooling systems (Rimm-Kaufman & Pianta, 2000). Thus, the quality of
the intersection between home and school becomes important. Many studies have
investigated children’s experiences within the home and school systems in an iso-
lated, unidimensional (school or home) fashion, or have explored unidirectional
(school to home) models. We believe that connections and experiences across home
and school systems provide the groundwork within which children’s developmental
trajectories are strengthened (Kim et al., 2012).
Meta-analytic findings have consistently revealed positive associations between
parent–school connections and children’s academic success (cf. Fan & Chen, 2001).
The positive outcomes of parental involvement are evident across diverse samples,
various academic skills and subjects, and differing intervention approaches. For
example, the impact of parental involvement on academic achievement has been
demonstrated for minority students in urban areas (Jeynes, 2003, 2005), as well as
for boys and girls (Jeynes, 2005). Documented positive effects of parent involve-
ment on children’s homework are evident, including an increase in the time students
spend on homework and assignment completion (Epstein & Sanders, 2002).
Long-term educational achievement has also been influenced by proximal
achievement-related outcomes, including homework completion rate and the fre-
quency of homework problems (Patall, Cooper, & Robinson, 2008). Parental pro-
motion of reading and writing has been found to be significantly related to the
development of children’s vocabulary, listening comprehension and early literacy
skills (Senechal & LeFevre, 2002), and parental involvement in mathematics educa-
tion improved mathematic achievement scores among students in elementary and
secondary schools (Sheldon & Epstein, 2005).
Collaborative approaches characteristic of family–school partnership programs
have also demonstrated positive outcomes for children. For example, Barry and
Santarelli (2000) found that tantrums across both home and school settings decreased
with the use of conjoint strategies. A parent–teacher collaborative team model
yielded increased social interaction with peers, student-initiated interactions, stu-
dent engagement in classroom activities, and academic skills for three students with
disabilities (Mortier, Hunt, Desimpel, & Van Hove, 2009). A responsive model that
shapes family–school partnerships based on individual decisions (Adolescent
Transition Program and Family Check-Up; Connell, Dishion, Yasui, & Kavanagh,
2007) was shown to effect and maintain decreased problem behaviors among ado-
lescents at risk for delinquent behaviors over time in comparison to a randomized
matched control group.
4 S.M. Sheridan et al.
The early years of a child’s life are critical for establishing a healthy developmental
trajectory. Early and meaningful experiences within the home, including positive
interactions between parents and children, are predictive of important cognitive (e.g.,
language), social–emotional (e.g., relationship skills), and behavioral (e.g., compli-
ance) outcomes. Furthermore, quality relationships between parents and caregivers
are associated with children’s learning and social competence during this pivotal
developmental period (Elicker, Wen, Kwon, & Sprague, 2013). Establishing family–
school partnerships during early childhood positions parents to develop beneficial
and constructive relationships with both their children and their children’s care
providers. That is, fostering family–school partnerships during early childhood
encourages positive child outcomes and sets the stage for families to feel valued in
ways that recognize and support their role in their child’s development and learning.
Much like in early childhood, during elementary school family–school partner-
ships play a critical role in children’s development and academic success. During
these years, partnerships are focused on fostering a positive relationship between
family members and educators. Family–school partnerships appear to be most
effective when they involve collaborative problem-solving around shared goals, and
incorporate evidence-based behavioral interventions (e.g., using positive reinforce-
ment techniques, maximizing structure, and providing clear expectations for
children; Guli, 2005). Two family–school partnership models (Getting Ready, CBC)
exemplify these ingredients and are being developed and tested within our own
research lab. These models, encompassing development from early childhood to
elementary age, will be discussed in detail.
We evaluated the efficacy of the Getting Ready intervention for promoting school
readiness among disadvantaged children aged birth to 5 and their families in a
longitudinal randomized clinical trial. Participants were involved in the Getting
Ready (experimental) or business as usual (comparison) condition based on their
ECP’s random assignment. The intervention was delivered via teacher use of Getting
Ready strategies during 60-min home visits including the parent, ECP, and the child.
Teachers who were participating in the intervention were trained in the use of the
strategies and received support regarding skill use during bimonthly individual and
small group coaching sessions. They were encouraged to use the collaborative and
triadic strategies in all of their interactions with parents (e.g., parent–teacher con-
ferences, drop-off and pick up times, and other structured and unstructured
communications).
In our first published study, we examined outcomes of Getting Ready on
children’s social–emotional skills among a sample of 220 Head Start children aged
36–53 months. Fifty one percent of the sample was boys. Approximately 33 % of
students were reported by parents to be White/non-Hispanic, 25 % Hispanic/Latino,
18 % Black, 3 % Indian, and 22 % “other” (Sheridan, Knoche, Edwards, Bovaird,
& Kupzyk, 2010). We found that Head Start children who participated in Getting
Ready exhibited more initiative, demonstrated greater improvements in their attach-
ment, and experienced larger gains in social competence relative to the control
group (Sheridan et al., 2010). Further, compared to teacher reports of children in the
comparison group, teachers of children in the Getting Ready condition reported
significantly greater decreases in children’s levels of anxiety/withdrawal. A related
study found significantly greater decreases in direct observations of overactivity in
parent–child interactions for Head Start children in the Getting Ready condition
compared to those in the control group (Sheridan et al., 2014). Interestingly,
parental depression moderated the effects of the intervention, such that the greatest
increases in positive affect and verbalizations were documented for children in the
Getting Ready condition whose parents were depressed at some point during the
implementation of the intervention relative to children whose parents were not
depressed (Sheridan et al., 2014).
Significant improvements in language and early literacy skills were also found
for children participating in the Getting Ready intervention (Sheridan, Knoche,
Kupzyk, Edwards, & Marvin, 2011). Teacher reports of language use, reading and
writing improved for all children in the study; however, larger gains were reported
over time for children in the treatment group compared to those in the control group.
Language use, as reported by teachers, increased at a higher rate for experimental
group children who had been identified as having developmental concerns, who did
not speak English upon entry of preschool and who resided with two adults in the
home. Additionally, on a direct measure of expressive language, the greatest rates of
improvement were documented for children participating in Getting Ready who had
been identified as having developmental concerns and whose parents had at least a
high school diploma or GED.
1 Complexities in Field-Based Research 7
Since its inception, empirical evidence has amassed supporting the efficacy of
CBC. Using the Procedural and Coding Manual of the Task Force on Evidence-
Based Interventions in School Psychology (Kratochwill & Stoiber, 2002), Guli
(2005) determined that CBC was an effective evidence-based model to address chil-
dren’s needs across both home and school settings. Single-case research using
experimental multiple baseline designs further supports the utility of CBC to address
various behavioral (e.g., tantrums across home and school; Barry & Santarelli,
1 Complexities in Field-Based Research 9
2000), social–emotional (e.g., enhancing play, social initiation; Colton & Sheridan,
1998; Sheridan, Kratochwill, & Elliott, 1990), academic (e.g., math completion;
Weiner, Sheridan, & Jenson, 1998) and health (e.g., Type 1 Diabetes; Lasecki,
Olympia, Clark, Jenson, & Heathfield, 2008) concerns. Positive outcomes have also
been found for different mental health diagnoses (e.g., autism; Ray, Skinner, &
Watson, 1999), in medical settings (Sheridan et al., 2009), for children and families
of various ethnic groups and cultural backgrounds (Sheridan, Eagle, & Doll, 2006),
and across developmental periods (Kratochwill, Elliott, Loitz, Sladeczek, &
Carlson, 2003). In fact, in a 4-year synthesis of single-case designs using multiple
linear regression we found that a model fitting client age and symptom severity
predicted school outcomes relatively well (Sheridan, Eagle, Cowan, & Mickelson,
2001). The study highlighted the importance of addressing severe behavior prob-
lems before the middle school years, as younger students (ages 5–7) with higher
behavioral severity ratings prior to CBC experienced better outcomes than those
experiencing less severe concerns and compared to older children (11 years and
older) at all severity levels.
Recently, our research team has been testing the efficacy of CBC using large-
scale randomized controlled trials (RCTs) (Sheridan et al., 2012; Sheridan, Ryoo,
Garbacz, Kunz, & Chumney, 2013). Participants in a recently completed RCT
examining the effects of CBC for early elementary age students (Kindergarten
through third grade) included 207 children and their parents and teachers. All
students were nominated for CBC by their classroom teacher due to presenting
concerns with disruptive behaviors. Approximately 74 % of children were males,
with an average age of 7.00 (SD = 1.08). An estimated 69 % of children were
reported by parents as White/non-Hispanic, 8 % were reported African-American,
and 5 % were Hispanic or Latino. Thirty-five percent of participants lived in house-
holds with a total income less than 150 % of the poverty threshold (based on 2008
poverty thresholds and household size) and 50 % of children met criteria for free
and reduced lunch. Approximately one quarter of children (23 %) lived in a home
with only one adult (Sheridan et al., 2013). Ninety percent of parent participants
were women with an average age of 34.74 (SD = 7.79) years (Sheridan et al., 2013).
Twenty-one percent earned a high school diploma (or equivalent) or less (Sheridan
et al., 2013).
CBC procedures for collaborative home–school problem-solving were followed
as described in Sheridan and Kratochwill (2008). Trained consultants facilitated a
series of meetings between teachers and parents. Over approximately eight weeks,
a consultant met with 2–3 parents and a teacher for approximately 3–4 conjoint
consultation sessions. During each meeting, many steps were taken to ensure that
consultants maintained integrity when implementing the CBC process (i.e., objec-
tive checklists, independent data collectors, permanent products of treatment plans;
Sheridan et al., 2013).
In the first published RCT of CBC, we evaluated the effects of CBC on behav-
ioral outcomes and explored the potential role of parent–teacher relationships as a
mediating variable (Sheridan et al., 2012). Specifically, in comparison to a
“business-as-usual” control group, we found that children in the CBC group showed
10 S.M. Sheridan et al.
significant improvements in parent and teacher reported social skills and teacher-
reported adaptive behaviors. Furthermore, CBC participants reported more positive
parent–teacher relationships relative to the control group, and the gains reported in
the parent–teacher relationships partially mediated the effects of CBC on several
child outcomes.
From the same RCT, we explored the effects of CBC in the home setting. Results
revealed that CBC parents reported more communication between home and school
and greater competence in problem-solving than control group parents. Furthermore,
relative to the comparison group, children in the CBC group showed significant
decreases in defiance, noncompliance, tantrums, and arguing at home. Degree of
family risk was found to moderate parents’ competence in problem-solving and
children’s total problem behaviors, teasing, and tantrums. Specifically, families with
significant levels of disadvantage (i.e., low parental education, low income status,
fewer than two adults in the household) were likely to see greater treatment gains
than those who were less disadvantaged (Sheridan et al., 2013).
We are currently conducting an RCT evaluating the utility of CBC in Midwestern
rural communities. One hundred and thirty-three teachers of students in grades kin-
dergarten through third grade, and a total of 238 students and their parents have
participated to date. Dependent variables in the current study include student behav-
ior and academic functioning, parent/teacher beliefs and practices, and the parent–
teacher relationship. Parent and teacher rating scales, direct observations of student
behavior, and permanent products generated from parents, teachers, and consultants
are also being collected. Mediation and moderation variables are also being explored
to determine various mechanisms through which CBC exerts its influence within
rural contexts.
In sum, CBC serves as a framework for supporting family–school partnerships
during the early elementary school years. Across a number of single-case studies,
CBC has been found to produce positive effects on behavioral, social, emotional,
and academic outcomes for children. Furthermore, CBC has shown positive out-
comes across multiple settings (e.g., medical settings) and across cultural and ethnic
groups. On a much larger scale, randomized controlled trials are continuing to dem-
onstrate a number of positive effects by utilizing various collaborative strategies
between home and school and incorporating evidence-based practices that concen-
trate on children’s social–emotional, behavioral, and academic needs.
An inherent challenge for all research conducted in the context of highly controlled
experimental settings is its difficulty informing specifically what can be expected
when implementation moves to more natural, uncontrolled settings. Research on the
efficacy of family–school partnership is no exception. There are “common culprits”
associated with the conduct of intervention research in highly applied practice set-
tings (see Stormshak et al., in press). For example, lack of resources or leadership,
limited personnel training and support, competing programs and demands, and
unexpected events can each preclude the ability of schools to roll out intervention
programs effectively. Nuances unique to partnership interventions add layers of
challenges that deserve special attention. We have found that culprits associated
with recruitment, attrition and mobility, relationships, and fit present unique challenges
for implementation in naturalistic school settings.
Recruitment
precludes the inclusion of data associated with the entire triad. Thus, demands for
recruitment and retention of triads, not individuals, introduce unique challenges.
Attrition from studies occurs for a wide range of reasons. Certain families are
mobile by definition (e.g., military families) or by circumstance (e.g., immigrants,
impoverished). Nearly four percent of school age children have one or both parents
in the military (Chandra, Martin, Hawkins, & Richardson, 2010) and as a result,
experience high rates of school mobility (Kerbow, 1996). Similarly, racial and ethnic
minority children living in poverty are more likely to experience recurrent moves
(i.e., five or more moves) than their white, middle class counterparts (Murphy,
Bandy, & Moore, 2012). This type of residential mobility puts children at risk for
behavioral, socioemotional, and academic difficulties (Engel, Gallagher, & Lyle,
2010; Jelleyman & Spencer, 2008). Despite the clear need for partnership practices
that address this mobility, most research in the area assumes participants will have
long-lasting relationships with schools and teachers; yet, the natural circumstances
created by mobility result in interactions between home and school that are often
brief and difficult to sustain. Demands for research on partnership practices that
generalize across schools, teachers, and parents create certain challenges for
researchers.
A related issue is concerned with longevity of family–school partnerships.
Partnerships comprise individuals in relationship with one another at a point in time.
Research investigating the efficacy of partnership interventions are capable of deter-
mining immediate, direct effects for all participants in the triad. However, difficul-
ties arise with efforts associated with evaluating long-term effects of family–school
partnerships. Any attempts to assess the maintenance of partnerships are fraught
with the reality that new triadic partnerships are formed each academic year.
Whereas it is possible to explore the development of new relationships or partner-
ships, the long-term effects of interventions on a specific partnership dynamic is not
possible because students (and thus, their parents) move on. This special case of
mobility, albeit a natural one, creates challenges when conducting research that is
grounded in specific relationships.
learned that researchers and field staff often have similar goals, but different means
for achieving them. Ultimately, they generally share a desire to identify methods
for helping students achieve the best of their ability, and to begin closing the
opportunity and achievement gap that exists for many children. However, different
vantage points encourage them to approach this goal using different methods.
Whereas school staff may feel pressured and look for an efficient solution to press-
ing needs, researchers generally prefer systematic, precise, rigorous implementa-
tion, observation, analysis, and interpretation of routinized practices while
controlling what is possible. Thus, there is often a disconnect between the needs of
researchers for imposing rigorous experimental control, and educational practitio-
ners needing timely solutions and ready information on effective practices for their
school context.
Fit
The issue of fit between the partnership program being evaluated in research and the
field site within which it is being implemented is an important one. Research
addressing interactions between partnership programs and the systems within which
they are implemented aims to identify specific contextual variables that may influ-
ence uptake and efficacy of partnership interventions. Efforts at determining what
works in the field need to also consider “for whom” and “in what context” effects
can be expected. Thus, in addition to asking what works, there is a need to consider
systemic variables in order to predict (a) whether what fits within natural and applied
implementation sites works, and (b) whether what works actually fits.
An issue associated with the identification of interactions between context and
intervention pertains to identifying potential levels or dosage of implementation
that may be sufficient (just enough), extensive (too much), or indicated (just right).
Policy and practice demands from inside and outside of the school system often
require selection of programs to be based not only on what is perceived as most
effective for meeting the goals of its constituents, but also what is most efficient or
least costly. Even when research points to specific partnership practices that may
result in positive outcomes for meeting identified needs, schools will not always be
positioned to implement a full intervention program. Adoption of a full partnership
model with families may be overwhelming to field sites in terms of human, finan-
cial, and material resources. Given that a system’s capacity to support an interven-
tion may determine its implementation, it seems important to know the amount of
an intervention that will produce desired effects so as not to produce undue pressure
or overtax the system’s capacity and resources for little gain or added value.
Research that uncovers the specific ingredients of partnership interventions and
how they predict desired outcomes will help to discern levels of implementation
that produce absolute (indicated), threshold (just enough), and saturation (too
much) effects.
16 S.M. Sheridan et al.
Research Agenda
determine the strategies used for specific areas of concern (e.g., disruptive behaviors,
academic skill deficits) and students of diverse backgrounds (e.g., race, language
spoken). Other statistical analyses can then be used to test the elements in relation
to outcomes (e.g., West, Walia, Hyder, Shahab, & Michie, 2010). Findings can
inform the understanding of the operative components of partnerships and each
component’s potency to help determine whether each ingredient is an essential,
desirable, or sufficient condition to produce treatment effects. An understanding of
each ingredient’s weight can clarify the components required to maximize the
potential for certain outcomes. More nuanced analyses of desirable partnership fea-
tures (e.g., quality, frequency) will uncover elements that can be adapted to be
responsive to particular relationship dynamics and local contexts.
Translational Methods
Fidelity Analysis
Cutting across each line of research is the need to clearly specify the family–school
partnership intervention being implemented. Drawing valid conclusions regarding
the effects of a partnership program relies on an understanding of what comprises
the intervention and how those components were delivered and received; that is,
how fidelity operates within family–school partnership interventions. As a result,
fidelity assessment depends on both an awareness of the critical elements of fam-
ily–school partnerships as well as the contextual features that influence their
implementation.
A necessary step in this line of research is to determine psychometrically sound
measures to assess the various dimensions of fidelity (e.g., adherence, quality, par-
ticipant responsiveness, dosage, and program differentiation; Dane & Schneider,
1998). Although critical elements of partnership interventions (and their operational
definitions) have not been identified, it is likely that relational features such as
shared responsibility, mutual decision making, and bidirectional communication are
components that contribute to a partnership intervention. To date, no generally
accepted tool or approach to assessing partnership intervention fidelity has been
identified, and as a result, basic reliability and validity evidence for measures of
partnership intervention fidelity have not been evaluated. Only when measures are
developed and subject to rigorous psychometric analyses will researchers be able to
fully explore fidelity within partnership programs.
Once measures of fidelity are determined, attention to the various dimensions
will allow researchers to carefully examine many facets of family–school partner-
ship interventions (Sheridan et al., 2014). It is necessary to begin empirically inves-
tigating fidelity (and each dimension) as an independent variable or a moderator of
treatment effects. Testing relationships between fidelity and relevant variables can
help determine the influence of variations of fidelity and the mechanisms through
which partnership interventions operate.
Another particularly valuable line of inquiry could examine threshold and satura-
tion levels of partnership programs (Sheridan et al., 2014). Assessment of threshold
levels can clarify the intensity required for partnership interventions to produce
optimal effects, and assessment of saturation levels can establish the point at
which effects may have reached and surpassed their peak. Information generated
from this line of research can help to determine how much deviation is permissible
20 S.M. Sheridan et al.
(i.e., allowable degree of drift) within partnership interventions and how much
support is needed by the system to implement the practices. Indeed, research of this
type and specificity would be beneficial for virtually all psychosocial and educa-
tional interventions.
Conclusion
Our research team has spent the last decade conducting family–school partnership
intervention research. Although our findings have added to the historic body of
evidence demonstrating the importance of family–school partnerships, careful
reflection and years of experience have uncovered the complexities associated with
the execution (e.g., fidelity of partnership practices), interpretation (e.g., defini-
tional variations), and translation (e.g., fit of partnership intervention) of partnership
research. We believe a viable research agenda is one that seeks to address these
unique challenges by empirically determining the active ingredients of family–
school partnership programs, using methods that support the translation of partner-
ship research into practice, and measuring and analyzing the fidelity with which
these interventions are put into practice to provide a more nuanced understanding of
how family–school partnership interventions operate.
Acknowledgments This study was supported by federal grants awarded to the first author by the
U.S. Department of Education Institute of Education Sciences (Grants #R305C090022,
R324A100115, R324A120153, R305A120144). The opinions expressed herein are those of the
authors and are not considered reflective of the funding agency. For correspondence regarding this
article, please contact Dr. Susan Sheridan.
References
Albright, M. I., & Weissberg, R. P. (2010). Family-school partnerships to promote social and emo-
tional learning. In S. L. Christenson & A. L. Reschly (Eds.), Handbook of school–family part-
nerships (pp. 246–265). New York, NY: Routledge.
Barry, L. M., & Santarelli, G. E. (2000). Making it work at school and home: A need based
collaborative, across settings, behavioral intervention. The California School Psychologist, 5,
43–51.
Bronfenbrenner, U. (1979). The ecology of human development: Experimental by nature and
design. Cambridge, MA: Harvard University Press.
Bronfenbrenner, U. (1992). Ecological systems theory. In R. Vasta (Ed.), Six theories of child
development: Revised formulations and current issues (pp. 187–248). Philadelphia, PA: Jessica
Kingsley.
Brown, J. R., Knoche, L. L., Edwards, C. P., & Sheridan, S. M. (2009). Professional development
to support parent engagement: A case study of early childhood practitioners. Early Education
and Development, 20, 482–506.
Chandra, A., Martin, L. T., Hawkins, S. A., & Richardson, A. (2010). The impact of parental
deployment on child social and emotional functioning: Perspectives of school staff. Journal of
Adolescent Health, 46, 218–223. doi:10.1016/j.jadohealth.2009.10.009.
1 Complexities in Field-Based Research 21
Chorpita, B. F., Becker, K. D., & Daleiden, E. L. (2007). Understanding the common elements of
evidence-based practice: Misconceptions and clinical examples. Journal of the American
Academy of Child & Adolescent Psychiatry, 46, 647–652. doi:10.1097/chi.0b013c318033ff71.
Chorpita, B. F., & Daleiden, E. L. (2009). Mapping evidence-based treatments for children and
adolescents: Application of the distillation and matching model to 615 treatments from 322
randomized trials. Journal of Consulting and Clinical Psychology, 77, 566–579. doi:10.1037/
a0014565.
Chorpita, B. F., Daleiden, E. L., & Weisz, J. R. (2005). Identifying and selecting the common ele-
ments of evidence based interventions: A distillation and matching model. Mental Health
Services Research, 7, 5–20. doi:10.1007/s11020-005-1962-6.
Colton, D., & Sheridan, S. M. (1998). Conjoint behavioral consultation and social skills training:
Enhancing the play behavior of boys with attention deficit-hyperactivity disorder. Journal of
Educational and Psychological Consultation, 9, 3–28.
Connell, A. M., Dishion, T. J., Yasui, M., & Kavanagh, K. (2007). An adaptive approach to family
intervention: Linking engagement in family-centered intervention to reductions in adolescent
problem behavior. Journal of Consulting and Clinical Psychology, 75, 568–579.
Damschroder, L. J., & Hagedorn, H. J. (2011). A guiding framework and approach for implementa-
tion research in substance use disorders treatment. Psychology of Addictive Behaviors, 25,
194–205. doi:10.1037/a0022284.
Dane, A. V., & Schneider, B. H. (1998). Program integrity in primary and early secondary preven-
tion: Are implementation effects out of control? Clinical Psychology Review, 18, 23–45.
doi:10.1016/S0272-7358(97)00043-3.
Downer, J. T., & Myers, S. S. (2010). Application of a developmental/ecological model to family-
school partnerships. In S. L. Christenson & A. L. Reschly (Eds.), Handbook of school–family
partnerships (pp. 3–29). New York, NY: Routledge.
Elicker, J., Wen, X., Kwon, K. A., & Sprague, J. B. (2013). Early Head Start relationships:
Association with program outcomes. Early Education & Development, 24, 491–516.
Engel, R. C., Gallagher, L. B., & Lyle, D. S. (2010). Military deployment and children’s academic
achievement: Evidence from Department of Defense Education Activity schools. Economics of
Education Review, 29, 73–82. doi:10.1016/jeconedurev.2008.12.003.
Epstein, J. L., & Sanders, M. G. (2002). Family, school, and community partnerships. In M. H.
Bornstein (Ed.), Handbook of parenting (Vol. 5, pp. 407–437). Mahwah, NJ: Lawrence
Erlbaum.
Fan, X., & Chen, M. (2001). Parental involvement and students’ academic achievement: A meta-
analysis. Educational Psychology Review, 13, 1–22.
Fantuzzo, J., Tighe, E., & Childs, S. (2000). Family involvement questionnaire: A multivariate
assessment of family participation in early childhood education. Journal of Educational
Psychology, 92, 367–376.
Forman, S. G., Shapiro, E. S., Codding, R. S., Gonzales, J. E., Reddy, L. A., Rosenfield, S. A., …
Stoiber, K. C. (2013). Implementation science and school psychology. School Psychology
Quarterly, 28, 77-100. doi:10.1037/spq0000019
Grolnick, W. S., & Slowiaczek, M. L. (1994). Parents’ involvement in children’s schooling: A
multidimensional conceptualization and motivational model. Child Development, 65,
237–252.
Guli, L. A. (2005). Evidence-based parent consultation with school-related outcomes. School
Psychology Quarterly, 20, 455–472.
Jelleyman, T., & Spencer, N. (2008). Residential mobility in childhood and health outcomes: A
systematic review. Journal of Epidemiology and Community Health, 62, 584–592. doi:10.1136/
jech.2007.060103.
Jeynes, W. H. (2003). A meta-analysis: The effects of parental involvement on minority children’s
academic achievement. Education and Urban Society, 35, 202–218.
Jeynes, W. H. (2005). A meta-analysis of the relation of parental involvement to urban elementary
school student academic achievement. Urban Education, 40, 237–269.
22 S.M. Sheridan et al.
Kerbow, D. (1996). Patterns of urban student mobility and local school reform. Journal of
Education for Students Placed at Risk, 1, 147–169. Retrieved from http://www.csos.jhu.edu/
crespar/techReports/Report5.pdf.
Kim, E. M., Coutts, M. J., Holmes, S. R., Sheridan, S. M., Ransom, K. A., Sjuts, T. M., & Rispoli,
K. M. (2012). Parent involvement and family-school partnerships: Examining the content, pro-
cesses, and outcomes of structural versus relationship-based approaches (CYFS Working
Paper No. 2012-6). Retrieved from cyfs.unl.edu
Kim, E. M., & Sheridan, S. M. (2014). Foundational aspects of family-school connections:
Definitions, conceptual frameworks, and research needs. In S. M. Sheridan & E. M. Kim
(Eds.), Research on family-school partnerships: An interdisciplinary examination of state of
the science and critical needs, 1, 1–14. New York, NY: Springer.
Knoche, L. L., Edwards, C. P., Sheridan, S. M., Kupzyk, K. A., Marvin, C. A., Cline, K. D., &
Clarke, B. L. (2012). Getting Ready: Results of a randomized trial of a relationship-focused
intervention on the parent-infant relationship in rural early head start. Infant Mental Health
Journal, 33, 439–458.
Knoche, L. L., Sheridan, S. M., Edwards, C. P., & Osborn, A. Q. (2010). Implementation of a
relationship-based school readiness intervention: A multidimensional approach to fidelity mea-
surement for early childhood. Early Childhood Research Quarterly, 25, 299–313.
Kratochwill, T. R., Elliott, S. N., Loitz, P. A., Sladeczek, I., & Carlson, J. S. (2003). Conjoint con-
sultation using self-administered manual and videotape parent–teacher training: Effects on
children’s behavioral difficulties. School Psychology Quarterly, 18, 269–302.
Kratochwill, T. R., Hoagwood, K. E., Kazak, A. E., Weisz, J. R., Hood, K., Vargas, L. A., & Banez,
G. A. (2012). Practice-based evidence for children and adolescents: Advancing the research
agenda in schools. School Psychology Review, 41, 215–235.
Kratochwill, T. R., & Stoiber, K. C. (2002). Evidence-based interventions in school psychology:
Conceptual foundations of the Procedural and Coding Manual of Division 16 and the Society
for the Study of School Psychology Task Force. School Psychology Quarterly, 17, 341–389.
Lasecki, K., Olympia, D., Clark, E., Jenson, W., & Heathfield, L. T. (2008). Using behavioral
interventions to assist children with type 1 diabetes manage blood glucose levels. School
Psychology Quarterly, 23, 389–406.
Lines, C., Miller, G. E., & Arthur-Stanley, A. (2011). The power of family-school partnering
(FSP): A practical guide for school mental health professionals and educators. New York, NY:
Routledge.
Mortier, K., Hunt, P., Desimpel, L., & Van Hove, G. (2009). With parents at the table: Creating
supports for children with disabilities in general education classrooms. European Journal of
Special Needs Education, 24, 337–354.
Murphy, D., Bandy, T., & Moore, K. A. (2012). Frequent residential mobility and young children’s
well-being. Child Trends Research Brief. Retrieved from www.childtrends.org
Power, T. J. (2003). Promoting children’s mental health: Reform through interdisciplinary and
community partnerships. School Psychology Review, 32, 3–16.
Ray, K. P., Skinner, C. H., & Watson, T. S. (1999). Transferring stimulus control via momentum to
increase compliance in a student with autism: A demonstration of collaborative consultation.
School Psychology Review, 28, 622–628.
Rimm-Kaufman, S. E., & Pianta, R. C. (2000). An ecological perspective on the transition to kin-
dergarten: A theoretical framework to guide empirical research. Journal of Applied
Developmental Psychology, 21, 491–511.
Senechal, M., & LeFevre, J. (2002). Parental involvement in the development of children’s reading
skill: A five-year longitudinal study. Child Development, 73, 445–460.
Sheldon, S. B., & Epstein, J. L. (2005). Involvement counts: Family and community partnerships
and mathematics achievement. Journal of Education Research, 98, 196–206.
Sheridan, S. M., Bovaird, J. A., Glover, T. A., Garbacz, S. A., Witte, A., & Kwon, K. (2012). A
randomized trial examining the effects of conjoint behavioral consultation and the mediating
role of the parent-teacher relationship. School Psychology Review, 41, 23–46.
1 Complexities in Field-Based Research 23
Sheridan, S. M., Eagle, J. W., Cowan, R. J., & Mickelson, W. (2001). The effects of conjoint
behavioral consultation: Results of a four-year investigation. Journal of School Psychology, 39,
361–385.
Sheridan, S. M., Eagle, J. W., & Doll, B. (2006). An examination of the efficacy of conjoint behav-
ioral consultation with diverse clients. School Psychology Quarterly, 21, 396–417.
Sheridan, S. M., Knoche, L. L., Edwards, C. P., Bovaird, J., & Kupzyk, K. A. (2010). Parent
engagement and school readiness: Effects of the Getting Ready intervention on preschool chil-
dren’s social–emotional competencies and behavioral concerns. Early Education and
Development, 21, 125–156.
Sheridan, S. M., Knoche, L. L., Edwards, C. P., Kupzyk, K. A., Clarke, B. L., & Kim, E. M. (2014).
Efficacy of the Getting Ready intervention and the role of parental depression. Early Education
and Development., 25, 746–769.
Sheridan, S. M., Knoche, L. L., Kupzyk, K. A., Edwards, C., & Marvin, C. A. (2011). A random-
ized trial examining the effects of parent engagement on early language and literacy: The
Getting Ready intervention. Journal of School Psychology, 49, 361–383.
Sheridan, S. M., & Kratochwill, T. R. (1992). Behavioral parent–teacher consultation: Conceptual
and research considerations. Journal of School Psychology, 30, 117–139.
Sheridan, S. M., & Kratochwill, T. R. (2008). Conjoint behavioral consultation: Promoting fam-
ily–school connections and interventions. New York, NY: Springer.
Sheridan, S. M., Kratochwill, T. R., & Bergan, J. R. (1996). Conjoint behavioral consultation: A
procedural manual. New York, NY: Plenum.
Sheridan, S. M., Kratochwill, T. R., & Elliott, S. N. (1990). Behavioral consultation with parents
and teachers: Delivering treatment for socially withdrawn children at home and school. School
Psychology Review, 19, 33–52.
Sheridan, S. M., Marvin, C. A., Knoche, L. L., & Edwards, C. P. (2008). Getting ready: Promoting
school readiness through a relationship-based partnership model. Early Childhood Services, 3,
149–172.
Sheridan, S. M., Rispoli, K., & Holmes, S. R. (2014). Treatment integrity in conjoint behavioral
consultation: Conceptualizing active ingredients and potential pathways of influence. In
L. Sanetti & T. Kratochwill (Eds.), Treatment integrity: Conceptual, methodological, and
applied considerations for practitioners (pp. 255–278). Washington, DC: American
Psychological Association.
Sheridan, S. M., Ryoo, J. H., Garbacz, S. A., Kunz, G. M., & Chumney, F. L. (2013). The efficacy
of conjoint behavioral consultation on parents and children in the home setting: Results of a
randomized controlled trial. Journal of School Psychology, 51, 717–733.
Sheridan, S. M., Warnes, E. D., Woods, K. E., Blevins, C. A., Magee, K. L., & Ellis, C. (2009). An
exploratory evaluation of conjoint behavioral consultation to promote collaboration among
family, school, and pediatric systems: A role for pediatric school psychologists. Journal of
Educational and Psychological Consultation, 19, 106–129.
Stormshak, E., Brown, K. L., Moore, K. J., Dishion, T., Seeley, J., & Smolkowski, K. (in press).
Going to scale with family-centered, school-based interventions: Challenges and future direc-
tions. In S. M. Sheridan & E. M. Kim (Eds.), Research on family-school partnerships: An
interdisciplinary examination of state of the science and critical needs. Vol. IV: Translating
family-school partnerships research into practice. New York, NY: Springer.
Weiner, R., Sheridan, S. M., & Jenson, W. R. (1998). Effects of conjoint behavioral consultation
and a structured homework program on math completion and accuracy in junior high students.
School Psychology Quarterly, 13, 281–309.
West, R., Walia, A., Hyder, N., Shahab, L., & Michie, S. (2010). Behavior change techniques used
by English stop smoking services and their associations with short-term quit outcomes.
Nicotine & Tobacco Research, 12, 742–747. doi:10.1093/ntr/ntq074.
Chapter 2
Going to Scale with Family-Centered,
School-Based Interventions: Challenges
and Future Directions
Students who display problem behaviors at school are at risk for a variety of
difficulties, including poor academic achievement, poor school attendance,
depression, and substance use (Barry, Lyman, & Grofer Klinger, 2002; Patterson,
Reid, & Dishion, 1992), all of which can be challenging for teachers and school
administrators to manage (Dishion & Stormshak, 2007; Walker, Colvin, & Ramsey,
1995). Many schools also lack the infrastructure necessary to systematically and
effectively support children and adolescents with academic, behavioral, or mental
health concerns (Eccles & Harold, 1993; Ringeisen, Henderson, & Hoagwood,
2003). However, schools are an ideal location to implement evidence-based prevention
and intervention programs to address problem behaviors because youths spend a
considerable amount of time there (Dishion, 2011). The World Health Organization
(WHO, 2008) and Centers for Disease Control (CDC, 2013) promote school set-
tings as particularly important for actions that target and improve outcomes for
child and adolescent health. Moreover, using schools as service delivery settings
may increase opportunities to provide health services to underserved populations,
such as rural populations, low-income families, and ethnically diverse youths. As
such, local, state, and federal policies have increasingly called for the use of
evidence-based practices in school settings.
Little theory or research exists regarding how to implement behavioral and mental
health interventions, such as family–school partnership programs, with fidelity
(Domitrovich & Greenberg, 2000; Elliott & Mihalic, 2004), yet high-quality imple-
mentation is directly linked to strong outcomes and improved effect sizes across inter-
vention models (Durlak & DuPre, 2008). Family-centered treatment models that
emphasize parent training and support for families show the largest effects over time
in nearly every review of interventions designed to reduce problem behavior and sub-
stance use (Kazdin, 2010; Prinz & Dumas, 2004). Despite this overwhelming evi-
dence, a high proportion of children and adolescents never receive treatment for these
problems, and a very small percentage of parents participate in parenting or family
interventions to address behavior problems (Prinz & Sanders, 2007; Zubrick et al.,
1995). Limited access to mental health treatment for children has fueled an increase in
the number of school-based mental health programs in the United States. Nevertheless,
there are many barriers to implementing the programs effectively (Weist, 2005).
The majority of interventions focus on the individual child or on the school context
(e.g., positive behavior support; Horner, Sugai, Todd, & Lewis-Palmer, 2005); few
are brief and target known risk factors for substance use, such as family management.
They typically consist of a response-to-intervention (RtI) framework that embeds ser-
vices in a model of universal, selected, and indicated interventions delivered in the
school, but they offer little coordination between school and home. This is a serious
shortcoming in that the nature of interactions between parents and their child’s school
becomes more formalized and less frequent in middle school (Rimm-Kaufman &
Pianta, 2000) and leads to less engagement by parents in their child’s overall adjust-
ment, when parenting and family management are critical to school success and
healthy adaptation. Data strongly suggest that motivating parents to engage in family
management will effect long-term change (Dishion & Kavanagh, 2003; Forgatch,
Bullock, & Patterson, 2004; Kazdin, 2002; Stormshak, Fosco, & Dishion, 2010), and
research supports the efficacy of interventions for high-risk students in the public
school environment that target parenting practices (e.g., Atkins et al., 2008).
Multiple barriers, such as time, money, and competing priorities, limit the ability
of schools to implement interventions that involve families (Forman, Olin,
Hoagwood, Crowe, & Saka, 2009), making most interventions that target parenting
practices unrealistic for schools (Christenson, 2003), despite their proven efficacy.
Given the dire economic situation currently facing many school systems, it is imper-
ative to find a cost-effective means of improving student success rates that is efficient
2 Challenges in Going to Scale 27
and effective, realistic, does not require extensive school staff time to implement
with fidelity, and integrates families and family-centered care into school systems.
Literature pertaining to the diffusion, implementation, and sustainability of
school-based interventions is sparse and leaves schools with little strategic support
regarding the use of evidence-based programs (Feldstein & Glasgow, 2008).
Additional research has found that evidence-based programs implemented outside
of controlled trials are generally not executed to proficient levels of quality
(Dusenbury, Brannigan, Hansen, Walsh, & Falco, 2005; Gottfredson & Gottfredson,
2002). This is unfortunate because program fidelity is strongly linked to positive
intervention outcomes (Durlak & DuPre, 2008). A more systematic understanding
of how to effectively and accurately implement evidence-based family–school part-
nership interventions in school settings is needed to ensure successful student out-
comes (Greenberg, Domitrovich, Graczyk, & Zins, 2001). Failure to fulfill this need
may lead to detrimental effects, such as the inability of schools to develop and sus-
tain systems of intervention that support struggling students (Dishion, 2011).
Translation of research to practice in community settings involves several phases,
including a preadoption phase, during which key stakeholders and markets are iden-
tified; the adoption phase, during which organizations get ready to implement the
program; the implementation phase, during which training and fidelity evaluation
occur; and a sustainability phase, during which structures and policies are identified
to enable continuation of the intervention (Spoth et al., 2013). Many contextual fac-
tors in schools influence their ability to implement family–school partnership
practices and sustain them, including teacher training, administrative support, finan-
cial resources, and school morale and organization. Few of these factors are taken
into consideration when these interventions are developed or disseminated, however
(Domitrovich et al., 2008). Protecting program fidelity is a primary goal when
evidence-based programs are embedded in existing school frameworks (Spoth,
Kavanagh, & Dishion, 2002). Even though few guidelines exist that demonstrate
how to integrate programs effectively and realistically (Dishion, 2011) and that
identify contextual and program structures that can make or break implementation
quality (Payne & Eckert, 2010), researchers and practitioners must understand the
conditions that both facilitate and impede high-quality implementation in schools.
This is a crucial next step in implementation science because schools connect daily
with large numbers of children and thus are valuable venues for dissemination of
prevention and intervention programs. In fact, schools are the largest provider of
child behavioral health services and the only community setting where many chil-
dren receive any behavioral health interventions at all (Bums et al., 1995; Hoagwood,
Bums, Kiser, Ringeisen, & Schoenwald, 2001).
Poor parenting practices and family relationships have been linked to the development
and maintenance of youths’ problem behaviors (Connell & Dishion, 2008; Spoth
et al., 2002; Stormshak, E. A., Bierman, K. L., McMahon, R. J., Lengua, L., &
28 E.A. Stormshak et al.
Conduct Problems Prevention Research Group, 2000). On the other hand, healthy
parenting practices and relationships have been associated with positive youth
outcomes, even in the presence of factors such as poverty and stress (Galambos,
Barker, & Almeida, 2003; Ryan, Martin, & Brooks-Gunn, 2006). It makes sense
that interventions targeting the development of positive parenting systems are effec-
tive for reducing youth problem behaviors (Dishion, Nelson, & Kavanagh, 2003;
Dishion & Stormshak, 2007; Forgatch, DeGarmo, & Beldavs, 2005). A particularly
salient time for intervening with parents may be during their child’s transition to
middle school, in that problem behaviors often amplify during adolescence (Dishion
& Patterson, 2006; Patterson, Capaldi, & Bank, 1991). For example, decreased
parental monitoring and parent–teacher communication and increased exposure to
peers make the middle school years a risk period for the development of adolescent
substance use, aggression, and violence (Dishion, Patterson, Stoolmiller, & Skinner,
1991; Eccles, Lord, & Roeser, 1995).
Even though parental involvement in school tends to decline when children enter
middle school, research has shown that parent involvement in education is associ-
ated with positive child outcomes, including higher grade point averages (Gutman
& Midgley, 2000), better self-regulation and social skills (Brody, Flor, & Gibson,
1999; McWayne, Hampton, Fantuzzo, Cohen, & Sekino, 2004), lower dropout rates
(Rumberger, 1995), fewer grade retentions and special education placements
(Miedel & Reynolds, 1999), and improvements in language and reading skills
(Sheridan, Knoche, Kupzyk, Pope Edwards, & Marvin, 2011). When parents are
involved in their child’s education, students also more readily adjust to the demands
of the classroom and show improved academic performance (Epstein, 1991;
Henderson & Berla, 1994; Henderson & Mapp, 2002; Reynolds, 1992). Despite the
advantages of involving parents in school settings in terms of primary school out-
comes, such as achievement, few schools develop or maintain organized systems for
positively intervening with them (Stormshak & Dishion, 2002).
Kavanagh, 2003; Dishion & Stormshak, 2007; Stormshak et al., 2011; Stormshak,
Dishion, Light, & Yasui, 2005). These trials with ethnically and socioeconomically
diverse young children and middle school-age youths have demonstrated interven-
tion effects on self-regulation, grade point average, attendance, school engagement,
and growth of teacher-rated child problem behavior over time (Fosco et al., 2012;
Stormshak et al., 2005, 2010), as well as a variety of nonacademic outcomes, such
as rates of depression, substance use, high-risk sexual behavior, and early-adult
obesity (Connell, Dishion, & Deater-Deckard, 2006; Stormshak et al., 2010; Van
Ryzin & Nowicka, 2013).
The PFS model is intended to be delivered by school personnel with relatively
little support from external consultants. Adaptations to the model for integration
into middle schools have included tiered intervention intensity (Myers & Nastasi,
1999), strategies to enhance motivation (Miller & Rollnick, 2002), and a tailored
intervention design (Collins, Murphy, & Bierman, 2004). As depicted in Fig. 2.1,
the core components of PFS have been matched and dovetailed to components of
school-wide positive behavior support systems, such as the Positive Behavior
Interventions and Supports (PBIS; Sugai et al., 2000) model, to build bridges
between school and home. The dovetailing of PFS with programs such as PBIS was
intentional with respect to a public health implementation perspective (Biglan,
1995; Biglan, Sprague, & Moore, 2006; Shaw, 1986). This perspective suggests that
by using an effective intervention model that makes pragmatic use of naturally
occurring ecological settings and well-established service delivery structures (e.g.,
PBIS), the rate of engagement will increase while some of the implementation
response cost to the setting will be reduced. In the scaling-up process, the ability of
PFS to improve family–school partnerships and student academic and social out-
comes was tested through dissemination and promotion of family support services
by existing school personnel. The program has been implemented in a range of
schools and economic conditions in Oregon; Table 2.1 shows the distribution of the
Fig. 2.1 Concatenation of school and family Positive Behavior Interventions and Supports (PBIS)
30
Lack of Resources
Lack of resources, money, and staffing is probably the most common reason that
researchers fail to successfully disseminate interventions, curricula, and other
empirically based models to schools. It is a key underlying factor that predicts poor
uptake and implementation, and it is related to staff turnover, lack of training, lack
of principal support, and other critical variables that predict implementation success
(Gingiss, Roberts-Gray, & Boerm, 2006; Payne & Eckert, 2010). Many schools are
serving growing numbers of students and managing increasing rates of mental
32 E.A. Stormshak et al.
health and behavioral problems with limited financial resources (Stormshak et al.,
2011). A diminished staff-to-student ratio can leave school personnel at all levels
feeling overworked and undersupported, with little energy or time left to implement
family–school partnership interventions. Assets are being depleted in an effort to
comply with existing public policy, such as achievement testing and complex
teacher and administrative evaluations. Amid financial cutbacks and reductions in
funding, many schools struggle to find the administrative focus, support services,
class time, or physical space necessary to support additional programs. Family–
school interventions may be seen as a luxury rather than a component of student
success and well-being (Stormshak et al., 2005). Principals who must cut key staff
members often find it politically challenging to prioritize interventions that are con-
sidered superfluous by influential stakeholders (Fosco et al., 2014). The overall lack
of resources and the increasing numbers of problematic students present substantial
barriers to successful implementation and sustainability of school-based prevention
and intervention programs.
Finding a balance between the need to sustain all pertinent intervention compo-
nents and the need to be brief and cost-effective is a potential difficulty in family–
school partnership research. Infusion of a multilevel, family-centered approach, that
is, a combination of universal, selected, and indicated levels, contributes to the effi-
cient management of resources, in that only those students who are most in need of
intensive intervention receive costly services (Stormshak et al., 2011). A tiered
design also dovetails more easily with other multilevel education programs, such as
PBIS and/or RtI. Programs can be executed gradually to further reduce burden and
support school staff who may be overwhelmed by the implementation of a new
intervention. This approach can also prove to be more economical. Similarly, when
it is not possible for one school staff member to devote the time needed to put the
intervention into action, it may become necessary to shift responsibilities to a range
of staff members. When this occurs, it is crucial that intervention components be
seamlessly integrated into the existing activities of school personnel, so as not to
overload already-pressured individuals (Gottfredson & Gottfredson, 2002). Finally,
user-friendly and streamlined implementation materials, such as manuals, bro-
chures, and videos, enable school staff to easily accommodate an intervention with-
out expending excessive amounts of time or energy. Materials should offer detailed
guidelines and explicit scripts yet remain flexible enough to match the unique needs
of individual schools and staff members (Turner & Sanders, 2006). In sum, it is vital
that intervention designers and implementers seriously consider each school’s avail-
able resources because attempts to support expensive and time-consuming pro-
grams often result in poor uptake, execution, and sustainability. Research that
focuses on adapting programs to fit into existing curricula and services in schools is
critical to understanding how to improve uptake in schools with few resources.
In our PFS project, schools in the intervention condition have weathered constant
budgetary changes during the course of implementation. Forty-eight percent of the
schools experienced a loss in overall operating expenditures per student between the
2007–2008 and 2011–2012 school years. In particular, one relatively small rural
school lost more than $1,500 of operating expenditures per student during this time
2 Challenges in Going to Scale 33
period. Expenditure per student for counselor, nurse, and support staff services was
reduced in 57 % of PFS intervention schools. A common result has been low teacher
and administrator morale brought on by multiple years of job loss, employment
insecurity, and turnover. Constant budgetary shortages have also compromised con-
tinuity in terms of training, intervention implementation, and staff expectations.
While the majority of school staff are involved in education out of genuine concern
for children and excel at being student centered, few understand or take into account
family contributions to problem behavior (Shirk & Jungbluth, 2008; Stormshak,
Connell, & Dishion, 2009). That said, schools are not likely to gather information
from parents regarding conditions at home or involve parents in school-based inter-
ventions. Furthermore, most school staff are trained according to individual models
of development (Stormshak et al., 2005), meaning few individuals working in
schools have the knowledge or skills necessary to consistently engage parents in a
manner that effectively and positively supports children’s academic and behavioral
success. Programs that require schools to enact structural changes to accommodate
the proactive involvement of parents in school-based student interventions often
necessitate a substantial shift in traditional paradigms (Fosco et al., 2014). Such
large-scale changes may impede the successful implementation and scale-up of
family–school partnership programs. If these changes can be embedded or dove-
tailed with other successful structural changes, such as PBIS, these impediments
may be diminished.
Positive family–school partnership is a key component in the behavioral, mental
health, and academic success of students, yet many school staff do not receive train-
ing in how to effectively engage parents (Stormshak et al., 2005, 2011). Not only
must scientists who are developing and disseminating family–school interventions
be cognizant of the need to empower through efficient multilevel program design,
but school administrators and teachers must learn how to proactively and positively
interact with parents. To begin, implementers should be ready to provide direction,
coaching, scaffolding, corrective feedback, and encouragement to school personnel
about their interactions with parents (Gottfredson & Gottfredson, 2002). This sup-
port opens an important channel of communication between scientists and schools
regarding best practices with respect to using family–school partnerships to enhance
student success and well-being. It may also be necessary to contextualize parenting
skills in terms of school-relevant tasks, such as homework routines and positive
behavior support, to overcome resistance and help staff recognize their own exper-
tise in providing parental support (Fosco et al., 2014). Similarly, materials must be
available that are concrete, behavioral, and positive in nature (Fosco et al., 2014).
Overall, all these exigencies must be addressed to overcome any resistance to alter-
ing existing school paradigms whose focus is on individual student development to
the exclusion of family participation.
34 E.A. Stormshak et al.
School Leadership
Support and leadership from school principals is a key element in the successful
implementation of evidence-based family–school partnership programs. Without it,
fidelity is not maintained long enough to fully integrate the program into school
policy and routines (Handler et al., 2007; McDougal, Clonan, & Martens, 2000).
When factors that predict successful uptake of interventions are studied, principals’
support often predicts successful implementation and maintenance of models in
schools (Payne, 2008; Payne & Eckert, 2010). Closely involving principals in the
training, consultation, implementation, and sustainability of these programs can be
difficult in the face of time constraints, limited resources, varying interest levels,
and individual differences in leadership ability. Yet, principals are crucial to estab-
lishing family–school partnerships as an overarching school norm and holding
school staff accountable for maintaining positive collaborations with families.
When principals do not consistently advocate for family collaboration, uptake and
maintenance of the intervention can be seriously constrained. Persistent administra-
tive turnover presents additional challenges, especially during difficult economic
times. In the PFS project, high levels of turnover occurred during the 3 years of the
study, with 45 % of schools turning over at least one principal and 20 % of schools
hiring a new principal every year of the study. In addition, vice principal turnover
occurred at 40 % of schools at least once during the project period. To accomplish
buy-in and support for the existing model required quick adjustment to working
with new school leadership in the middle of the project.
Gaining the support and buy-in of school principals can be a challenging task.
Tremendous demands have been placed on them in this age of achievement account-
ability, instructional leadership, and federal and state requirements for new, time-
consuming teacher evaluation methods. As a result, they may not appreciate
additional expectations to foster a family-friendly school culture. Implementation
2 Challenges in Going to Scale 35
School Climate
The overall school climate can significantly affect the staff’s ability to successfully
implement evidence-based family–school partnership programs. School climate,
defined in the literature in multiple ways, often refers to supportive administration
and endorsement of program implementation (Beets et al., 2008). Schools that fos-
ter a sense of respect, collaboration, support, and active problem solving at all staff-
ing levels to effectively sustain implementation requirements may be the most
successful at long-term uptake of programs (Greenberg, Domitrovich, & Bumbarger,
2001). Constructive and encouraging interpersonal relationships among staff mem-
bers can promote a sense of community that is critical to promoting positive student
outcomes (King & Newmann, 2000). Unfortunately, for a multitude of reasons,
such as high turnover rates, poor leadership, staff shortages, and inadequate com-
munication systems, the school climate does not always appear supportive or opti-
mistic. The result may be an absence of collegiality and insufficient motivation to
implement the intervention, which can require significant amounts of focus, energy,
and openness to change. Poor school climates almost guarantee a lack of buy-in.
Staff members’ readiness and motivation to increase collaborative family
involvement with the school can be assessed before a partnership intervention is
begun. This evaluation helps pinpoint what additional support may be needed and
which strategies may be used to increase buy-in by school personnel and ultimately,
to facilitate implementation (Gottfredson & Gottfredson, 2002). Particularly in
inadequate school climates, it is helpful if implementers work closely with the
school’s key opinion holders and develop working relationships with administrators
and teachers that facilitate positive family–school practices (Stormshak et al., 2005).
We have found it necessary for implementers to align the key components of an
intervention with the school’s current mission (e.g., the family involvement and
partnership requirement in federal and state regulations, such as Title 1 and IDEA)
and change capacity (Feldstein & Glasgow, 2008). Although it can be quite chal-
lenging to implement and sustain family–school interventions in hostile or resistant
school climates, implementers may be able to combat some of these difficulties by
first prioritizing effective and supportive relationships with staff members at all
36 E.A. Stormshak et al.
levels. Research in this area could broaden the definition of school climate to include
factors such as teacher support, parent involvement, and community-level support.
These factors may buttress successful uptake of programs in schools.
Critical Events
Critical events are serious incidents at a school that may affect training, implementa-
tion, or the program itself. These events are disruptive to the school, learning envi-
ronment, and overall school climate. When these events occur, leadership are
compelled to focus almost entirely on them, and the school enters a “crisis mode”
that interrupts learning and programming until the school recovers from the inci-
dent. During the course of the PFS project, critical incidents were tracked and docu-
mented. The number of incidents was surprising: they occurred in nearly 50 % of
schools. Following are examples of some of these incidents that led to disruption of
the implementation of the model.
• School A principal was involved in a career-altering motorcycle accident.
• Teacher at School B committed suicide.
• Physical fight between parents occurred in the School C family resource center.
• Several project schools experienced strikes; strikes occurred statewide.
• 204 teachers were laid off in the school district that included two middle schools
in the study.
• School E was restructured from sixth to eighth grades to seventh and eighth
grades.
• Eighth grader in School F lost a parent in a multiple homicide.
• Teacher at School G died unexpectedly.
• Sixth grader at School H died in a bus accident.
Although some of these crises were disruptive and divisive and caused a setback
to positive program implementation, some schools were able to use these crises as a
positive opportunity to increase support and collaboration with parents (e.g., using
the universal-level family resource center as a safe room for staff, students, and
parents). Implementation research has not addressed the issue of critical events.
More research is needed in this area to understand the impact that critical events
have on schools, their support of students and families, and their ability to continue
implementing programs.
School staff seldom receive adequate training regarding the influence of family
factors on student outcomes or how to include families in student interventions
(Stormshak et al., 2005). For example, one of the staff members in the PFS study
2 Challenges in Going to Scale 37
Program Integration
Schools may have access to a number of promising intervention programs, yet few
possess a map for how to realistically integrate programs into their daily norms and
routines (Dishion, 2011). Consequently, schools often feel burdened by their
attempts to implement too many individual interventions and may never uptake any
particular program to fidelity. School staff can also be wary of new intervention
efforts because other daily requirements are seldom reduced to accommodate them,
and most often the workload is increased with no commensurate increase in pay. To
make matters worse, because few schools are able to sustain evidence-based pro-
grams with a high degree of fidelity, program effectiveness is thereby decreased
(Durlak & DuPre, 2008), and school staff may not observe positive changes in their
students as a result of intervention efforts. Rarely seeing clear and consistent posi-
tive student outcomes may reduce the likelihood that schools continue to devote
time to applying new interventions. Successful uptake of family–school partnership
programs suffers because developing positive, proactive, and collaborative relation-
ships with families is a potentially difficult endeavor and can seem quite distal to
student achievement.
To reduce the burden of implementation experienced by schools and potentially
increase uptake, program developers must design and test interventions that can
easily be integrated into other efforts and existing school structures. There are sev-
eral ways to address the challenge of increasing the ease of integration and usability
of these programs. For example, family-centered interventions that offer a range of
services, from brief but effective parent contacts to more intensive involvement,
often fit well with referral and intervention systems that already exist in schools
(Stormshak et al., 2005). Similarly, family–school partnership programs that offer a
menu of empirically supported interventions (e.g., brief-focused consultation, two
38 E.A. Stormshak et al.
to three sessions about a parenting topic, multisession parenting group) that can be
accomplished using diverse delivery methods are often accessible to a greater num-
ber of families and school personnel (Stormshak & Dishion, 2009). Offering a range
of intervention options and multiple delivery methods enables schools to more eas-
ily integrate new interventions into their existing routines, thereby increasing the
uptake, penetration of families served, and sustainability of family-centered prac-
tices. Using a bottom-up collaborative approach to intervention design that recog-
nizes the expertise of school staff enables developers and implementers to
successfully integrate intervention efforts and motivate school personnel toward
positive change (Cappella, Jackson, Bilal, Hamre, & Soulé, 2011; Shernoff et al.,
2011). For example, building on the skill and experiences that schools already pos-
sess regarding parent interactions (e.g., a well-attended parent topic night, assign-
ment completion and attendance records proactively provided to parents via
technology) can increase buy-in for integrating new methods of family involvement,
as well as encourage innovation and risk taking among administrators and teachers.
Finally, future research must continue to investigate how schools both struggle and
succeed in managing the educational, social, behavioral, and mental health out-
comes of students (Dishion, 2011). Such information is pivotal to increased under-
standing of how to develop, implement, and sustain family–school partnership
programs that are meaningful and successful for students, parents, and school staff
alike.
Research in the area of implementation has been growing during the past decade,
and multiple studies have been examining circumstances such as principal support,
resources, teacher training, and school climate as primary factors that predict suc-
cessful uptake and implementation. The challenge is that with declining financial
support for schools, research must find a way for programs to be implemented in
spite of few resources and for these models to be sustained over time. Molloy,
Moore, Trail, Epps, and Hopfer (2013) examined schools that had implemented
PBIS to understand factors related to sustaining the model. They found that full
implementation was related to reduced rates of problem behavior, which provides
meaningful support for the model. Smaller schools, elementary schools, and those
with higher SES parents had the best quality implementation; on the other hand,
only 37–49 % of schools implemented the model fully. If we can implement pro-
grams effectively only in high-SES, well-resourced schools, a nation-wide improve-
ment in family–school partnership and quality of education will not occur. Measures
of school capacity are commonly used to evaluate whether schools are “ready” to
implement programs with fidelity (Gingiss et al., 2006). This practice eliminates
schools that are the most disadvantaged and would most benefit from implementing
family-centered practices. Research that focuses on understanding how to imple-
ment family-centered programs in schools with few resources and limited stability
will be important for the future of implementation science.
2 Challenges in Going to Scale 39
Acknowledgment This work was supported by grant R324A09011 from the Institute of
Education Sciences to John Seeley (PI) and Thomas Dishion (PI).
References
Atkins, M. S., Frazier, S. L., Leathers, S. J., Graczyk, P. A., Talbott, E., Jakobsons, L., … Bell,
C. C. (2008). Teacher key opinion leaders and mental health consultation in low-income urban
schools. Journal of Consulting and Clinical Psychology, 76, 905–908. doi:10.1037/a0013036.
Barry, T. D., Lyman, R. D., & Grofer Klinger, L. (2002). Academic underachievement and attention-
deficit/hyperactivity disorder: The negative impact of symptom severity on school performance.
Journal of School Psychology, 40, 259–283. doi:10.1016/S0022-4405(02)00100-0.
Beets, M. W., Flay, B. R., Vuchinich, S., Acock, A. C., Li, K.-K., & Allred, C. (2008). School cli-
mate and teachers? Beliefs and attitudes associated with implementation of the positive action
program: A diffusion of innovations model. Prevention Science, 9(4), 264–275. doi:10.1007/
s11121-008-0100-2.
40 E.A. Stormshak et al.
Biglan, A. (1995). Translating what we know about the context of antisocial behavior into a lower
prevalence of such behavior. Journal of Applied Behavior Analysis, 28, 479–492. doi:10.1901/
jaba.1995.28-479.
Biglan, A., Sprague, J., & Moore, K. J. (2006). A functional contextualist framework for affecting
peer influence practices. In K. A. Dodge, T. J. Dishion, & J. E. Lansford (Eds.), Deviant peer
influences in programs for youth: Problems and solutions (pp. 342–365). New York, NY: The
Guilford Press.
Biglan, A., & Taylor, T. K. (2000). Increasing the use of science to improve child-rearing. The
Journal of Primary Prevention, 21, 207–226. doi:10.1023/A:1007083203280.
Brody, G. H., Flor, D. L., & Gibson, N. M. (1999). Linking maternal efficacy beliefs, developmen-
tal goals, parenting practices and child competence in rural single-parent African American
families. Child Development, 70, 1197–1208. doi:10.1111/1467-8624.00087.
Bums, B. J., Costello, E. J., Angold, A., Tweed, D. L., Stangl, D. K., Farmer, E. M. Z., & Erkanli,
A. (1995). Children’s mental health service use across service sectors. Health Affairs, 14,
147–159.
Cappella, E., Jackson, D. R., Bilal, C., Hamre, B. K., & Soulé, C. (2011). Bridging mental health
and education in urban elementary schools: Participatory research to inform intervention devel-
opment. School Psychology Review, 40, 486–508.
Centers for Disease Control and Prevention. (2013, February 27). Coordinated School Health
Program. Retrieved from http://www.cdc.gov/HealthyYouth/CSHP/
Christenson, S. L. (2003). The family–school partnership: An opportunity to promote the learning
competence of all students. School Psychology Quarterly, 18, 454–482. doi:10.1521/
scpq.18.4.454.26995.
Collins, L., Murphy, S., & Bierman, K. (2004). A conceptual framework for adaptive preventive
interventions. Prevention Science, 5, 185–196. doi:10.1023/B:PREV.0000037641.26017.00.
Connell, A. M., & Dishion, T. J. (2008). Reducing depression among at-risk early adolescents:
Three-year effects of a family-centered intervention embedded within schools. Journal of
Family Psychology, 22, 574–585. doi:10.1037/0893-3200.22.3.574.
Connell, A. M., Dishion, T. J., & Deater-Deckard, K. (2006). Variable- and person-centered
approaches to the analysis of early adolescent substance use: Linking peer, family, and inter-
vention effects with developmental trajectories [Special Issue]. Merrill-Palmer Quarterly, 52,
421–438. doi:10.1353/mpq.2006.0025.
Dishion, T. J. (2011). Promoting academic competence and behavioral health in public schools: A
strategy of systemic concatenation of empirically based intervention principles. School
Psychology Review, 40, 590–597.
Dishion, T. J., & Kavanagh, K. (2003). Intervening with adolescent problem behavior: A family-
centered approach. New York, NY: Guilford.
Dishion, T. J., Nelson, S. E., & Kavanagh, K. (2003). The Family Check-Up with high-risk young
adolescents: Preventing early-onset substance use by parent monitoring. Behavior Therapy, 34,
553–571. doi:10.1016/S0005-7894(03)80035-7.
Dishion, T. J., & Patterson, G. R. (2006). The development and ecology of antisocial behavior in
children and adolescents. In D. Cicchetti & D. J. Cohen (Eds.), Developmental psychopathol-
ogy: Vol. 3. Risk, disorder, and adaptation (pp. 503–541). New York, NY: Wiley.
Dishion, T. J., Patterson, G. R., Stoolmiller, M., & Skinner, M. (1991). Family, school and behav-
ioral antecedents to early adolescent involvement with antisocial peers. Developmental
Psychology, 27, 172–180. doi:10.1037/0012-1649.27.1.172.
Dishion, T. J., Shaw, D. S., Connell, A. M., Gardner, F., Weaver, C. M., & Wilson, M. N. (2008).
The Family Check-Up with high-risk indigent families: Preventing problem behavior by
increasing parents’ positive behavior support in early childhood. Child Development, 79, 1395–
1414. doi:10.1111/j.1467-8624.2008.01195.x.
Dishion, T. J., & Stormshak, E. A. (2007). Intervening in children’s lives: An ecological family-
centered approach to mental health care. Washington, DC: American Psychological Association.
Domitrovich, C. E., Bradshaw, C. P., Poduska, J. M., Hoagwood, K., Buckley, J. A., Olin, S.,
Romanelli, L. H., … Ialongo, N. S. (2008). Maximizing the implementation quality of
2 Challenges in Going to Scale 41
Greenberg, M. T., Domitrovich, C., Graczyk, P., & Zins, J. (2001). The study of implementation in
school-based preventive interventions: Theory, research, and practice. Washington, DC:
Center for Mental Health Services, Substance Abuse and Mental Health Administration,
U.S. Department of Health and Human Services.
Greenberg, M. T., Domitrovich, C., & Bumbarger, B. (2001). The prevention of mental disorders
in school-aged children: Current state of the field. Prevention and Treatment, 4, 1–62.
doi:10.1037/1522-3736.4.1.41a.
Gutman, L. M., & Midgley, C. (2000). The role of protective factors in supporting the academic
achievement of poor African American students during the middle school transition. Journal of
Youth and Adolescence, 29, 233–248. doi:10.1023/A:1005108700243.
Handler, M. W., Rey, J., Connell, J., Their, K., Feinberg, A., & Putnam, R. (2007). Practical con-
siderations in creating school-wide positive behavior support in public schools. Psychology in
the Schools, 44, 29–39. doi:10.1002/pits.20203.
Henderson, A. T., & Berla, N. (1994). A new generation of evidence: The family is critical to stu-
dent achievement (p. 174). Washington, DC: National Committee for Citizens in Education.
Henderson, A. T., & Mapp, K. L. (2002). A new wave of evidence: The impact of school, family
and community connections on student achievement. Austin, TX: Southwest Educational
Laboratory.
Hoagwood, K., Bums, B. J., Kiser, L., Ringeisen, H., & Schoenwald, S. K. (2001). Evidence-based
practice in child and adolescent mental health services. Psychiatric Services, 52, 1179–1189.
doi:10.1176/appi.ps.52.9.1179.
Horner, R. H., Sugai, G., Todd, A. W., & Lewis-Palmer, T. (2005). School-wide positive behavior
support. In L. Bambara & L. Kern (Eds.), Individualized supports for students with problem
behaviors: Designing positive behavior plans (pp. 359–390). New York, NY: Guilford Press.
Kazdin, A. E. (2002). Psychosocial treatments for conduct disorder in children and adolescents. In
P. E. Nathan & J. M. Gorman (Eds.), A guide to treatments that work (2nd ed., pp. 57–85).
London, United Kingdom: Oxford University Press.
Kazdin, A. E. (2010). Problem-solving skills training and parent management training for opposi-
tional defiant disorder and conduct disorder. In J. R. Weisz & A. E. Kazdin (Eds.), Evidence-
based psychotherapies for children and adolescents (pp. 211–226). New York, NY: Guilford
Press.
King, M. B., & Newmann, F. M. (2000). Will teacher learning advance school goals? Phi Delta
Kappan, 81, 576–580.
McDougal, J. L., Clonan, S. M., & Martens, B. K. (2000). Using organizational change procedures
to promote the acceptability of prereferral intervention services: The School-Based Intervention
Team Project. School Psychology Quarterly, 15, 149–171. doi:10.1037/h0088783.
McWayne, C., Hampton, V., Fantuzzo, J., Cohen, H. L., & Sekino, Y. (2004). A multivariate exam-
ination of parent involvement and the social and academic competencies of urban kindergarten
children. Psychology in the Schools, 41, 363–377. doi:10.1002/pits.10163.
Miedel, W. T., & Reynolds, A. J. (1999). Parent involvement in early intervention for disadvan-
taged children: Does it matter? Journal of School Psychology, 37, 379–402. doi:10.1016/
S0022-4405(99)00023-0.
Miller, W. R., & Rollnick, S. (2002). Motivational interviewing: Preparing people for change (2nd
ed.). New York, NY: Guilford.
Molloy, L. E., Moore, J. E., Trail, J., Epps, J. J., & Hopfer, S. (2013). Understanding real-world
implementation quality and “active ingredients” of PBIS. Prevention Science, 14(6), 593–605.
doi:10.1007/s11121-012-0343-9.
Myers, J., & Nastasi, B. (1999). Primary prevention in school settings. In C. R. Reynolds & T. B.
Gutkin (Eds.), Handbook of school psychology (3rd ed.). New York, NY: Wiley.
Patterson, G. R., Capaldi, D., & Bank, L. (1991). An early starter model for predicting delin-
quency. In D. J. Pepler & K. H. Rubin (Eds.), The development and treatment of childhood
aggression (pp. 139–168). Hillsdale, NJ: Erlbaum.
Patterson, G. R., Reid, J. B., & Dishion, T. J. (1992). A social learning approach: Vol 4. Antisocial
boys. Eugene, OR: Castaglia.
2 Challenges in Going to Scale 43
Payne, A. A. (2008). A multilevel analysis of the relationships among communal school organiza-
tion, student bonding, and delinquency. Journal of Research in Crime and Delinquency, 45,
429–455. doi:10.1177/0022427808322621.
Payne, A. A., & Eckert, R. (2010). The relative importance of provider, program, school, and com-
munity predictors of the implementation quality of school-based prevention programs.
Prevention Science, 11, 126–141. doi:10.1007/s11121-009-0157-6.
Prinz, R. J., & Dumas, J. E. (2004). Prevention of oppositional-defiant disorder and conduct disor-
der in children and adolescents. In P. Barrett & T. H. Ollendick (Eds.), Handbook of interven-
tions that work with children and adolescents: From prevention to treatment (pp. 475–488).
Chichester, UK: Wiley.
Prinz, R. J., & Sanders, M. R. (2007). Adopting a population-level approach to parenting and
family support interventions. Clinical Psychology Review, 27, 739–749. doi:10.1016/j.cpr.
2007.01.005.
Reynolds, A. J. (1992). Comparing measures of parental involvement and their effects on aca-
demic achievement. Early Childhood Research Quarterly, 7, 441–462. doi:10.1016/0885-
2006(92)90031-S.
Rimm-Kaufman, S. E., & Pianta, R. C. (2000). Patterns of family–school contact in preschool and
kindergarten. School Psychology Review, 28, 426–438.
Ringeisen, H., Henderson, K., & Hoagwood, K. (2003). Context matters: Schools and the “research
to practice” gap in children’s mental health. School Psychology Review, 32, 153–168.
Rogers, E. M. (2003). Diffusion of Innovations (5th ed.). New York, NY: Free Press.
Rumberger, R. W. (1995). Dropping out of middle school: A multilevel analysis of students and
schools. American Educational Research Journal, 32, 583–625. doi:10.2307/1163325.
Ryan, R. M., Martin, A., & Brooks-Gunn, J. (2006). Is one good parent good enough? Patterns of
mother and father parenting and child cognitive outcomes at 24 and 36 months. Parenting:
Science and Practice, 6, 211–228. doi:10.1207/s15327922par0602&3_5.
Shaw, M. C. (1986). The prevention of learning and interpersonal problems. Journal of Counseling
and Development, 64, 624–627. doi:10.1002/j.1556-6676.1986.tb01024.x.
Sheridan, S. M., Knoche, L. L., Kupzyk, K. A., Pope Edwards, C., & Marvin, C. A. (2011). A
randomized trial examining the effects of parent engagement on early language and literacy:
The Getting Ready intervention. Journal of School Psychology, 49, 361–383. doi:10.1016/j.
jsp.2011.03.001.
Shernoff, E. S., Maríñez-Lora, A., Frazier, S. L., Jakobsons, L., Atkins, M. S., & Bonner, D.
(2011). Teachers supporting teachers in urban schools: What iterative research designs can
teach us. School Psychology Review, 40, 465–485.
Shirk, S. R., & Jungbluth, N. J. (2008). School-based mental health checkups: Ready for practical
action. Clinical Psychology: Science and Practice, 5, 217–223. doi:10.1111/j.1468-
2850.2008.00131.x.
Spoth, R. L., Kavanagh, K. A., & Dishion, T. J. (2002). Family-centered preventive intervention
science: Toward benefits to larger populations of children, youth, and families. Prevention
Science, 3, 145–152. doi:10.1023/A:1019924615322.
Spoth, R., Trudeau, L., Shin, C., Ralston, E., Redmond, C., Greenberg, M., & Feinberg, M. (2013).
Longitudinal effects of universal preventive intervention on prescription drug misuse: Three
randomized controlled trials with late adolescents and young adults. American Journal of
Public Health, 103, 665–672. doi: 10.2105/AJPH.2012.301209.
Stormshak, E. A., Connell, A. M., Véronneau, M-H., Myers, M. W., Dishion, T. J., Kavanagh, K.,
& Caruthers, A. S. (2011). An ecological approach to promoting early adolescent mental health
and social adaptation: Family-centered intervention in public middle schools. Child
Development, 82, 209–225. doi: 10.1111/j.1467-8624.2010.01551.x.
Stormshak, E. A., Connell, A. M., & Dishion, T. J. (2009). An adaptive approach to family-centered
intervention in schools: Linking intervention engagement to academic outcomes in middle and
high school. Prevention Science, 10, 221–235. doi:10.1007/s11121-009-0131-3.
Stormshak, E. A., & Dishion, T. J. (2002). An ecological approach to child and family clinical and
counseling psychology. Clinical Child and Family Psychology Review, 5, 197–215. doi:10.102
3/A:1019647131949.
44 E.A. Stormshak et al.
Steven B. Sheldon
Focusing on Partnerships
Schools and school districts need to collaborate with and support students’ families
and communities because of their significant impact on student learning and aca-
demic performance. Family involvement and parenting, for example, consistently
predicts stronger student outcomes such as higher academic achievement, more
regular school attendance, and greater student engagement (Epstein et al., 2010;
Henderson & Mapp, 2002; Pomerantz, Moorman, & Litwack, 2007). Looking at
school performance through the decentralization reform in Chicago, Bryk and his
colleagues found that schools that flourished tended to be characterized by strong
relationships with students’ families and community (Bryk, Seabring, Allensworth,
Luppescu, & Easton, 2010). Their extensive analyses demonstrate that strong, posi-
tive school–family–community relationships are an essential element of school
improvement. Efforts to improve schools and student achievement need to devote
attention to the manner in which schools are communicating and engaging families
and to consider these practices an important part of any movement to reform and
improve school systems.
For schools’ efforts to engage families and community partners to be successful,
they need to be coordinated, systematic, and organized. A programmatic and orga-
nized approach to school, family, and community partnerships contributes to the
ability of these efforts to produce meaningful student and family outcomes (Epstein
et al., 2010). Sheldon (2005), for example, showed that schools using an organiza-
tional approach that includes teamwork, annual planning, and evaluation tend to
conduct stronger outreach to families; and elsewhere found that these schools tend
to report higher levels of family involvement at school (Sheldon, 2007a, 2009;
Sheldon & Van Voorhis, 2004). Also, compared to schools that are not working to
engage families in an organized and systematic manner, those using this organizational
framework had greater gains in student daily attendance from 1 year to the next
(Sheldon, 2007b). Together these studies illustrate the importance of schools using a
programmatic approach to family engagement focused on meeting students’ goals.
Schools’ approach to family and community engagement is strongly influenced
by the surrounding administrative context and community. This chapter focuses on
how the context in which schools are embedded shape educators’ approach to
implementing family and community engagement practices. Specifically, I focus on
3 Moving Beyond Monitoring: A District Leadership Approach… 47
the role of the central school district in developing and sustaining school programs
for school, family, and community partnerships. I provide a framework from which
to understand the different ways central district leaders can promote family engage-
ment in schools. In doing so, this chapter offers a framework that provides practitio-
ners examples of strong district practices. Additionally, the framework provides
researchers a starting point from which studies might further develop theoretical
and empirical understandings about how districts can help strengthen school–family
relationships and, ultimately, student and school outcomes.
Schools that are not intentional about their efforts to engage families will not
persist with these efforts and will not engage all families equally (Epstein &
Sheldon, 2006). At Johns Hopkins University, the National Network of Partnership
Schools (NNPS) has worked with schools and school districts around the United
States and internationally to create organized programs of school, family, and
community partnerships (see Epstein et al., 2010). Schools are encouraged to
implement four organizational components that help embed partnerships into the
fabric of schools: Teamwork, Goal-Oriented Plans, Responsive Implementation,
and Program Evaluation.
Teamwork
One reason school-based programs for school, family, and community partnerships
are not implemented and sustained more widely is because the responsibility of this
work often rests with a single individual in a school. In these cases the most likely
result is that the partnership work is not school-wide, the person responsible
becomes burnt out and leaves after a short time, or both. Instead, school leaders
need to establish a committee or team to take ownership of the partnership work and
help teachers, staff, and families coordinate their efforts to engage all families and
community members.
As a first step in establishing a school-wide partnership program, NNPS recom-
mends that schools form a team dedicated to coordinate the partnership efforts
throughout the school—An Action Team for Partnerships (ATP). The ATP members
include teachers, school administrators, parents, community members, and, at the
high school level, students. To ensure that the partnership efforts at the school rein-
force, rather that distract from, other school improvement efforts, at least one mem-
ber of the ATP should also be a member of the school improvement team or council
(SIT). The ATP chair attends SIT meetings and communicates with the school prin-
cipal, teacher leaders, and the PTO/PTA president about ATP plans and goals. Also,
ATPs need parental representation from all groups throughout the school
community.
48 S.B. Sheldon
Goal-Oriented Plans
Responsive Implementation
Effective family and community engagement activities are responsive and sensitive
to the factors that limit family member’s participation. Researchers have demon-
strated variation in family engagement according to the education levels of the
child, educational attainment of the parents, family structure, and language spoken
at home (Astone & McLanahan, 1991; Dauber & Epstein, 1993; Eccles & Harold,
3 Moving Beyond Monitoring: A District Leadership Approach… 49
Evaluation
Finally, school and action team leaders need to conduct on-going and end-of-year
evaluations of their partnership program and practices. In the current context of
education, whatever gets measured gets done. Evaluating a partnership program,
therefore, is essential to the development and sustainability of family engagement
efforts. ATP members, by evaluating their partnership program, are able to identify
the strengths and weaknesses of their program; demonstrate outcomes from the
activities implemented; and send a message that partnerships are valued at the
school. Studies demonstrate that partnership programs are more likely to improve
and maintain a higher level of quality if the Action Team participates in an end-of-
year evaluation of the program, and if feedback is obtained from families participat-
ing in family engagement activities (Sheldon, 2007a).
Contextual Influences
Extensive research has been conducted to understand the school-level factors that
predict the development of strong partnership programs. Repeatedly, studies have
found that partnership programs need the support of the school principal to provide
school staff the time, space, and encouragement to reach out and engage families in
their children’s schooling (Sanders & Harvey, 2002; Sanders & Sheldon, 2009;
Sheldon & Van Voorhis, 2004). In this way, school partnership programs are like
most other school improvement strategies.
Another important influence on school practices to engage family and community
members is the support from the central district offices. Numerous studies show that
50 S.B. Sheldon
schools reporting greater support for partnerships from the district tend to have
stronger partnership programs and are better able to sustain those programs (Epstein
et al., 2010; Sanders, 2012; Sheldon, 2005, 2007a). Epstein, Galindo, and Sheldon
(2011) found district reports of their efforts to promote partnerships in schools cor-
responded to stronger school reports of program implementation and outreach to
families. Also, they found that district facilitation helped predict schools’ imple-
mentation of practices most responsive to the challenges families face when trying
to support their children’s schooling. The studies, however, provide little guidance
about how district personnel can provide that support.
District support has also been shown to be vital to improve classroom instruc-
tion. Drawing on organizational and social-learning theories, researchers have
explored the ways school district personnel enable school leaders to implement
school reform practices to improve teaching and learning (Burch & Spillane, 2004;
Honig, 2003, 2006, 2012; Knapp, 2008). This work demonstrates that practices of
school district personnel have the ability to improve classroom instruction and
shape educators’ approaches to using data to improve student achievement.
For educators to improve instruction and student learning, it may be necessary
for district personnel to change the nature of their relationship with schools.
Historically, and predominantly still, district personnel have maintained a relation-
ship with schools founded on monitoring and oversight (Honig, 2006, 2012). Rather
than work with principals or other school leaders, and consistent with the sanction-
based policy approach, central district personnel have tended to focus their efforts
on ensuring policy compliance. As a result, districts are often perceived to be an
additional burden for schools and not a resource to spur school improvement.
In some cases, however, central district personnel have changed the nature of
their relationship with school leaders, helping them improve instructional leader-
ship in schools. Studying several of these districts, Honig (2012) was able to cate-
gorize specific practices used by district personnel to promote principals’
instructional leadership: (1) Joint Work—helping principals value their role as
instructional leaders at their school, (2) Differentiation—providing support to meet
the individual needs of principals, (3) Modeling—demonstrating practices and
meta-cognitive strategies principals can use at their school, (4) Developing Tools—
providing principals with documents and templates that help guide principals’
actions as instructional leaders, and (5) Brokering—connecting principals to new
information and/or individuals, as well as buffering them from competing demands
to help them focus on instructional leadership in their schools. These practices
illustrate how the central district offices, moving beyond monitoring, can support
school leaders and facilitate practices to help promote strong instruction and stu-
dent learning. One question that researchers need to explore, however, is the extent
to which these types of district practices can be applied to the development of
strong school, family, and community relationships; and the degree to which this
capacity-building approach to partnerships translates into educational practice and
student outcomes.
3 Moving Beyond Monitoring: A District Leadership Approach… 51
Similar to instructional leadership, the district role with regard to school, family,
and community partnerships tends to emphasize monitoring over supporting school
practices. In working with and researching schools and districts at NNPS, we have
discovered that district support and outreach to schools is essential to the develop-
ment and scale-up of school partnership programs. The NNPS provides a frame-
work to help district personnel understand how to reorient their work and interactions
with schools and school leaders. This framework is based on the work by research-
ers and facilitators who collaborate with and support districts across the United
States. Schools and districts that join NNPS follow the research-based approach
described above to involve all families in their children’s education.
Researchers at NNPS collect an annual UPDATE survey from its members,
which measures program implementation and support of school, family, and com-
munity partnerships. As stated previously, analyses have shown that schools nested
within districts that support the implementation of comprehensive partnership pro-
grams receive more and better support from families than schools trying to involve
parents without district support (Hutchins & Sheldon, 2013). These findings have
spurred greater attention to the work of district personnel and how they can support
partnership work in schools.
The UPDATE survey data collected from district leaders show that their prac-
tices generally fall under two broad categories: district-level leadership and direct
facilitation of schools. District-level leadership helps to ensure that home-school
collaboration is not siloed but integrated throughout other offices within the district,
and that it is coordinated across schools throughout the district. Examples of district-
level leadership activities include: reviewing a district policy on family involve-
ment, conducting staff development on partnerships, coordinating a District
Advisory Council on family and community involvement, and collecting best prac-
tices from schools to share throughout the district. Direct facilitation of schools
provides support to school and ATP leaders who work to implement comprehensive
school, family, and community partnership programs for student success. Examples
of facilitation initiatives include: conducting one-day workshops for ATPs, making
monthly visits or contacts with school ATPs, scheduling an annual meeting with
building principals, and helping ATPs evaluate their programs and progress. Both
types of district leadership are important for schools to implement and sustain
school, family, and community partnership programs (Epstein et al., 2010).
Studies and fieldwork with diverse districts across the United States indicate six
strategies (see Table 3.1) that help district leaders to organize their partnership
work: create awareness, align program and policy, guide learning and program
development, share knowledge, celebrate milestones, and document progress and
evaluate outcomes (Epstein, 2008; Epstein et al., 2010). These six objectives weave
through the two primary categories of district leaders’ work: district-level leader-
ship and direct facilitation of schools’ efforts. Below are examples of activities
52 S.B. Sheldon
central district personnel conduct around the six strategies of district leadership and
facilitation of schools.1 These strategies are not mutually exclusive of one another,
as one district practice might capture two or more strategies.
Create Awareness
Creating awareness occurs when district personnel are actively promoting the dis-
trict’s partnership program to all key stakeholders, including teachers, administra-
tors, families, and community groups. Activities that create awareness include
convening a one-on-one meeting between the district leader for family and com-
munity engagement and the district’s superintendent to discuss the goals of the part-
nership program and disseminating a press release announcing the launch of the
district’s partnership initiative. Creating awareness is an important step to help
potential partners understand the district’s goals for its partnership program.
District Leadership
Direct Facilitation
Many school districts create awareness through monthly newsletters to school per-
sonnel. Pasco School District, in Pasco, Washington, disseminates its monthly ATP
Connection to school leaders and other interested stakeholders. The two-page
newsletter has five sections: District News, Spotlight On, Read All About It,
Resource Corner, and Important Dates. ATP Connection helps Pasco’s 19 schools to
stay connected, coordinate dates of activities, and share promising practices.
1
The National Network of Partnership Schools solicits examples of promising partnership prac-
tices from its members annually. The examples of activities in this section are taken from various
collections of Promising Partnership Practices. To read the latest edition of Promising Partnership
Practices, go to www.partnershipschools.org.
54 S.B. Sheldon
District leaders for partnerships are encouraged to work with other district and
school leaders to integrate the partnership program with other district policies,
requirements, and procedures. Examples of activities aimed at aligning program
and policy include identifying a budget to implement the district’s partnership pro-
gram and developing district policies so that work and progress on family and com-
munity involvement is one component of the evaluations of principals and teachers.
Aligning program and policy necessitates cross-departmental collaboration.
District Leadership
Francis Howell School District, in Saint Charles, Missouri, aligns program and
policy through its District Parent Involvement Advisory group. The group meets
three times a year and brings together representatives of partnership programs from
the district’s 23 schools. Members of the advisory group share promising practices,
discuss challenges, and make joint plans. In addition, the district superintendent,
chief academic officer, and chief financial officer report on current conditions and
answer questions. School board members also participate. District leaders consider
the Parent Involvement Advisory meetings crucial to student success, especially in
a district small enough for everyone to meet occasionally in one place.
Direct Facilitation
With many states adopting the Common Core, district leaders are developing
resources for families to understand how to support student learning at home. St.
Paul Public Schools, located in St. Paul, Minnesota, addressed this need through its
Learning Standards for Families booklets. Teachers in St. Paul volunteered their
time and resources to create the booklets, which were then translated from English
into Spanish, Hmong, and Somali. The booklets help to meet a number of parent
involvement goals, including defining parental roles in supporting the academic
standards and promoting consistency across schools and programs in support of the
standardized curriculum. The resource also helps to ease the transition between
schools for highly mobile students.
3 Moving Beyond Monitoring: A District Leadership Approach… 55
Guiding learning and program development involves the district leader organizing
and conducting professional development activities to develop school-based
partnership programs. Many districts in NNPS conduct an initial one-day action
team training and then hold “refresher” courses for new ATP members in the follow-
ing years or semesters. District leaders may also attend the schools’ monthly ATP
meetings and conduct quarterly cluster meetings with ATP chairs or co-chairs.
Some districts conduct virtual meetings via Skype or Adobe Connect if they span a
large geographic distance.
District Leadership
In Cecil County Public Schools, in Elkton Maryland, the district awards mini-
grants to schools to help fund family engagement practices. Schools apply for a
mini-grant by mid-June to fund a practice they will implement the following year.
The program is sponsored by the Board of Education. The Board expanded the
award program after seeing the successful results from year to year. District per-
sonnel are responsible for promoting the program to schools, reading proposals
and making awards, monitoring the awards, and evaluating the family engagement
projects.
Direct Facilitation
Share Knowledge
Districts that are able to sustain their partnership programs often help diverse groups
of stakeholders exchange knowledge with one another. This strategy involves fos-
tering ongoing communication throughout the district to build knowledge about
practices and programs of school, family, and community partnerships. As with
guiding learning and program development, districts share knowledge virtually and
through face-to-face interactions.
District Leadership
The Arizona State Parent Information and Resource Center (PIRC), located in
Gilbert, Arizona, worked with school sites across the entire state. In order to build a
statewide network of partnerships, the PIRC developed a virtual Share Center. The
Share Center was an online forum that allowed school leaders in Arizona to share
ideas with each other and to access information posted by schools and services from
outside the state. Before items are available for public viewing, PIRC staff review
the submissions. The Share Center includes a growing catalog of partnership ideas
for activities and events, presentations, handouts, newsletters, templates, activities,
and other materials.
For many years, Fort Worth Independent School District, located in Fort Worth,
Texas, had a Roving Resource Center that served parent liaisons and counselors
with a lending library of books and media for use in workshops with parents. To
improve the Center, the district found space in a centrally located middle school for
a stationary Parent Engagement Outreach Center and can now also serve families
directly. In addition to the traditional book and media library, the Center has a
“make and take” station where parents gain ideas about how to help their children
with learning at home; computer and audio stations; and workshops for parents
about nutrition, parenting, college planning, and other important topics.
3 Moving Beyond Monitoring: A District Leadership Approach… 57
Direct Facilitation
Celebrate Milestones
District Leadership
Direct Facilitation
Francis Howell School District, in Saint Charles, Missouri, makes sure that
celebrations are a community affair. Each year, the district conducts “Howell-a-
Palooza” to showcase student work, promote community resources, and strengthen
school–community bonds. The first celebration, held in 2008, drew over 5000 peo-
ple. The event offers an array of activities for all ages, including high school bands
performances, choice performances, school showcases, and more than 100 booths
featuring community organizations. The annual celebration promotes partnerships
within and across the district and promotes family engagement as well as highlights
community resources.
District Leadership
Direct Facilitation
Some districts distribute data notebooks for each school to document progress and
evaluation outcomes. In Pasco Public Schools, in Pasco Washington, the district
leaders created an organizational and record-keeping notebook for each ATP to help
the school keep track of its partnership plans and progress. Each binder included the
following sections: Training Materials, One-Year Action Plan, Meeting Minutes,
Newsletters, ATP Contact Information, Resources (Help!), Evaluation Tools and
Correspondence & Miscellaneous. At the end of each year, the district collects the
binders and assesses the progress on school, family, and community partnerships at
each school. They also prepare the binder for the next year, adding and updating
information for the new and returning ATP members.
Studies are also needed that explore the impact of this district approach to family
and community engagement on principals and other school leaders. Because school
principals are so important to the implementation of school, family, and community
partnership programs (Sanders & Sheldon, 2009), understanding how district lead-
ership shapes principal leadership within schools is vital. This research might adopt
a cognitive or interpretive framework (Spillane, Diamond, et al. 2002; Spillane,
Reiser & Reimer, 2002) to study whether and how school leaders change their
beliefs and practices related to the engagement of family and community members
in student learning through district facilitation of partnership programs. An
interpretive framework approach would not only examine district leaders’ work, but
may help uncover how school leaders are making sense and implementing family
engagement practices in light of that district support.
In addition to examining how district leadership and facilitation for partnerships
affect school leader attitudes and behaviors, studies are needed to understand how
organizational dynamics within schools promote the development of strong school,
family, and community partnerships. In particular, studies might examine how dis-
trict leaders can use the social networks among educators to promote the diffusion
of partnership efforts by teachers and develop a school climate supportive of part-
nerships. Teachers’ social networks have been shown to affect their implementation
of instructional practices and adoption of technology (Diamond, 2012; Frank, Zhao,
& Borman, 2004); however, how teachers’ formal and informal interactions shape
their attitudes and practice toward students’ families requires additional research.
To more fully understand the impact of school districts on family and community
engagement, studies are needed that look at ways in which districts are engaging
families directly. Currently, many school districts implement a parent education/
leadership training course hoping to engage families in their children’s school or
learning at home. These workshops, sometimes called Parent Universities, are
aimed at educating family members about child development and the school sys-
tem, as well as empowering them to advocate for their children at home and at
school (Henderson, 2010). Recently, the Harvard Family Research Project reported
that over 140 parent university groups exist across the United States (http://www.
hfrp.org/hfrp-news/news-announcements/parent-university-network retrieved on
May 27, 2014). Very few studies, however, have examined the extent to which these
empowerment programs are motivating family members to become involved as
leaders at their children’s school, district, or in city governance. As this form of
direct district engagement with families extends to more and more locales, better
research is needed to understand the extent to which these efforts are associated
with family, student, and school outcomes.
Also, increasingly central district offices are collaborating with community orga-
nizations to support students and families. Some important research has begun to
explore the nature and potential of district-community collaborations, demonstrating
3 Moving Beyond Monitoring: A District Leadership Approach… 61
the benefits of these relationships. For example, in their book, A Match on Dry
Grass, Warren and Mapp (2011) provide examples of how community organizing
groups have worked with schools and districts to promote equity and student
achievement in districts across the country. Similarly, Sanders (2009) showed how
community groups can work with and support districts to maintain an emphasis on
family and community engagement. These studies, however, have not attempted to
look at the impact of district-community collaborations on family engagement at
home or at the school, nor has there been research connecting these efforts to school
or student outcomes.
Across districts, increasing emphasis is being placed on using data and data-
driven processes to improve classroom instruction and student outcomes. District
offices are using early warning indicators such as attendance, behavior, and course
credits to identify those student most at risk of dropping out (Allensworth &
Easton, 2007; Mac Iver & Messel, 2013). Being at risk, however, does not mean
that a student cannot succeed in school. Mac Iver and Messel (2013) found that
eighth graders who exhibited an early warning indicator but who, then, did not in
ninth grade were as likely to graduate as students who never exhibited an early
warning indicator. Combining district or school practices of family and community
engagement with this data-driven educational approach, however, has not been
attempted and represents an important method by which schools and district offices
can collaborate to improve student outcomes like attendance, behavior, grades, and
graduation rates.
Finally, in addition to basic research on the impact of district strategies on
schools, researchers are encouraged to investigate the effectiveness of specific prac-
tices and strategies across districts. Studies are needed, for example, that investigate
whether and which practices within the six strategies are most effective for rural,
large urban, and/or suburban districts; which are more or less effective for districts
with varying degrees of racial or ethnic heterogeneity; and which strategies are most
effective for districts with high percentages of immigrant families. These types of
studies would significantly help district leaders choose practices most likely to suc-
ceed in their community.
Conclusions
It is clear that without on-going support from school district personnel, school lead-
ers will struggle to develop and maintain strong programs of school, family, and
community partnerships. In presenting the framework of district strategies, I offer
methods for structuring partnership work in applied settings and add to policymak-
ers’ understandings about the role of district leadership for school reform and
improvement. Most importantly, perhaps, I argue for the need to expand research
into the role of the central district office; studying the impact of district leadership
and collaboration among colleagues, facilitation of school programs, and district
programs intended to directly empower families. How district personnel construct
their work with school leaders around partnerships, moving beyond monitoring, is
62 S.B. Sheldon
also an aspect of education policy and practice that needs greater attention from
researchers. District leadership for partnerships is likely to have the greatest impact
on whether and how school leaders approach their engagement with families and
the community when district personnel operate as an active and supportive resource
for schools.
References
Allensworth, E., & Easton, J. (2007). What matters for staying on-track and graduating in Chicago
public high schools. Chicago, IL: Consortium on Chicago School Research.
Astone, N. M., & McLanahan, S. S. (1991). Family structure, parental practices, and high school
completion. American Sociological Review, 56, 309–320.
Bryk, A. S., Seabring, P., Allensworth, E., Luppescu, S., & Easton, J. Q. (2010). Organizing
schools for improvement: Lessons from Chicago. Chicago, IL: University of Chicago Press.
Burch, P., & Spillane, J. (2004). Leading from the middle: Mid-level district staff and instructional
improvement. Chicago, IL: Cross City Campaign for Urban School Reform.
Dauber, S. L., & Epstein, J. L. (1993). Parents’ attitudes and practices of involvement in inner-city
elementary and middle schools. In N. Chavkin (Ed.), Families and schools in a pluralistic
society (pp. 53–71). Albany, NY: SUNY Press.
Diamond, J. B. (2007). Where the rubber meets the road: Rethinking the connection between high
stakes accountability policy and classroom instruction. Sociology of Education, 80, 285–313.
Diamond, J. B. (2012). Accountability policy, school organization, and classroom practice: Partial
recoupling and educational opportunity. Education and Urban Society, 44, 151–182.
Eccles, J. S., & Harold, R. D. (1996). Family involvement in children’s and adolescents’ schooling.
In A. Bloom & J. F. Dunn (Eds.), Family-school links: How do they affect educational out-
comes? (pp. 3–34). Mahwah, NJ: Erlbaum.
Epstein, J. L. (2008). Research meets policy and practice: How are school districts addressing
NCLB requirements for parental involvement? In A. R. Sadovnik, J. O’Day, G. Bohrnstedt, &
K. Borman (Eds.), No child left behind and the reduction of the achievement gap: Sociological
perspectives on federal educational policy (pp. 267–279). New York, NY: Routledge.
Epstein, J. L., Galindo, C. L., & Sheldon, S. B. (2011). Levels of leadership: Effects of district and
school leaders on the quality of school programs of family and community involvement.
Educational Administration Quarterly, 47, 462–495.
Epstein, J. L., Sanders, M. G., Sheldon, S. B., Simon, B. S., Salinas, K. C., Jansorn, N. R., …,
Williams, K. J. (2010). School, family, and community partnerships: Your handbook for action
(3rd ed.). Thousand Oaks, CA: Corwin Press.
Epstein, J. L., & Sheldon, S. B. (2006). Moving forward: Ideas for research on school, family, and
community partnerships. In C. F. Conrad & R. Serlin (Eds.), SAGE handbook for research in
education: Engaging ideas and enriching inquiry (pp. 117–137). Thousand Oaks, CA: Sage.
Frank, K. A., Zhao, Y., & Borman, K. (2004). Social capital and the diffusion of innovations within
organizations: Application to the implementation of computer technology in schools. Sociology
of Education, 77, 148–171.
Henderson, A. T. (2010). Building local leadership for change: A national can of parent leadership
training programs. Providence, RI: Annenberg Institute for School Reform at Brown University.
Henderson, A. T., & Mapp, K. L. (2002). A new wave of evidence: The impact of school, family,
and community connections on student achievement. Austin, TX: Southwest Educational
Development Laboratory.
Honig, M. I. (2003). Building policy from practice: District central office administrators’ roles and
capacity for implementing collaborative education policy. Educational Administration
Quarterly, 39, 292–338.
3 Moving Beyond Monitoring: A District Leadership Approach… 63
Honig, M. I. (2006). Street-level bureaucracy revisited: Frontline district central office administra-
tors as boundary spanners in education policy implementation. Educational Evaluation and
Policy Analysis, 28, 357–383.
Honig, M. I. (2012). District central office leadership as teaching: How central office administra-
tors support principals’ development as instructional leaders. Educational Administration
Quarterly, 48, 733–774.
Hutchins, D. J., & Sheldon, S. B. (2013). Summary 2012 school data. Baltimore, MD: National
Network of Partnership Schools at Johns Hopkins University.
Knapp, M. S. (2008). How can organizational and sociocultural learning theories shed light on
district instructional reform? American Journal of Education, 114, 521–539.
Lareau, A. (2003). Unequal childhoods: Class, race, and family life. Berkeley, CA: University of
California Press.
Mac Iver, M. A., & Messel, M. (2013). The ABCs of keeping on track to graduation: Research
findings from Baltimore. Journal of Education for Students Placed at Risk, 18, 50–67.
Mintrop, H., & Sunderman, G. L. (2009). Predictable failure of federal sanction-driven account-
ability for school improvement and why we maintain it anyway. Educational Researcher, 38,
355–364.
Pomerantz, E. M., Moorman, E. A., & Litwack, S. D. (2007). The how, whom, and why of parents’
involvement in children’s schooling: More is not necessarily better. Review of Educational
Research, 77, 373–410.
Sanders, M. G. (2009). Collaborating for change: How an urban school district and a community-
based organization support and sustain school, family, and community partnerships. Teachers
College Record, 111, 1693–1712.
Sanders, M. G. (2012). Sustaining programs of school, family, and community partnerships:
A qualitative longitudinal study of two districts. Educational Policy, 26, 845–869.
Sanders, M. G., & Harvey, A. (2002). Beyond the school walls: A case study of principal leader-
ship for school–community collaboration. Teachers College Record, 104, 1345–1368.
Sanders, M. G. & Sheldon, S. B. (2009). Principals Matter: A Guide to School, Family, and
Community Partnerships. Thousand Oaks, CA: Corwin Press.
Sheldon, S. B. (2005). Testing a structural equation model of partnership program implementation
and parent involvement. The Elementary School Journal, 106, 171–187.
Sheldon, S. B. (2007a). Improving student attendance with a school-wide approach to school–family–
community partnerships. Journal of Educational Research, 100, 267–275.
Sheldon, S. B. (2007b). Getting families involved with NCLB: Factors affecting schools’ enact-
ment of federal policy. In A. R. Sadovnik, J. O’Day, G. Bohrnstedt, & K. Borman (Eds.),
No child left behind and the reduction of the achievement gap: Sociological perspectives on
federal educational policy. New York, NY: Routledge.
Sheldon, S. B. (2009). Using evaluation to prove and improve the quality of partnership pro-
grammes in schools. In R. Deslandes (Ed.), International perspectives on contexts, communi-
ties and evaluated innovative practices: Family-school-community partnerships (pp. 126–142).
New York, NY: Routledge.
Sheldon, S. B. & Epstein, J. L. (2007a). Elementary and Middle School Family and Community
Involvement Surveys: Student Questionnaires. Report from the Center on School, Family, and
Community Partnerships. Johns Hopkins University: Baltimore, MD.
Sheldon, S. B. & Epstein, J. L. (2007b). Elementary and Middle School Family and Community
Involvement Surveys: Parent Questionnaires. Report from the Center on School, Family, and
Community Partnerships. Johns Hopkins University: Baltimore, MD.
Sheldon, S. B., & Van Voorhis, F. L. (2004). Partnership programs in U. S. Schools: Their develop-
ment and relationship to family involvement outcomes. School Effectiveness and School
Improvement, 15, 125–148.
Spillane, J. P., Diamond, J. B., Burch, P., Hallett, T., Jita, L., & Zoltner, J. (2002). Managing in the
middle: School leaders and the enactment of accountability policy. Educational Policy, 16,
731–762.
Spillane, J. P., Reiser, B., & Reimer, T. (2002). Policy implementation: A cognitive model. Review
of Educational Research, 72, 387–431.
Warren, M. R., & Mapp, K. L. (2011). A match on dry grass. New York, NY: Oxford University Press.
Chapter 4
Research Issues to Forward a Policy Agenda
in Support of Family–School Partnerships
Lisa L. Knoche
Introduction
Within the research community, there is general consensus on how research findings
are assessed, and how they contribute to future research directions. The intent of
family–school partnership research, for example, is to systematically build a knowl-
edge base for understanding behaviors and actions of individuals and schools that
contribute to engagement and partnership practices. Researchers are steeped in the
intricacies of content related to family–school partnerships. Researchers can com-
mit lengthy periods of time to investigating a single research question. Policy
makers, and their intermediaries, are also accessing research, but using it for an
alternative purpose. Their intent is not to build a knowledge base, but instead is
primarily focused on action. The timeline for policy makers is often swift; informa-
tion is needed quickly and decisions might be rushed. Given these different orienta-
tions, it is worthwhile to first specify policy makers’ approach for using research
information and detail how information is processed to contribute to policy
decisions.
1
Public policy is generally defined as a system of laws, regulatory measures, and funding priorities
concerning a given topic put in place by governmental entities to cause action in an effort to
achieve some social goal (Smith & Larimer, 2009). Policy makers work in a variety of settings, and
include government representatives at all levels of local, state and federal government. Policy
makers include individuals such as school board members, city council representatives, adminis-
trative leaders within state or federal organizations, and legislators at all levels of government.
4 Policy Agenda 67
Examining the ways that research findings might be used by policy makers is a first
step in understanding the interface between policy and research (Nutley, Walter, &
Davies, 2007; Tseng, 2012). First, information can be used by policy makers for
instrumental purposes. That is, research findings are used to directly inform a policy
decision. A decision maker might seek out relevant research information and shape
his legislation or funding priority according to the available research findings.
Information can also be used in a political way to reinforce or refute a given position
on a policy-related decision. That is, a policy maker uses the findings from research
studies to back up a position she already holds—not to establish a new position, or
explore possible options. When used for developing a greater understanding of
issues, policy makers are using research conceptually. Research is contributing to
the overall perspective of the policy maker on a given topic, but not directly contrib-
uting to any action. Finally, the imposed use of research evidence (Weiss, Murphy-
Graham, & Birkeland, 2005) is becoming increasingly prevalent as evidence-based
practices and programs are being incorporated into many government programs.
Policy makers are demanding that evidence be a criterion in determining actions.
For researchers, it is important to understand that research information can and
will be used in these different ways. It underscores the importance of providing data
in multiple forms, for multiple audiences. Ideally, the research community would
like rigorous research evidence to play a foundational role in policy decisions; how-
ever, evidence is only one aspect considered by decision makers. There are many
competing priorities; policy makers are responsible to key stakeholders and con-
stituent bases with numerous interests and varying goals (Shonkoff, 2000). The
current ideological climate is often highly influential; research evidence may con-
tribute little to shape policy decisions (Shonkoff, 2000; Weiss, Murphy-Graham,
Petrosino, & Gandhi, 2008).
Understanding how research findings might influence the thoughts, behaviors, and
actions of policymakers and their constituents is relevant. Research findings can be
influential at the individual/person level, the interpersonal level, and/or at a collec-
tive level (Henry & Mark, 2003). At the individual level, research information may
cause a change in the thoughts or actions of a single policy maker. At this level,
decision makers are likely to imply a “truth test” and a “utility test” to assess
research findings; these are considered their frames of reference (Weiss & Bucuvalas,
1980). Decision makers will assess the rigor and reliability of the research findings
and determine if they are trustworthy. Decision makers will also assess findings to
determine if they suggest feasible change or provide guidance for policy direction.
These frames of reference reinforce the value and necessity of high-quality research
68 L.L. Knoche
to inform policy. At the interpersonal level, change results from interaction among
individuals. Data might be shared to influence the behavior of others. For example,
one school board member might share research information with another member
to affect a discussion on effective school practices. Finally, at the collective level,
change in action can also occur within organizations—this is the level at which
policy change is likely to occur. A school district, for example, might institute the
application of a district-wide drop-out prevention program, or at the federal level,
policy-makers might make programs to support parent engagement a funding prior-
ity. The levels at which behavior change and action may take place are interdepen-
dent (i.e., the conversation between the school board members could in turn
contribute to a change in the district’s policies). Thus, affecting attitudes and behav-
iors of individual policy makers is a likely first step for researchers interested in
promoting change in the collective.
Collaboration between family–school partnership scholars and policy makers
could provide an instructive backdrop for further informing the process by which
research evidence is utilized in policy-making contexts. The topic of knowledge
utilization has been studied previously (Weiss & Bucuvalas, 1980), but evolution in
political ideologies, funding priorities and limitations, and the active involvement of
intermediary agencies could certainly affect how information is used for policy
making in today’s political landscape. Compelling and informative research ques-
tions might include the following: What data form is most effective at directly
informing policy? What type of research information is likely to yield change in the
thoughts, behaviors, and/or actions of individual policymakers? What levels of
engagement are needed at the individual level before policy change can occur at the
collective level? How can researchers support engagement? While these questions
are not exclusive to family–school partnership research, the content area could
provide an opportunity to explore these meaningful associations via interdisciplin-
ary partnerships.
A research agenda to inform and influence policy may or may not directly align
with research emanating from the scientific community. Indeed, compelling, policy-
guiding research can result from an iterative process involving both policy makers
and practitioners who work alongside researchers to generate a research agenda.
The framework of agenda setting is most effective when it is bi-directional—that is
the process reflects both “research to policy” and “policy to research.” A uni-
directional approach where research is “pushed” out to the policy community with-
out their input or desires will not likely be effective at promoting change.
It is customary, and often comfortable, for researchers to serve as “suppliers” of
information to the policy and practice communities. In such an approach, research-
ers are working in a uni-directional capacity. The expectation is that the information
being disseminated will be utilized and ultimately influence decisions. Alternatively,
4 Policy Agenda 69
that will strengthen the likelihood of information being influential in a policy arena.
These include the need to develop relevant questions, establish strong evidence of
effectiveness, replicate findings, and incorporate varied research designs.
The quality and type of research questions addressed in studies are among the first
considerations when research is intended to influence policy. A clear statement
about the importance of the work and relevancy to policy is needed from the begin-
ning. Often, the discussion and application of research is relegated to a small section
in discussion sections of manuscripts. If one of the intended purposes of the research
is to inform policy, its potential utility in this regard should be initially specified
(Guerra, Graham, & Tolan, 2011).
Furthermore, researchers are advised to frame research questions around issues
that policies can actually address and consider targets that can be regulated through
policy change (Huston, 2005). Often, the questions of primary interest to research-
ers highlight associations between variables. For example, how does the parent–
child relationship evolve over a developmental spectrum? Or, what teacher variables
predict effective family–school partnerships? These questions, while highly rele-
vant for advancing the understanding of family–school partnerships, are not easily
regulated through policy change. Researchers often talk about conditions that pro-
mote outcomes; policy makers want to know how to create these conditions.
Examples of the types of questions relevant to policy makers include how is an
effective parent–child relationship established? Or, how can effective teachers be
identified? Policy making is action-oriented and is about change—how would a
change in behavior lead to improvement? What would be gained? Associations
alone do not provide this type of action-focused information (Huston, 2005).
In research, the focus is often on “what we do not know” and is less likely target-
ing “what we should do” (Shonkoff, 2000). Researchers tend to be more concerned
with issues of internal validity, whereas external validity is of utmost priority for the
practice and policy communities. To meet the needs of policy makers, research must
attend to organizational or setting level changes that promote or relate to outcomes
as opposed to a typical focus on individual differences (Huston, 2005). Fortunately,
analytic techniques have been identified and are available to address these contex-
tual variables (see Beretvas, Volume II of this series).
initiatives. There are ways, however, that these studies would be enhanced to provide
even greater information for policy decisions. First, economic information is a pri-
ority element for policy makers. Thus, a focus by researchers on cost, as well as
effectiveness, is needed (Guerra et al., 2011; Huston, 2005, 2008). An intervention
program might have excellent results, but the cost for implementing the program
could prohibit uptake of the program. Policy makers are balancing multiple priori-
ties and cost is essential information in determining decisions. In the same spirit of
efficiency, policy makers are interested in thresholds of services associated with
desired outcomes. Thresholds concern identifying how much of a given service or
program is enough to promote adequate, or expected, change. A focus on thresholds
and cost effectiveness helps policy makers identify efficiencies in programming and
maximize funding opportunities (Huston, 2005).
Second, attention to issues of fidelity is critically important for understanding
program effects. Researchers are often concerned about Type 1 error or detecting an
effect when it is not present. Avoiding Type 2 error, however, is also of primary
concern (Huston, 2005). Measures of implementation fidelity can provide assur-
ance that an intervention was in fact implemented and that a lack of change in tar-
geted outcomes was a result of program ineffectiveness, and not due to lack of
implementation. The field of implementation science provides guidance on effec-
tive methods for measuring indicators of implementation fidelity (Halle, Metz, &
Martinez-Beck, 2013).
Third, research that identifies and assesses the effectiveness of evidence-based
strategies, principles, or practices for use across varying contexts is needed (Guerra
et al., 2011). Context is a primary focus of family–school partnership research.
Contextual variables describe and differentiate children, families, and schools and
help illustrate and understand associations with partnership outcomes. Likewise,
context is critical for child and family policy. Complete context descriptions must
be collected, analyzed, and reported to improve the chance that evidence-based pro-
grams can be appropriately scaled up from efficacy trials. One of the common con-
cerns about evidence-based curricula or programs is that they can be difficult to
implement beyond the bounds of the settings in which they were developed and
initially tested. Policy makers are interested in supports via programs that can be
implemented in their full constituent body. The primary questions that confront
family–school partnership researchers revolve around “what works for whom under
what conditions” (Tseng, 2012). Researchers must consider mediators and modera-
tors of impact (Guerra et al., 2011; Shonkoff, 2000). Information on how a program
operates in the field is of critical importance to decision makers (Dodge, 2011).
Fourth, researchers are advised to consider uptake rates in randomized controlled
trials. This is an extension of the recommendation to include context descriptors in
all research. The uptake rate is the percentage of participants who agree to engage
in a study out of those who are invited to participate. Understanding characteristics
and selection factors among children, families, and teachers who opt out of partici-
pation in the trial, as well as those who opt in, is relevant and important. Results
from research studies are based on subjects who elect to participate in intervention
activities; reported findings do not generally consider subjects who opt out of
72 L.L. Knoche
Replicating Findings
Evidence resulting from single studies is helpful for informing and advancing the
knowledge base of family–school partnership research. Each study independently
contributes to a repository of information on effective programs, practices, and
behaviors to support positive outcomes. It is this accumulation of consistent evi-
dence that is essential for use in policy decisions (Huston, 2008). Of even greater
value is information provided from a multidisciplinary perspective. For example,
data from economists and psychologists that illustrate intervention effectiveness in
terms of student outcomes, as well as economic impact, across multiple trials, would
prove very useful for policy makers. In many instances, findings from studies are
inconsistent and may be contradictory. Such a scenario provides an opportunity for
researchers to dig deeper into the research and attempt to understand the processes
that may be contributing to the variation. However, from a policy making perspec-
tive, this scenario is challenging. Inconsistent and contradictory findings violate the
desire for a well-specified action plan.
Meta-analysis is a useful methodology for compiling available data on constructs
of interest and dealing with conflicting findings. However, for meta-analyses to be
of benefit to decision makers, researchers must provide sufficient data on the con-
text in which studies are taking place. Details on location, age of participants, racial/
ethnic background, language, education, and other demographic characteristics are
essential. Presenting associations and impacts of intervention without adequate con-
textual detail is of minimal benefit (Weiss et al., 2008).
4 Policy Agenda 73
There is a growing consensus in the research community that the use of mixed
method designs—including both qualitative and quantitative methodologies—is
valuable and important. Different methodologies generate distinct types of data to
inform similar research questions. Randomized experimental designs yield infor-
mation that indicates if a particular program, policy, or intervention resulted in
change. Results indicate the amount of change that might be expected, and can even
specify the conditions under which these changes occurred. However, such a design
in isolation can rarely describe how that change occurred (Guerra et al., 2011;
McCall & Green, 2004; Tseng, 2012). Qualitative data can illustrate and describe
how participants experienced the intervention. The combination of methodologies
provides useful information on intervention effects and process. Additionally,
designs that gather effectiveness information from a community sample are also
needed (Dodge, 2011). Family–school partnership scholars are encouraged to look
towards creative, alternative evaluation designs to assess program effectiveness that
consider population-level effects (e.g., use of administrative data).
To influence and inform policy, researchers need to connect with policy makers and
affiliated key stakeholders, including intermediary organizations, and build rela-
tionships that are essential for information exchange. As a first step, researchers
must determine the most appropriate targets or outlets for research findings. Once
the information targets are identified and partnerships formed, the connections can
provide an opportunity to educate policy makers. Furthermore, this identification
allows researchers to tailor dissemination efforts and products to improve under-
standing and reinforce significance of research findings.
Information Targets
pathway towards enhancing dissemination efforts and improving the likelihood for
influence is through targeted relationships with intermediary agencies and offices
that contribute to policy efforts. There are several relevant national organizations
that are connected to policy work and also have an interest in family–school partner-
ship issues. A short list of these organizations include the (a) National Community
Education Association, (b) National Education Association, (c) National Coalition
for Parent Involvement in Education, (d) National Parent Teacher Association, (e)
National Center for Children in Poverty (NCCP), and (f) Center for Law and Social
Policy (CLASP). This is only a selected few organizations with an interest in this
topic. Researchers are advised to give consideration to intermediary groups and
individuals at all levels—local, state, and national.
Ensuring that research findings are available to these organizations will enhance
the opportunity for research findings to make their way to policy discussions. For
example, a guide for policy makers regarding parent engagement from preschool to
grade three was recently released by NCCP (Smith, Robbins, Stagman, & Mahur,
2013). This summary document was developed with policy makers as a target audi-
ence and includes research findings on parent engagement, promising models to
support engagement, examples of policy efforts in several states, and specific policy
recommendations related to parent engagement. The authors at NCCP synthesized
relevant research to produce this easy-to-read, appealing guide that includes direct
calls for policy action. Another example that highlights the work of an intermediary
in prompting policy change is that of CLASP, an organization that considers policy
solutions for low-income individuals. A recently released report includes a synthe-
sis of relevant research related to parent engagement and also provides targeted
policy recommendations and resources for state representatives (Johnson-Staub,
2013). These are two examples of how intermediary organizations have become
involved in translating scientific findings into practical and palatable products and
solutions for a varied group of key stakeholders. Targeted relationships with inter-
mediaries, therefore, are essential and can prompt and support dialogue between the
research, practice, and policy communities.
Dissemination of Products
Conclusions
References
Dearing, J., & Kreuter, M. (2010). Designing for diffusion: How can we increase uptake of cancer
communication innovations? Patient Education and Counseling, 81, S100–S110. doi:10.1016/j.
pec.2010.10.013.
Dodge, K. A. (2011). Context matters in child and family policy. Child Development, 82, 433–442.
doi:10.1111/j.1467-8624.2010.01565.x.
Granger, R. (2005). Connecting research and practice. Annual Report Essay, W.T. Grant
Foundation.
Guerra, N. G., Graham, S., & Tolan, P. H. (2011). Raising healthy children: Translating child
development research into practice. Child Development, 82, 7–16. doi:10.1111/j.1467-8624.
2010.01537.x.
Halle, T. G., Metz, A. J., & Martinez-Beck, I. (Eds.). (2013). Applying implementation science in
early childhood programs and systems. Baltimore, MD: Paul H. Brookes.
Henry, G. T., & Mark, M. M. (2003). Beyond use: Understanding evaluation’s influence on
attitudes and actions. American Journal of Evaluation, 24, 293–314. doi:10.1177/
109821400302400302.
Huston, A. C. (2005). Connecting the science of child development in public policy. SRCD Social
Policy Report, 19(4), 3–18.
Huston, A. C. (2008). From research to policy and back. Child Development, 79, 1–12.
doi:10.1111/j.1467-8624.2007.01107.x.
4 Policy Agenda 77
Johnson-Staub, C. (2013). Charting progress for babies in child care project: Promote family
engagement. Retrieved from http://www.clasp.org/admin/site/babies/make_the_case/files/
Promote-Family-Engagement.pdf
McCall, R. B., & Green, B. L. (2004). Beyond the methodological gold standards of behavioral
research: Considerations for practice and policy. SRCD Social Policy Report, 18(2), 3–19.
Nutley, S., Walter, I., & Davies, H. T. O. (2007). Using evidence: How research can inform public
services. Bristol, UK: Policy Press.
Phillips, D. A. (2005). Commentary. In A. C. Huston (Ed.), Connecting the science of child devel-
opment in public policy. SRCD Social Policy Report, 19(4), 3–18.
RAND Corporation. (2011). Standards for high quality research and practice. Santa Monica, CA:
Author.
Sheridan, S. M., Bovaird, J. A., Glover, T. A., Garbacz, S., Witte, A., & Kwon, K. (2012). A ran-
domized trial examining the effects of conjoint behavioral consultation and the mediating role
of the parent-teacher relationship. School Psychology Review, 41, 23–46.
Shonkoff, J. P. (2000). Science, policy, and practice: Three cultures in search of a shared mission.
Child Development, 82, 181–187. doi:10.1111/1467-8624.00132.
Shonkoff, J. P., & Bales, S. (2011). Science does not speak for itself: Translating child develop-
ment research for the public and its policymakers. Child Development, 82, 17–32.
doi:10.1111/j.1467-8624.2010.01538.x.
Smith, K. B., & Larimer, C. W. (2009). The public policy theory primer. Boulder, CO: Westview
Press.
Smith, S., Robbins, T., Stagman, S., & Mahur, D. (2013). Parent engagement from preschool
through grade 3: A guide for policymakers. New York, NY: National Center for Children in
Poverty.
Tseng, V. (2012). The uses of research in policy and practice. SRCD Social Policy Report, 26(2),
3–16.
Weiss, C. H., & Bucuvalas, M. J. (1980). Truth tests and utility tests: Decision-makers’ frames of
reference for social science research. American Sociological Review, 45, 302–313.
Weiss, C. H., Murphy-Graham, E., & Birkeland, S. (2005). An alternate route to policy influence:
How evaluations affect D.A.R.E. American Journal of Evaluation, 26, 12–29.
doi:10.1177/1098214004273337.
Weiss, C. H., Murphy-Graham, E., Petrosino, A., & Gandhi, A. G. (2008). The Fairy Godmother–
and her warts. American Journal of Evaluation, 29, 29–47. doi:10.1177/1098214007313742.
Chapter 5
Testing Statistical Moderation in Research
on Home–School Partnerships: Establishing
the Boundary Conditions
Demographic Characteristics
were 29, 46, and 32 % in Journal of School Psychology, Elementary School Journal,
and Journal of Applied Developmental Psychology, respectively. Among the 36
studies, close to one third of them (31 %) have involved multilevel data and ana-
lyzed the data with multilevel models. The sample size of these studies ranged from
62 to 33,311. However, only 10 (28 %) out of the 36 studies have reported the actual
estimation method and the statistical software used for the analyses. Most impor-
tantly, none of these 36 studies used any latent variables techniques (e.g., the use of
structural equation models; Brown, 2006; Kline, 2011) or took into account the
issue of potential measurement error in their analyses. In other words, all the
(observed) variables in the analyses were assumed to have perfect reliability (or no
measurement error). The issue of measurement error and the advantage of account-
ing for measurement error in the analyses are discussed below with a real data
demonstration.
In our example, we test a potential moderation effect of hyperactivity (as the mod-
erator) on the relation between teacher-rated home–school relationship (as the
focal variable related to parent academic involvement) and peer acceptance (as the
outcome variable). Peer acceptance was selected as the outcome variable based on
its noted importance to students’ academic engagement and achievement (Buhs,
Ladd, & Herald, 2006). Hyperactivity was selected as the moderator variable
based on the finding that hyperactive behaviors place a child at risk for low peer
acceptance (Ronk, Hund, & Landau, 2011). Thus, a finding that a positive home–
school relationship buffers children with hyperactive behaviors from low peer
acceptance would suggest the importance of this type of parent involvement for
improving social and academic adjustment of students with elevated levels of
hyperactive behaviors.
Our demonstration is based on a portion of data from Project Achieve (Hughes
& Kwok, 2007) and contained 409 (54 % male) fourth grade children attending
one of three school districts (one urban, two small cities) in southeast and central
Texas. The ethnic composition of these 409 students was 37 % White, 21 %
Black, 38 % Hispanic, and 4 % Other. Children’s mean age was 6.56 years
(SD = .35), and 59 % of participants were identified by school records as eligible
for free or reduced lunch. In 45.4 % of households, at least one parent had com-
pleted high school. Children’s mean Broad Reading and Broad Math age standard
scores on the Woodcock Johnson III (Woodcock, McGrew, & Mather, 2001), or
the comparable Spanish language test of achievement, the Batería-III (Muñoz-
Sandoval, Woodcock, McGrew, & Mather, 2005) were 97.92 (SD = 16.25) for
Reading and 101.50 (SD = 12.56) for Math. These 409 students were enrolled in
159 classrooms.
5 Testing Statistical Moderation 83
Multilevel data are very common in educational studies; for instance, students are
often nested within higher level clusters such as classrooms and schools. Given that
students from the same classroom share the same environment (e.g., with the same
home-room teacher), they are more likely to respond or react in a manner similar to
classmates than to children from other classrooms. Hence, the responses of students
from the same cluster are likely to not be independent from each other, which is a
very important assumption (i.e., independent observations) for traditional statistical
84
Table 5.1 Descriptive statistics and zero-order correlations of the variables used in the demonstration
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
Correlations
1. PA 1.000
2. HSR .142 1.000
3. HA −.373 −.207 1.000
4. HSR × HA .225 .148 −.346 1.000
5. X1 .066 .893 −.128 .086 1.000
6. X2 .107 .877 −.130 .069 .803 1.000
7. X3 .151 .874 −.249 .159 .724 .684 1.000
8. X4 .070 .847 −.146 .040 .726 .764 .679 1.000
9. X5 .165 .872 −.218 .165 .716 .708 .854 .681 1.000
10. X6 .105 .840 −.227 .171 .670 .710 .688 .797 .686 1.000
11. X7 .068 .830 −.157 .116 .765 .722 .678 .665 .635 .677 1.000
12. X8 .204 .739 −.139 .173 .641 .579 .580 .494 .593 .511 .515 1.000
Descriptive statistics
Mean 3.288 .004 .079 −.163 .026 −.184 .034 −.056 .030 .028 .001 .033
SD .708 .777 1.017 .856 .933 .885 .997 .788 .975 .863 .866 1.055
Underscored correlations were significant at p < .05
Note: PA peer acceptance, HSR home–school relationship, HA hyperactivity, X1–X8 Items for HSR (the description of individual item is presented in
Appendix 1), SD standard deviation
O.-M. Kwok et al.
5 Testing Statistical Moderation 85
a
X γ1 e
γ2
M Outcome
γ3
X×M
b
δ1 X1
δ2 X2
δ3 X3 Φ11
δ4 X4
X
δ5 X5
δ6 X6
δ7 X7
Φ12 Φ13
γ1
δ8 X8
Φ22 e
γ2
M Outcome
δ9 X1× M
γ3
δ10 X2× M
Φ23
δ12 X4× M
X×M
δ13 X5× M
δ14 X6× M
δ15 X7× M
δ16 X8× M
Fig. 5.1 (a) Moderation model with composite scores. Note. X (Focal variable): Home–school
relationship (HSR); M (Moderator): Hyperactivity as the moderator; Outcome: Peer acceptance.
(b) Moderation model with latent factors. Note. X: Home–school relationship (HSR); X1–X8:
c
δ1 X1
δ2 X2
δ3 X3
Φ11
δ4 X4
X
δ5 X5
γ1 e
δ6 X6
γ3
δ7 X7 Outcome
Φ12
δ8 X8
Φ22 γ2
eX X X
γ1
e
γ2
M Outcome
γ3
eX×M X×M X× M
Fig. 5.1 (continued) the eight items of the home–school relationship scale; M: Hyperactivity as
the moderator; X × M: the interaction effect between home–school relationship and hyperactivity;
(X1 × M) to (X8 × M): the eight product indicators between the eight items of HSR scale and the
hyperactivity; Outcome: Peer acceptance. (c) Moderation model with the use of the latent moder-
ated structural equations (LMS) approach. Note. X: Home–school relationship (HSR); X1–X8: the
eight items of the home–school relationship scale; M: Hyperactivity as the moderator; Outcome:
Peer acceptance. The interaction effect between X and M is represented by a filled circle. (d)
Moderation model with the reliability-adjusted composite score. Note. X: Teacher-rated home–
school relationship; M: Hyperactivity as the moderator; Outcome: Peer acceptance.
Variance of eX : V ( eX ) = V ( X ) ´ (1 - r XX ) where V(X) is the variance of Teacher-rated home–
school relationship composite score and ρXX is the reliability of Teacher-rated home–school rela-
tionship (based on the eight-item teacher-rated home–school relationship scale);
( )
Variance of eX´M : V ( eX´M ) = V ( X ´ M ) ´ 1 - r( X´M )( X´M ) where V(X × M) is the variance of the
interaction effect term and ρ(X × M)(X × M) is the reliability of the interaction effect term based on Aiken
and West’s (1991) equation 8.12 (p. 144)
5 Testing Statistical Moderation 87
methods such as the ordinary least squares (OLS) regression. Without adequately
taking this dependency issue into account, the standard errors of the parameter esti-
mates can be underestimated, which in turn, can lead to inflated type I error rates
and incorrect statistical conclusions.
Given the multilevel structure of our data (with students nested within class-
rooms), multilevel models (MLMs; Hox, 2010; Raudenbush & Bryk, 2002) are
needed to analyze this type of data due to the potential non-independent observa-
tions. To examine whether our data were completely independent, we first fit the
random intercept model (i.e., a model without any predictors; Raudenbush & Bryk,
2002) to obtain the intra-class correlation (ICC) of the outcome variable, peer
acceptance. ICC can be viewed as the average correlation between a pair of observa-
tions (students) within a cluster (classroom). The ICC of peer acceptance was .32
which was substantial and further supported the need of using multilevel models for
analyzing our data.
We then fit our hypothesized model as shown in Fig. 5.1a with the traditional
multilevel models which assume that all the observed variables are perfectly mea-
sured. We analyzed our data using the MIXED routine in SPSS (V.22; SPSS Inc.,
2013). The corresponding annotated SPSS MIXED syntax for this model is
presented in Appendix 2. The restricted maximum likelihood (REML) estimation
method was used given that it has been the default estimation method in most of the
commonly used multilevel modeling-related programs (e.g., SPSS MIXED, HLM,
STATA xtmixed). The parameter estimates are presented in Table 5.2. As shown in
Table 5.2, the target interaction effect between teacher-rated home–school relation-
ship (HSR) and hyperactivity was marginally significant (p = .053) while the two
main effects were significant (p < .05).
One of the potential causes of the marginal significant effect is the lack of control
of the plausible measurement error in the variables. As Aiken and West (1991)
pointed out, moderation or interaction effects generally carry relatively low statisti-
cal power, and the occurrence of measurement error can further introduce potential
bias in the parameter estimates and the corresponding standard errors. There are
many potential sources of measurement error in educational studies, including the
physical and mental condition of the participants or test takers such as fatigue, the
testing situation such as lighting and noisiness, and the instruments such as the
wording of items and equipment-related issues. The occurrence of measurement
error can sometimes lead to underestimated effects (a.k.a. attenuation due to mea-
surement error). One of the major advantages of using structural equation modeling
(SEM) over the traditional approaches such as OLS regression is that SEM can flex-
ibly analyze models with both observed and unobserved/latent variables (or con-
structs) simultaneously while taking into account the potential measurement error
by isolating the corresponding variance out from the model (as the measurement
part of the model), thus allowing researchers to directly model the relations among
the error-free latent variables (as the structural part of the model) and reduce bias in
parameter estimates (Cham, West, Ma, & Aiken, 2012). Following, we will reana-
lyze the hypothesized model using the structural equation model. Details of how to
set up the model and conduct the analysis are described below.
88
One of the commonly used approaches for analyzing models with latent interaction
effects is the product-indicator approach (Marsh, Wen, & Hau, 2004; Marsh, Wen,
Hau, & Nagengast, 2013), which includes several variations: the constrained
approach (Algina & Moulder, 2001; Jöreskog & Yang, 1996; Moulder & Algina,
2002), the partially constrained approach (also named as the generalized appended
product-indicator approach; Wall & Amemiya, 2001), and the unconstrained
approach (Marsh et al., 2004; Marsh, Wen, & Hau, 2006). As stated in the name
(product-indicator approach), all these approaches require the creation of the latent
interaction factor(s) by creating the corresponding interaction/product terms from
the observed variables. In our real data example, hyperactivity (M) is a single stan-
dardized score and the teacher-rated home–school relationship is an eight-item
scale. As shown in Fig. 5.1b, the eight indicators of the latent interaction factor
(X × M) are the product terms between the standardized hyperactivity score (M) and
each of the eight items (X1–X8) of the home–school relationship scale.
There are some guidelines on how to create the latent interaction factors when
involving two latent factors. For example, suppose that we have two latent factors
(F1 and F2) and each latent factor has three items uniquely loaded on it (X1, X2,
and X3 exclusively on F1; and X4, X5, and X6 exclusively on F2). As Marsh et al.
(2004) pointed out, there is no need to create all possible product terms (in our
example, the maximum number of product terms between the indicators of the two
latent factors is nine) given that some of the product terms carry overlapping or
redundant information. They propose the use of the matched pair product term by
pairing up the indicators with similar (standardized) factor loadings. Following our
previous example with two latent factors (F1 and F2), suppose that the standardized
loadings for the three indicators in F1 are .8 (X1), .6 (X2), and .4 (X3) and the stan-
dardized loadings for the three indicators in F2 are .5 (X4), .9 (X5), and .3 (X6). The
total number of matched pairs of indicators needed for creating the latent interaction
factor (F1 × F2) is three: (X1 × X5), (X2 × X4), and (X3 × X6). When the latent fac-
tors carry uneven numbers of indicators (e.g., F1 has six indicators while F2 has
four indicators), we only need to create four distinctive matched pairs given that
pairs with reused indicators will only contain redundant information. In addition to
the magnitude of the factor loadings, we can also create the matched pairs based on
the actual description/meaning of the indicators. Again, instead of creating all pos-
sible pairs, the matched pair approach creates the pairs containing unique informa-
tion which eases the estimation while yields the optimal results in terms of standard
errors and statistical power.
As pointed out previously, there are three different product-indicator approaches,
and the commonly used one is the unconstrained approach proposed by Marsh et al.
(2004). Compared with the other approaches which require many nonlinear con-
straints for the interaction latent factor, the unconstrained approach is the one with
the simplest specification which only requires imposing a single constraint (i.e.,
constrain the mean of the interaction latent factor equal to the covariance between
90 O.-M. Kwok et al.
the two main latent factors). Nevertheless, this constraint will not be necessary if the
double-mean-centering strategy (Lin, Wen, Marsh, & Lin, 2010) is used. That is, all
the observed indicators are mean-centered1 before creating the product terms, and
the product terms are then mean-centered before fitting the model with the latent
interaction factor.
We adopted the double-mean-centering strategy to create all the product indica-
tors and fit the model as shown in Fig. 5.1b. As shown in Fig. 5.1b, the indicators of
the interaction latent factor (X × M) are from the product terms between the eight
items (i.e., X1–X8) of the home–school relationship scale and the standardized
hyperactivity score (i.e., M). Given that we used the unconstrained approach with
doubly mean-centered variables, nonlinear constraint was not necessary and we did
not impose any nonlinear constraints in the model. We have used the Type = Complex
routine in Mplus (V7.11; Muthén & Muthén, 1998–2012) which is the model-based
approach to analyze multilevel data (Wu & Kwok, 2012). Robust maximum likeli-
hood estimation method (i.e., MLR as the default estimation method under
Type = Complex) was used. MLR produces “maximum likelihood parameter esti-
mates with standard errors and a chi-square test statistic that are robust to non-
normality and non-independence of observations” (Muthén & Muthén, 1998–2012,
p. 603). The corresponding Mplus syntax for this model is presented in Appendix 2.
We first fit the model as shown in Fig. 5.2 but the model did not fit the data well
based on the overall model chi-square test and other commonly used fit indices2
(χ2(133) = 463.215, p < .001; RMSEA= .078; CFI = .869), which were off from the
generally recommended cutoff values of fit indices (i.e., RMSEA≤ .05 and CFI ≥ .95
indicate good fit; Kline, 2011). Based on the modification indices, we then added
three correlated residuals: between X3 (parent has shared goals with school) and X5
(similar expectations of child) of the home–school relationship (HSR) scale,
between X4 (parent respects teacher) and X6 (teacher respects parent) of the HSR
scale, and between the product indicators of (X3 × M) and (X5 × M). Although the
overall model chi-square test of the modified model was still significant
(χ2(130) = 260.462, p < .001), the model fit indices did indicate that this modified
model fit the data adequately (RMSEA= .050; CFI = .948).
1
Mean-centering of a variable is to subtract each observed value by the corresponding mean of that
variable. For example, for Xi, the observed value of the X variable for the i-th person, the mean-
centered value for this i-th person is:
Xi* ( the mean centered Xi ) = Xi - X ,
Low Hyperactivity
0.5
Average Hyperactivity
Peer Acceptance (in Standard Deviation (SD) unit)
0.4
High Hyperactivity
0.3
0.2
The difference The difference
between the between the
0.1
high and low high and low
hyperactivity hyperactivity
0
groups is close groups is less
to one SD of than one-half
-0.1
the peer SD of the peer
acceptance acceptance
-0.2
-0.3
-0.4
-0.5
Lower-level Higher-level
Home-School Relationship
Fig. 5.2 Decomposing the moderation effect of hyperactivity on the relation between home–
school relationship and peer acceptance. (a) Single-group approach. (b) Multiple-group approach
For the measurement part of the model (i.e., the latent factors of X and X × M),
all the observed indicators and the product terms were significantly loaded on the
corresponding latent factors, with standardized loadings ranged from .672 to .901
for the home–school relationship factor (i.e., the X latent factor in Fig. 5.1b) and
from .729 to .900 for the interaction latent factor (i.e., the X × M latent factor in
Fig. 5.1b). For the structural part of the model, both hyperactivity (γ2 = −.219,
p < .001) and the interaction effect factor (γ3 = .092, p = .039) were statistically sig-
nificant in predicting peer acceptance whereas home–school relationship was not
significant (γ1 = .051, p = .196). Specifically, compared with the results in model 1a
(γ3 = .068, p = .053), the target interaction effect, hyperactivity by home–school rela-
tionship, is now statistically significant (γ3 = .092, p < .05).
To further understand the interaction effect, we have adopted the Aiken and
West’s (1991) approach to decompose the interaction effect by substituting some
meaningful values of the moderator to the final model as shown below (also see
Table 5.2 model 1b):
which has been mean-centered and completely standardized in (5.1), we can substitute
three commonly used values for hyperactivity, including: mean of hyperactivity (i.e.,
0; also labeled as the average hyperactivity group in Fig. 5.2), one standard deviation
(SD) above the mean of hyperactivity (i.e., 0 (mean) + 1 (SD) = 1; also labeled as the
high hyperactivity group in Fig. 5.2), and one standard deviation below mean of
hyperactivity (i.e., 0 (mean)−1 (SD) = −1; also labeled as the low hyperactivity group
in Fig. 5.2). Here are the three equations for the three hyperactivity groups:
% of explained variance
Y model without interaction factor - Y model with interaction factor
= ´ 100%
Y model with interaction factor
.458 - .452
= ´ 100% = 1.3%
.458
5 Testing Statistical Moderation 93
where Ψ is the residual variance of the target outcome variable (i.e., peer acceptance
in our example). This change in the residual variance is similar to the R-square
change in multiple regression models, which can be viewed as the proportion of the
variance explained in the outcome variable (i.e., peer acceptance) solely by adding
the latent interaction factor. Accordingly, 1.3 % of the variance in peer acceptance
has been explained by the latent interaction factor between hyperactivity and home–
school relationship. Based on Cohen’s (1988) guideline, this (1.3 %) is a small
effect size which is quite common (especially for interaction effects) in the social
sciences literature.
Another commonly used approach for analyzing latent interaction effects is the
distribution-analytic approaches (Kelava et al., 2011; Marsh et al., 2013; Marsh,
Wen, Nagengast, & Hau, 2012), which include: Latent Moderated Structural
Equations (LMS; Klein & Moosbrugger, 2000), Quasi-Maximum Likelihood
(QML; Klein & Muthén, 2007), and other similar approaches such as marginal
maximum likelihood (Cudeck, Harring, & du Toit, 2009). Compared with the
product-indicator approaches, the distribution-analytic approaches directly estimate
the latent interaction effects by taking the non-normality of the interaction effects
into account without creating any product indicators. Both LMS and QML are more
commonly discussed than the other distribution-analytic approaches (Kelava et al.,
2011; Marsh et al., 2012; Marsh et al., 2013). LMS uses a full information maxi-
mum likelihood-based approach with the expectation-maximization (EM) algo-
rithm, whereas QML uses a quasi-log-likelihood function with two-stage
maximization (i.e., single-step iteration method followed by the Newton–Raphson
algorithm).
There are some major differences between LMS and QML. First, LMS has a
very restrictive distributional assumption (i.e., all the observed and latent predictor
variables and the corresponding measurement error/residual variances are normally
distributed) whereas QML has a less restrictive distributional assumption and more
robust against the violations of the normality assumptions. Second, LMS can
become computationally intensive when multiple (three or more) interaction effects
are simultaneously estimated in the model, and QML is generally computationally
feasible (i.e., the number of latent interaction effects is less of a concern with QML).
A third difference is that LMS is built-in to Mplus (V7.11) and QML is a stand-
alone program available from Andreas Klein (klein@psych.uni-frankfurt.de). Given
that only LMS is available in Mplus, we analyzed the hypothesized latent interac-
tion effect (as presented in Fig. 5.1c) with the use of LMS. The corresponding anno-
tated Mplus syntax for this model is presented in Appendix 2. For more information
regarding QML, readers can consult Klein and Muthén (2007) which contained the
technical details, or Kelava et al. (2011).
94 O.-M. Kwok et al.
Based on the previous 1b model, we have analyzed the latent interaction effect
model using the LMS approach by including the same two correlated residuals
(i.e., X3 [parent has shared goals with school] and X5 [similar expectations of
child] of the home–school relationship (HSR) scale, X4 [parent respects teacher]
and X6 [teacher respects parent] of the HSR scale). Given the multilevel nature
of our data, we used the “Type = Complex Random” routine along with
“Algorithm = Integration” in which the “Type = Complex” part could address the
multilevel data while the “Type = Random” and “Algorithm = Integration” parts
were to initiate the LMS procedure for analyzing the latent interaction effect. The
robust estimation method, MLR, was the default estimation method for the LMS
approach in Mplus.
The results are presented in Table 5.2. Mplus does not produce any overall model
fit chi-square test or related fit statistics when using the “Type = Random” routine.
Additionally, neither standardized solutions nor modification indices are available
under the “Type = Random” routine. As shown in Table 5.2, model 1c had the exact
same pattern of significances and very similar (unstandardized) parameter estimates
as those from model 1b. That is, both hyperactivity and the interaction effect
between hyperactivity and home–school relationship (HSR) were significant in pre-
dicting peer acceptance whereas HSR was not significant. The conclusion of the
findings was the same as model 1b.
In summary, both the unconstrained approach (one of the product-indicator
approaches as used in model 1b) and the LMS approach (one of the distribution-
analytic approaches as used in model 1c) have their own advantages and disadvan-
tages for analyzing latent interaction effects. For example, the unconstrained
approach can be easily implemented in most of the SEM programs (e.g., AMOS,
EQS, LISREL, Mplus, and Stata’s SEM routine) while LMS can only be estimated
in Mplus. Similarly, the unconstrained approach is in general computationally fea-
sible and can produce overall model chi-square test and other fit indices, as well as
the standardized solutions and modification indices whereas the LMS can be com-
putationally intensive without producing the same set of model fit information as
the unconstrained approach. On the other hand, the LMS approach does not require
the creation of the product indicators which can become an issue when there is a
large difference in the number of observed indicators among the latent factors used
for creating the latent interaction effects (Wu, Wen, Marsh, & Hau, 2013). Moreover,
as shown in previous simulation studies (Cham et al., 2012; Wu et al., 2013), when
the normality assumption is met, the LMS approach produces more accurate and
efficient parameter estimates and standard errors which can lead to higher statistical
power than the unconstrained approach. Whereas, the unconstrained approach is
relatively more robust against the non-normal conditions and still yields unbiased
latent interaction effect estimates. Again, each one of these approaches offers differ-
ent advantages (and disadvantages) on estimating the latent interaction effects and
readers may select one (or both) of these approaches depending on their needs and
the availability of the statistical software.
5 Testing Statistical Moderation 95
Although we have shown the advantages of analyzing interaction effects with the
use of the latent factor model, sometimes it may not be feasible to include all the
observed and latent variables in the same model simultaneously. Moreover, struc-
tural equation models are generally estimated using the maximum likelihood or
related estimation methods which require relatively large sample sizes to produce
unbiased and efficient parameter estimates. Hence, trying to estimate a complex
model which contains many observed and latent variables with a relatively small
sample sizes can lead to potential convergence difficulties. To avoid the conver-
gence issue, one may modify the model by creating composite scores (e.g., sum-
ming or averaging the items of a latent construct) to reduce the number of variables
and parameters in the model. In our demonstration, this will involve converting
model 1b (with original observed items and latent factors) back to model 1a
(observed composite scores only). Nevertheless, we have already noted that the
target interaction effect, home–school relationship by hyperactivity, was statisti-
cally significant (p = .039) under model 1b whereas marginally significant under
model 1a (p = .053).
Instead of directly using the observed composite scores, there is an alternative
approach to incorporate measurement errors into the composites-only model
(Brown, 2006; Kline, 2011). As shown in Fig. 5.1d, each of the two latent factors,
home–school relationship and the interaction effect, contains only one indicator,
respectively. This model is not identifiable and some constraints have to be imposed
to make it identifiable. First, the factor loadings are both fixed to be 1.0 for the two
latent factors. For home–school relationship, the variance of the corresponding
residual (eX) is constrained based on the following equation:
V ( eX ) = V ( X ) ´ (1 – r XX ) , (5.2)
where V(X) is the variance of the home–school relationship composite score (.604)
and ρXX is the reliability of the home–school relationship scale (.941) based on the
eight-item HSR scale. As shown in Table 5.1, the standard deviation of HSR is .777
and the variance of this variable is (.777)2 = .604. Given these two pieces of informa-
tion, we can then obtain the residual variance of home–school relationship:
(
V ( eX´M ) = V ( X ´ M ) ´ 1 - r( X´M )( X´M ) , ) (5.3)
96 O.-M. Kwok et al.
where V(X × M) is the variance of the interaction effect and ρ(X × M)(X × M) is the reli-
ability of the interaction effect, which can be further calculated based on Aiken and
West’s (1991) equation3 8.12 (p. 144):
(g XM ) + rXX ´ rMM
2
r( X´ M )( X´ M ) = (5.4)
(g XM ) + 1
2
where γXM is the zero-order correlation between home–school relationship (X) and
hyperactivity (M), ρXX is the reliability of home–school relationship, and ρMM is the
reliability of hyperactivity. Given that hyperactivity is a single standardized score,
we assume it has perfect reliability (i.e., ρMM = 1.00) to ease the calculation. The
zero-order correlation between home–school relationship and hyperactivity is −.207.
With this information, we can calculate the reliability of the interaction effect:
r( X´ M )( X´ M ) = = = .943
(g XM ) + 1 ( -.207 ) + 1
2 2
Given that the variance of the interaction effect is .732, we can then calculate the
residual variance of the interaction effect using (5.3):
( )
V ( eX´M ) = V ( X ´ M ) ´ 1 - r( X´M )( X´M ) = .732 ´ (1 - .943 ) = .042
Thus, we can constrain the two residual variances (of HSR and the interaction
effect) to .036 and .042, respectively. The corresponding annotated Mplus syntax
for this model is presented in Appendix 2.
As presented in Table 5.2, this is a saturated model which fits the data perfectly.
The parameter estimates and the corresponding tests of significance of model 1d are
very similar to the ones from model 1b (estimated with the latent interaction factor).
Specifically, the target interaction effect is still significant (p = .035). Hence, based
on these results, we can again reach the same conclusion as the findings for model
1b. That is, hyperactivity has a substantial negative effect on peer acceptance while
this negative effect can be reduced by more positive home–school relationships.
Regression coefficients can be biased (attenuated) due to measurement error in
predictors and the underestimated effects may lead to low statistical power and incor-
rect statistical conclusions (Aiken & West, 1991). Given that interaction effects in
general have relatively low statistical power, the occurrence of the measurement
error in predictors which are used for creating the interaction effects introduces more
measurement error in the interaction effects which may lead to more substantial
attenuation of the interaction effects and further lower the statistical power for detect-
ing the effects. Hence, when analyzing interaction effects, taking the measurement
error into account can reduce the bias in the parameter estimates, which in turn, may
sometimes help to increase the statistical power for detecting the interaction effects.
3
The variables are assumed to be mean-centered.
5 Testing Statistical Moderation 97
To ease our demonstration example, we only used the students with complete data.
In reality, researchers may have missingness in their data. The traditional approaches,
such as listwise deletion which only includes cases without missing data in the
analysis or mean imputation which replaces missing values with the arithmetic
mean of the variable based on the available observations, generally result in sub-
stantial reduction of statistical power and produce biased estimates of the effects
(Enders, 2010). There are two promising modern approaches to handle data that are
either missing completely at random (i.e., MCAR; the missingness does not depend
on the complete data that can potentially be observed) or missing at random (i.e.,
MCAR; the missingness does not depend on any unobserved data): full information
maximum likelihood (FIML) and multiple imputation (MI) (cf. Enders, 2010; Little
& Rubin, 2002; Schafer & Graham, 2002). FIML is the default method in Mplus to
handle missing data. However, these two approaches require the multivariate nor-
mality assumption and interaction effect is likely not normally distributed even
though both the predictor and the moderator are bivariate normally distributed.
Enders, Baraldi, and Cham (2014) showed that, while both FIML and MI performed
well when the interaction effect was close to normally distributed, with moderate to
severe non-normality, FIML and MI could result in substantial bias on the interac-
tion effect estimate. Nevertheless, these two modern approaches still outperformed
the listwise deletion method. Future research is needed for better procedures to
handle missing data with interaction effects. Readers are encouraged to consult with
Enders et al. (2014) for discussions on how to analyze interaction effects with the
two modern approaches when missing data are present, and Von Hippel (2009) for
guidance on multiple imputation with nonlinear effects.
In our example, we tested a simple interaction effect with two continuous predic-
tors. We can also test interaction effects with the product of continuous and non-
continuous or categorical variables (e.g., gender, intervention conditions). Below
we discuss a few other issues researchers may face when they examine interaction
effects.
a Single-group approach
X e
F-M Outcome
X × F-M
b Multiple-group approach
e e
_ _
X Outcome X Outcome
Fig. 5.3 Different approaches for testing interaction effect with noncontinuous/categorical vari-
able. Note. X: Home–school relationship; F–M: The dichotomized gender variable (i.e., F = Female
group; M = Male group); Outcome: Peer acceptance
variables and the ways to test and interpret the interaction effects created between
the coded categorical variables and the continuous variables. Nevertheless, some
researchers may be unfamiliar with how to treat continuous variables and decide to
categorize at least one of the predictors (e.g., dichotomizing a variable with the use
of median split). As shown previously (MacCallum, Zhang, Preacher, & Rucker,
2002; Maxwell & Delaney, 1993), under some circumstances, splitting (or dichoto-
mizing) a continuous variable into distinct groups can produce misleading results
such as spurious effects with inflated Type I error rate. Hence, researchers should
avoid this practice and keep their continuous variables as they are in the analysis.
Concluding Remarks
In this chapter, we offer some alternative approaches for testing moderation effects
with an example of hyperactivity moderating the effect of home–school relationship
on peer acceptance. In our demonstration, we showed that the latent interaction
approaches (models 1b and 1c) and the reliability-adjusted approach (model 1d)
resulted in significant interaction effects, whereas the traditional approach (model
1a) resulted in a marginally significant effect. We have discussed the potential
advantage of taking the measurement error into account in testing interaction effects
which can in general reduce the bias in the parameter estimates. Additionally, we
have discussed issues related to testing interaction effects including: centering vari-
ables, probing (graphing) interaction effects, obtaining effect sizes of the interaction
effects, handling missing data, and testing more complex interaction effects (e.g.,
100 O.-M. Kwok et al.
interaction effects with categorical variables). We hope that this chapter can shed
some light on testing moderation effects in family–school partnership research.
Note:
Mixed—SPSS Mixed routine
Mixed (Outcome) with (Predictors)
PA: Peer Acceptance; Hyper: Hyperactivity; HSR: home–school relationship;
H_H: the interaction effect term (i.e., product between HSR and Hyper)
/Fixed = (Predictors): estimate the regression coefficients
/Random = Intercept: estimate the level-2 variance
|Subject (ClassID): indicate the higher level cluster variable (i.e., class-
room ID in our example)
/Method = REML: estimation method (REML: Restricted Maximum
Likelihood as the default estimation method)
/Print = Solution: print out the parameter estimates
Testcov: test the random effect variance
5 Testing Statistical Moderation 101
Note:
Title: Title of the syntax
Data: File is (the file name containing the data)
Variable: Names are (the variable names in the data file)
Usevariables are (the variables used in the analysis/model)
Cluster = (cluster ID variable)
Analysis: Type = Complex (Mplus routine which takes the dependency issue
into account by adjusting the standard errors of the estimates)
Model: (specifying the model as shown in Fig. 5.1b)
HSR (the latent factor of HSR) BY the corresponding 8 observed
indicators (i.e., X1–X8)
H_H (the latent interaction factor) BY the corresponding 8
observed product indicators (i.e., HX1–HX8)
PA (the observed Peer Acceptance variable) ON (predicted by)
Hyper, HSR, and H_H
X3 WITH X5 (correlated the residuals between observed items
X3 and X5)
Output: Stdyx (request for standardized solutions)
102 O.-M. Kwok et al.
Note:
Title: Title of the syntax
Data: File is (the file name containing the data)
Variable: Names are (the variable names in the data file)
Usevariables are (the variables in the analysis/model)
Cluster = (cluster ID variable)
Analysis: Type = Complex Random (Complex is for taking the multilevel
structure/dependency into account while Random is required for
the use of the LMS routine)
Algorithm = integration (This command line is required for the use
of the LMS routine)
Model: (specifying the model as shown in Fig. 5.1c)
HSR (the latent factor of HSR) BY the corresponding 8 observed
indicators (i.e., X1–X8)
PA (the observed Peer Acceptance variable) ON (predicted by)
Hyper and HSR
(continued)
5 Testing Statistical Moderation 103
Note:
Title: Title of the syntax
Data: File is (the file name containing the data)
Variable: Names are (the variable names in the data file)
Usevariables are (the variables in the analysis/model)
Cluster = (cluster ID variable)
Analysis: Type = Complex (Complex is for taking the multilevel structure/
dependency into account)
(continued)
104 O.-M. Kwok et al.
References
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions.
Newbury Park, CA: Sage.
Algina, J., & Moulder, B. C. (2001). A note on estimating the Jöreskog–Yang model for latent vari-
able interaction using LISREL 8.3. Structural Equation Modeling, 8, 40–52.
Brown, T. A. (2006). Confirmatory factor analysis for applied research. New York, NY: Guilford
Press.
Buhs, E. S., Ladd, G. W., & Herald, S. L. (2006). Peer exclusion and victimization: Processes that
mediate the relation between peer group rejection and children’s classroom engagement and
achievement? Journal of Educational Psychology, 98, 1–13. doi:10.1037/0022-0663.98.1.1.
Cham, H., West, S. G., Ma, Y., & Aiken, L. S. (2012). Estimating latent variable interactions with
nonnormal observed data: A comparison of four approaches. Multivariate Behavioral Research,
47, 840–876. doi:10.1080/00273171.2012.732901.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ:
Erlbaum.
Cudeck, R., Harring, J. R., & du Toit, S. H. C. (2009). Marginal maximum likelihood estimation
of a latent variable model with interaction. Journal of Educational and Behavioral Statistics,
34, 131–144. doi:10.3102/1076998607313593.
Eccles, J. S., & Harold, R. D. (1996). Family involvement in children’s and adolescents’ schooling.
In A. Booth & J. F. Dunn (Eds.), Family-school links: How do they affect educational out-
comes? (pp. 3–34). Mahwah, NJ: Erlbaum.
Enders, C. K. (2010). Applied missing data analysis. New York, NY: Guilford Press.
Enders, C. K., Baraldi, A. N., & Cham, H. (2014). Estimating interaction effects with incomplete
predictor variables. Psychological Methods. 19(1):39-55
Enders, C. K., & Tofighi, D. (2007). Centering predictor variables in cross-sectional multilevel
models: A new look at an old issue. Psychological Methods, 12, 121–138. doi:10.1037/
1082-989X.12.2.121.
Epstein, J. L., & Sanders, M. G. (2002). Family, school, and community partnerships. In M. H.
Bornstein (Ed.), Handbook of parenting (2nd ed., Vol. 5, pp. 407–437). Mahwah, NJ: Lawrence
Erlbaum.
5 Testing Statistical Moderation 105
Fan, X., & Chen, M. (2001). Parental involvement and students' academic achievement:
A meta-analysis. Educational Psychology Review, 13, 1–22. doi:10.1023/A:1009048817385.
Hamre, B. K., & Pianta, R. C. (2001). Early teacher-child relationships and the trajectory of chil-
dren’s school outcomes through eighth grade. Child Development, 72, 625–638.
doi:10.1111/1467-8624.00301.
Hill, N. E. (2001). Parenting and academic socialization as they relate to school readiness: The
roles of ethnicity and family income. Journal of Educational Psychology, 93, 686–697.
doi:10.1037/0022-0663.93.4.686.
Hill, N E., Castellino, D. R., Lansford, J E., Nowlin, P., Dodge, K. A., Bates, J. E., & Pettit, G. S.
(2004). Parent academic involvement as related to school behavior, achievement, and aspira-
tions: Demographic variations across adolescence. Child Development, 75, 1491–1509.
doi:10.1111/j.1467-8624.2004.00753.x.
Hill, C. R., & Hughes, J. N. (2007). An examination of the convergent and discriminant validity of
the Strengths and Difficulties Questionnaire. School Psychology Quarterly, 22, 380–406.
doi:10.1037/1045-3830.22.3.380.
Hill, N. E., & Tyson, D. F. (2009). Parental involvement in middle school: A meta-analytic assess-
ment of the strategies that promote achievement. Developmental Psychology, 45, 740–763.
doi:10.1037/a0015362.
Hox, J. J. (2010). Multilevel analysis: Techniques and applications (2nd ed.). New York, NY:
Routledge.
Hsu, H-Y., Kwok, O., Acosta, S., & Lin, J-H. (2015). Detecting misspecified multilevel SEMs
using common fit indices: A Monte Carlo study. Multivariate Behavioral Research, 50,
197–215.
Hughes, J. N., & Kwok, O. (2007). The influence of student–teacher and parent–teacher relation-
ships on lower achieving readers’ engagement and achievement in the primary grades. Journal
of Educational Psychology, 99, 39–51. doi:10.1037/0022-0663.99.1.39.
Jöreskog, K. G., & Yang, F. (1996). Nonlinear structural equation models: The Kenny–Judd model
with interaction effects. In G. A. Marcoulides & R. E. Schumacker (Eds.), Advanced structural
equation modeling: Issues and techniques (pp. 57–88). Mahwah, NJ: Erlbaum.
Kelava, A., Werner, C. S., Schermelleh-Engel, K., Moosbrugger, H., Zapf, D., Ma, Y., … & West,
S. G. (2011). Advanced nonlinear latent variable modeling: Distribution analytic LMS and
QML estimators of interaction and quadratic effects. Structural Equation Modeling, 18, 465–
491. doi:10.1080/10705511.2011.582408.
Klein, A. G., & Moosbrugger, H. (2000). Maximum likelihood estimation of latent interaction
effects with the LMS method. Psychometrika, 65, 457–474. doi:10.1007/BF02296338.
Klein, A. G., & Muthén, B. O. (2007). Quasi-maximum likelihood estimation of structural equa-
tion models with multiple interaction and quadratic effects. Multivariate Behavioral Research,
42, 647–673. doi:10.1080/00273170701710205.
Kline, R. B. (2011). Principles and practice of structural equation modeling (3rd ed.). New York,
NY: Guilford Press.
Kreft, I. G. G., de Leeuw, J., & Aiken, L. S. (1995). The effect of different forms of centering in
hierarchical linear models. Multivariate Behavioral Research, 30, 1–21. doi:10.1207/
s15327906mbr3001_1.
Lin, G. C., Wen, Z., Marsh, H. W., & Lin, H. S. (2010). Structural equation models of latent inter-
actions: Clarification of orthogonalizing and double-mean-centering strategies. Structural
Equation Modeling, 17, 374–391. doi:10.1080/10705511.2010.488999.
Little, R. J. A., & Rubin, D. B. (2002). Statistical analysis with missing data (2nd ed.). Hoboken,
NJ: Wiley.
MacCallum, R. C., Zhang, S., Preacher, K. J., & Rucker, D. D. (2002). On the practice of dichoto-
mization of quantitative variables. Psychological Methods, 7, 19–40.
doi:10.1037/1082-989X.7.1.19.
106 O.-M. Kwok et al.
Marsh, H. W., Wen, Z., Hau, K. T., & Nagengast, B. (2013). Structural equation models of latent
interaction and quadratic effects. In G. R. Hancock & R. O. Muller (Eds.), Structural equation
modeling: A second course (2nd ed., pp. 267–308). Greenwich, CT: Information Age.
Marsh, H. W., Wen, Z., Nagengast, B., & Hau, K.-T. (2012). Structural equation models of latent
interaction. In R. H. Hoyle (Ed.), Handbook of structural equation modeling (pp. 436–458).
New York, NY: Guilford Press.
Marsh, H. W., Wen, Z., & Hau, K.-T. (2004). Structural equation models of latent interactions:
Evaluation of alternative estimation strategies and indicator construction. Psychological
Methods, 9, 275–300. doi:10.1037/1082-989X.9.3.275.
Marsh, H. W., Wen, Z., & Hau, K. T. (2006). Structural equation models of latent interaction and
quadratic effects. In G. R. Hancock & R. O. Mueller (Eds.), Structural equation modeling:
A second course (pp. 225–265). Greenwich, CT: Information Age.
Masten, A. S., Morrison, P., & Pelligrini, D. S. (1985). A revised class play method of peer assess-
ment. Developmental Psychology, 21, 523–533. doi:10.1037/0012-1649.21.3.523.
Maxwell, S. E., & Delaney, H. D. (1993). Bivariate median splits and spurious statistical signifi-
cance. Psychological Bulletin, 113, 181–190. doi:10.1037/0033-2909.113.1.181.
Meehan, B. T., Hughes, J. N., & Cavell, T. A. (2003). Teacher-student relationships as compensa-
tory resources for aggressive children. Child Development, 74, 1145–1157. doi:10.1111/
1467-8624.00598.
Moulder, B. C., & Algina, J. (2002). Comparison of methods for estimating and testing latent vari-
able interactions. Structural Equation Modeling, 9, 1–19. doi:10.1207/S15328007SEM0901_1.
Muñoz-Sandoval, A. F., Woodcock, R. W., McGrew, K. S., & Mather, K. (2005). Bateria III
Woodcock–Muñoz. Itasca, IL: Riverside.
Muthén, L. K., & Muthén, B. O. (1998–2012). Mplus user’s guide (6th ed.). Los Angeles, CA:
Muthén & Muthén.
Raudenbush, S., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis
methods (2nd ed.). Thousand Oaks, CA: Sage.
Ronk, M. J., Hund, A. M., & Landau, S. (2011). Assessment of social competence of boys with
attention-deficit/Hyperactivity disorder: Problematic peer entry, host responses, and evalua-
tions. Journal of Abnormal Child Psychology, 39, 829–840. doi:10.1007/s10802-011-9497-3.
Ryu, E., & West, S. G. (2009). Level-specific evaluation of model fit in multilevel structural equa-
tion modeling. Structural Equation Modeling, 16, 583–601. doi:10.1080/10705510903203466.
Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the state of the art. Psychological
Methods, 7, 147–177. doi:10.1037/1082-989X.7.2.147.
Serpell, Z. N., & Mashburn, A. J. (2012). Family–school connectedness and children’s early social
development. Social Development, 21, 21–46. doi:10.1111/j.1467-9507.2011.00623.x.
Shieh, G. (2011). Clarifying the role of mean centering in multicollinearity of interaction effects.
British Journal of Mathematical and Statistical Psychology, 64, 462–477.
SPSS Inc. (2013). IBM SPSS statistics 22 core system user’s guide. Chicago, IL: Author.
Tang, S., Dearing, E., & Weiss, H. B. (2012). Spanish-speaking Mexican-American families’
involvement in school-based activities and their children's literacy: The implications of having
teachers who speak Spanish and English. Early Childhood Research Quarterly, 27, 177–187.
doi:10.1016/j.ecresq.2011.09.001.
Von Hippel, P. T. (2009). How to impute interactions, squares, and other transformed variables.
Sociological Methodology, 39, 265–291. doi:10.1111/j.1467-9531.2009.01215.x.
Wall, M. M., & Amemiya, Y. (2001). Generalized appended product indicator procedure for non-
linear structural equation analysis. Journal of Educational and Behavioral Statistics, 26, 1–29.
doi:10.3102/10769986026001001.
West, S. G., Aiken, L. S., & Krull, J. L. (1996). Experimental personality designs: Analyzing cat-
egorical by continuous variable interactions. Journal of Personality, 64, 1–48. doi:10.1111/
j.1467-6494.1996.tb00813.x.
5 Testing Statistical Moderation 107
Wong, S. W., & Hughes, J. N. (2006). Ethnicity and language contributions to dimensions of par-
ent involvement. School Psychology Review, 35, 645–662.
Woodcock, R. W., McGrew, K. S., & Mather, N. (2001). Woodcock-Johnson III tests of achieve-
ment. Riverside, CA: Riverside.
Wu, J., & Kwok, O. (2012). Using SEM to analyze complex survey data: A comparison between
design-based single-level and model-based multilevel approaches. Structural Equation
Modeling, 19, 16–35. doi:10.1080/10705511.2012.634703.
Wu, Y., Wen, Z., Marsh, H. W., & Hau, K.-T. (2013). A comparison of strategies for forming prod-
uct indicators for unequal numbers of items in structural equation models of latent interactions.
Structural Equation Modeling, 20, 551–567. doi:10.1080/10705511.2013.824772.
Zhang, Y., Haddad, E., Torres, B., & Chen, C. (2011). The reciprocal relationships among parents’
expectations, adolescents’ expectations, and adolescents’ achievement: A two-wave longitudi-
nal analysis of the NELS data. Journal of Youth and Adolescence, 40, 479–489. doi:10.1007/
s10964-010-9568-8.
Chapter 6
Contexts of Family–School Partnerships:
A Synthesis
Eva N. Patrikakou
If the quest for a common definition is viewed as intangible or even futile given the
multidimensional nature of home–school partnerships, the need for using an inte-
grative theoretical framework is tangible and valuable. Having an integrative frame-
work within which home–school partnerships can be explored, implemented, and
evaluated is fundamental to alleviate some of the definitional and methodological
issues that have afflicted this field of study since its inception.
Bronfenbrenner’s broader bioecological framework has been proposed and used
over the years as it encompasses the multidimensional nature of home–school part-
nerships, and also puts the spotlight on the uniqueness of interactions between two
or more settings contributing to development (Bronfenbrenner, 1979, 1994).
Although calls for the need and critical function of integrating such a theoretical
infrastructure have been raised both in empirical and theoretical work in the past
two decades, close attention to this aspect has not been paid (e.g., Patrikakou, 1996;
6 Contexts of Family–School Partnerships: A Synthesis 111
Patrikakou, Weissberg, Redding, & Walberg, 2005; Sheridan, Cowan, & Meegan,
1999; Sheridan, Knoche, Edwards, Bovaird, & Kupzyk, 2010).
The bioecological framework of development acknowledges both individual and
contextual characteristics, which is a necessary assumption when studying the
effects of any factor in human development. This theory also posits that individual
characteristics affect and are affected by systems, institutions, and programs. In its
original form, there were four concentric systems included (i.e., microsystem,
mesosystem, exosystem, and macrosystem)1 one of which, the mesosystem, has
been exclusively dedicated to the interaction of immediate environments in an indi-
vidual’s life, such as family and school (Bronfenbrenner, 1979). In later reformula-
tions of his theory, Bronfenbrenner (1986, 1994) added a fifth system (chronosystem)
which extended the concept of context into the dimension of time, not in terms of
chronological age and developmental changes in a person, which are nonetheless
inherent in the study of development, but rather in terms of the changes that occur
in various environments in which a person lives and grows (see Fig. 6.1). This key
addition makes the bioecological theory even more pertinent as a framework to
discuss home–school partnerships, especially in light of rapid and massive contex-
tual changes that have occurred, such as technology and media advances that have
influenced individuals and all surrounding systems.
1
Microsystem: proximal influences from immediate settings such as home and school; mesosys-
tem: interactions between two or more microsystems; exosystem: settings that may not contain the
individual, but which influence the individual’s microsystems; and macrosystem: broader societal,
cultural, and ideological influences.
112 E.N. Patrikakou
research, programming, and policy which is not always triggered by the researchers’
agenda, but by policy needs, whether to inform a decision, enhance the policy-
makers’ understanding on a given topic, or directly target opportunities to actively
influence public policy. Such a view significantly expands the sometimes narrow
consideration of family–school partnerships as only a mesosystemic area of study,
to highlighting its impact on, and interaction with, broader systems of influence.
The author’s argument that scholars in the area of home–school partnerships can
engage in the process of policy-making through a variety of ways further highlights
the multidimensionality of the field, not only in terms of inquiry, but also in terms
of practical applications and broad dissemination.
Another aspect of the multidimensional nature of home–school partnerships is
showcased by Kwok, Im, Hughes, Wehrly, and West (2015). The authors discuss
methodological issues illuminating the bioecological theory’s premise of bidirec-
tionality of influences in that individual and microsystemic family and school char-
acteristics not only are affected by other systems, institutions, and programs, but
they affect those spheres of influence. The issues discussed in this chapter are also
pertinent to the study of issues encompassed in the chronosystem by illuminating
contextual changes or consistency of conditions and characteristics in one’s imme-
diate and broader environment. The example of exceptionality characteristics inter-
acting with environmental properties and processes, including family–school
relations, also points to how microsystems, and their mesosystemic interactions are
vital in, and are affected by, development and individual characteristics. As
Bronfenbrenner and Ceci (1994) noted “which features of the environment become,
or are made, salient plays a critical role in determining which of a multitude of
innate possibilities have the most chance of finding realization” (p. 583).
Multidimensionality
parents can play in their child’s development and the need to support families in
maximizing the benefits of parent–child interactions. Stormshak et al. (2015) shed
light on the importance of school leadership and targeted training from pre-service,
and in-service, teachers and other school personnel on home–school partnership
issues. The impact of school leadership, perceptions of leadership on professionals’
attitudes toward interventions, and broader policy initiatives are immense. This has
become especially evident for policy initiatives, such as Response of Intervention
(RTI), also known as Multi-Tiered System of Supports (MTSS), that have been
proposed to serve as a structural backbone of home–school interventions. School
professionals, who have confidence in their leadership, including a positive, knowl-
edgeable principal and other informed leaders, display more favorable attitudes
about RTI and its intended benefits (Feiker-Hollenbeck & Patrikakou, 2014). Such
perceived importance of school-based leaders contributes to the critical role school
climate and fidelity of implementation play in the implementation of home–school
interventions.
Sheldon (2015) expands the discussion of roles in the function that districts can
play in facilitating schools working with families, and in this way, moving beyond
the mere monitoring of partnership activities in which schools are engaged.
Expanding the discussion on this topic, Knoche (2015) further extends the examina-
tion of various roles to the broader level of policy-makers and researchers, and
emphasizes the importance of forging ways to enhance each other’s work and create
(a) a more solid basis for policies involving families and schools, as well as (b) more
targeted research to enhance our understanding of contextual perspectives in home–
school partnerships.
As Stormshak et al. (2015) suggest to enhance the chances of home–school pro-
gramming to be implemented with fidelity—an issue that several authors discuss as
a recurring problem for partnership interventions—it is best if it is couched within
existing, naturally occurring ecological configurations, or established multi-tiered
service-delivery structures, such as RTI. Response to Intervention is a multi-tiered
assessment and intervention-delivery model designed to improve educational out-
comes via research-based instructional methods aligned with student data. Although
the earliest conceptualization of RTI involved four tiers of instruction, with the forth
tier being special education, RTI is now typically considered a general education
reform framework involving assessment and intervention that occurs prior to spe-
cial education referral (Feiker-Hollenbeck & Patrikakou, 2014). Teachers’ account-
ability for student progress has been key to the framework, while school leadership
and collaboration with families are also fundamental aspects of this model.
Stormshak et al.’s (2015) recommendation that implementation of evidence-based
practices within existing service-delivery settings and mandated initiatives already
in place can minimize impediments inherent in scale-up implementations is an
important one.
Inherent in the premise of utilizing existing service-delivery models is the assump-
tion that home–school partnerships are an integral part of broader interventions.
Indeed, as Sheldon (2015) denotes, robust school–family–community partnerships
6 Contexts of Family–School Partnerships: A Synthesis 117
are essential in fostering school improvement. Family involvement has long been
part of federal policy (e.g., the Elementary and Secondary Education Act, Individuals
with Disabilities Education Act) and recently calls for its integration into broader
educational reform have been intensified (e.g., The Southwest Educational
Development Laboratory, 2013).
Contextual Perspectives
methods to parse out the not-so-obvious nuances is essential to inform further program
design efforts and provide a more comprehensive description of the contextual
factors under investigation.
Moving Forward
The chapters in this volume have detailed the importance of various contextual and
methodological aspects in the study of home–school partnerships. The call made by
authors to clearly design, implement, and evaluate partnership interventions is espe-
cially challenging, as it requires to address concretely the multidimensional nature
of home–school relations. Papers in this volume offer distinct and tangible ways to
advance research in this area. Specifically, chapter authors draw attention to (a) criti-
cal issues involved in the fidelity of implementation of intervention programs;
(b) the analysis and interpretation of results, methodologically addressing the multi-
dimensionality of home–school partnerships; (c) the training of the next generation of
teachers to effectively integrate home–school partnerships in their practice; and
(d) reaching out to policy-makers and policy-enforcers to garner support for develop-
ing and maintaining strong home–school partnership programs. Most importantly, the
work presented in this volume highlights the fact that a viable and valuable research
agenda in the area of home–school partnerships can be methodologically robust and
successfully address the complexities inherent in this line of inquiry.
As the variables, processes, and effects (direct, indirect, and moderated) associ-
ated with family involvement and home–school partnerships have become more
clear over the past three decades, a major new factor has rapidly taken traction and
wedged itself among all these factors, and will take the field swiftly into a new era.
Easy access to technology and the Internet is changing the way that families con-
duct their daily lives, connect with schools and the community, and enhance their
children’s learning opportunities. This rapidly changing nature of relationships will
affect the field of home–school partnerships in profound ways. Given the socio-
historical impact of technology and media infusion on daily life, such a profound
contextual change serves as a good example of the bioecological framework’s
importance of chronosystemic influences.
The prospect of exploring the impact of new forms of connectedness both among
family members, as well as between home and school is an exciting prospect. Some
of the critical issues raised in the chapters of this volume as impeding implementa-
tion of effective programs may be mediated and potentially moderated by the use of
technology (e.g., recruitment, attrition, mobility, or to a certain extent, fidelity);
other issues may become pronounced and would need to be further addressed (e.g.,
lack or resources or decreased family time); and yet others will be significantly
enhanced (e.g., the development of the learner’s personal competencies).
The basic principles underlining good research approaches or effective program-
ming that were discussed in the chapters of this volume will not shift in the same
way that responsive, caring parenting and the responsive, caring education have not
6 Contexts of Family–School Partnerships: A Synthesis 119
References
American Academy of Pediatrics Committee on School Health. (2004). School-based health ser-
vices. Pediatrics, 113, 1839–1845.
Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and
design. Cambridge, MA: Harvard University.
Bronfenbrenner, U. (1986). Ecology of the family as a context for human development: Research
perspectives. Developmental Psychology, 22, 723–742.
Bronfenbrenner, U. (1994). Ecological models of human development. In T. Husen & T. N.
Postlewaite (Eds.), International encyclopedia of education (Vol. 3, pp. 1643–1647). Oxford,
United Kingdom: Elsevier Sciences.
Bronfenbrenner, U., & Ceci, S. J. (1994). Nature-nurture reconceptualized in developmental per-
spective: A bioecological model. Psychological Review, 101, 568–586.
Elias, M. J., Patrikakou, E. N., & Weissberg, R. P. (2007). A competence-based framework for
parent-school-community partnerships in secondary schools. School Psychology International,
28, 540–554.
Feiker-Hollenbeck, A., & Patrikakou, E. N. (2014). Response to Intervention in Illinois: An explo-
ration of school professionals’ attitudes and beliefs. Mid-Western Educational Researcher, 26,
58–82.
Hoekstra, A., & Korthagen, F. (2011). Teacher learning in a context of educational change:
Informal learning versus systematically supported learning. Journal of Teacher Education, 62,
76–92.
Knoche, L. L. (2015). Research issues to forward a policy agenda in support of family–school
partnerships. In S. M. Sheridan & E. M. Kim (Eds.), Family-school partnerships in context.
New York, NY: Springer.
Kwok, O.-M., Im, M., Hughes, J. N., Wehrly, S. E., & West, S. G. (2015). Testing statistical
moderation in research on home-school partnerships: Establishing the boundary conditions.
In S. M. Sheridan & E. M. Kim (Eds.), Family-school partnerships in context. New York, NY:
Springer.
Patrikakou, E. N. (1996). Investigating the academic achievement of adolescents with learning
disabilities: A structural modeling approach. Journal of Educational Psychology, 88, 435–450.
Patrikakou, E. N. (2004). Adolescence: Are parents relevant to students’ high school achievement
and post-secondary attainment? The Harvard Family Research Project. Family Involvement
Network of Educators: Research Digests. From http://gseweb.harvard.edu/hfrp/projects/fine/
resources/digest/adolescence.html
120 E.N. Patrikakou
Patrikakou, E. N. (2015). Relationships among parents, students, and teachers: The technology
wild card. Procedia Journal of Social and Behavioral Science, 174, 2253–2258.
Patrikakou, E. N., Weissberg, R. P., Redding, S., & Walberg, H. J. (Eds.). (2005). School-family
partnerships: Fostering children’s school success. New York, NY: Teachers College Press.
Redding, S. (2014). Personal competency: A framework for building students’ capacity to learn.
Center on Innovations in Learning, Temple University. Retrieved from http://www.centeril.
org/publications/Personal_Compentency_Framework.pdf
Sheldon, S. B. (2015). Moving beyond monitoring: A district leadership approach to school, fam-
ily, and community partnerships. In S. M. Sheridan & E. M. Kim (Eds.), Family-school part-
nerships in context. New York, NY: Springer.
Sheridan, S., Cowan, R., & Meegan, T. (1999). The times, they are a changing: A review of raising
children in a socially toxic environment. School Psychology Quarterly, 14, 428–432.
Sheridan, S. M., Holmes, S. R., Smith, T. E., & Moen, A. L. (2015). Complexities in field-based
partnership research: Exemplars, challenges, and an agenda for the field. In S. M. Sheridan &
E. M. Kim (Eds.), Family-school partnerships in context. New York, NY: Springer.
Sheridan, S., Knoche, L., Edwards, C., Bovaird, J., & Kupzyk, K. (2010). Parent engagement and
school readiness: Effects of the getting ready intervention on preschool children’s social-
emotional competencies. Early Education and Development, 21, 125–156.
Southwest Educational Development Laboratory. (2013). Partners education in a dual capacity-
building framework for family–school partnerships. Retrieved March, 2014, from http://
www2.ed.gov/documents/family-community/partners-education.pdf
Spillane, J. P., Reiser, B. J., & Reimer, T. (2002). Policy implementation and cognition: Reframing
and refocusing implementation research. Review of Educational Research, 72, 387–431.
Stein, B., Sontag-Padilla, L., Chan Osilla, K., Woodbridge, M., Kase, C., Jaycox, L., … & Golan,
S. (2012). Interventions to improve student mental health. Retrieved September, 2014, from
http://www.rand.org/content/dam/rand/pubs/technical_reports/2012/RAND_TR1319.pdf
Stormshak, E. A., Brown, K. L., Moore, K. J., Dishion, T., Seeley, J., & Smolkowski, K. (2015).
Going to scale with family-centered, school-based interventions: Challenges and future direc-
tions. In S. M. Sheridan & E. M. Kim (Eds.), Family-school partnerships in context. New York,
NY : Springer.
Twyman, J. S. (2014). Competency-based education: Supporting personalized learning. Connect:
Making Learning Personal. Center on Innovations in Learning, Temple University. Retrieved
from http://www.centeril.org/connect/resources/Connect_CB_Education_Twyman-2014_11.12.
pdf
Index
A Centering variables, 99
Administration, 35 CLASP. See Center for Law and Social Policy
An Action Team for Partnerships (ATP), (CLASP)
47, 51, 55, 56, 59 Community engagement and family
The Arizona State Parent Information and data-driven educational approach, 61
Resource Center (PIRC), 56 district approach, 60
goal-oriented plans, 48
B A Match on Dry Grass, 67
Bioecological model responsive implementation, 48
Bronfenbrenner’s broader bioecological school–family partnerships, 112
framework, 110 schools’ approach, 46–47
children’s learning and behavior, 112 Conjoint behavioral consultation (CBC)
chronosystem, 111 educational problem-solving, 8
empirical and theoretical work, 110–111 empirical support, 8–10
family–school partnerships, 110, 111 families and schools, 69
home–school–community partnership family–school partnership, 7
programs, 112 and Getting Ready, 10, 112
home–school partnerships, 110 operationalized stages, 7
human development, 112 stages, 7–8
individual and contextual characteristics, 111 Conjoint behavioral consultation stages, 8
Local Education Agencies, 112 Contextual perspectives, 113, 116–18
macrosystemic and exosystemic policy
factors, 112–113
mesosystem, 111 D
mesosystemic interactions, 113 Developmental considerations, 14–15
microsystem of family, 112 Dissemination
multidimensionality, 113 contextual perspectives, 117
policy-making, 113 family support services, 29
Bronfenbrenner’s broader bioecological prevention and intervention programs, 27
framework, 110 products, 75
District leadership and direct facilitation
C strategies, 52
CBC. See Conjoint behavioral consultation
(CBC)
Center for Law and Social Policy E
(CLASP), 74 Early childhood professionals (ECPs), 4–6