Content uploaded by Nancy E. Hill
Author content
All content in this area was uploaded by Nancy E. Hill on Jan 07, 2015
Content may be subject to copyright.
Content uploaded by Nancy E. Hill
Author content
All content in this area was uploaded by Nancy E. Hill on Nov 12, 2014
Content may be subject to copyright.
Parental Involvement in Middle School:
A Meta-Analytic Assessment of the Strategies That Promote Achievement
Nancy E. Hill and Diana F. Tyson
Duke University
Early adolescence is often marked by changes in school context, family relationships, and developmental
processes. In the context of these changes, academic performance often declines, while at the same time
the long-term implications of academic performance increase. In promoting achievement across elemen-
tary and secondary school levels, the significant role of families, family–school relations, and parental
involvement in education has been highlighted. Although there is a growing body of literature focusing
on parental involvement in education during middle school, this research has not been systematically
examined to determine which types of involvement have the strongest relation with achievement. The
authors conducted a meta-analysis on the existing research on parental involvement in middle school to
determine whether and which types of parental involvement are related to achievement. Across 50
studies, parental involvement was positively associated with achievement, with the exception of parental
help with homework. Involvement that reflected academic socialization had the strongest positive
association with achievement. Based on the known characteristics of the developmental stage and tasks
of adolescence, strategies reflecting academic socialization are most consistent with the developmental
stage of early adolescence.
Keywords: academic achievement, family-school relationships, meta-analysis, middle school, parental
involvement in education
Early adolescence and entry into middle school reflect change
on multiple levels. The middle school years coincide with key
changes in adolescent development, including biological and cog-
nitive growth, social development, and renegotiations of family
relationships, especially the parent–adolescent relationship
(Adams & Berzonsky, 2003; Grolnick, Price, Beiswenger, &
Sauck, 2007; Keating, 2004; Lerner & Steinberg, 2004; Smetana,
Campione-Barr, & Daddis, 2004; Steinberg & Silk, 2002). Further,
the middle school context reflects a significant change compared
to elementary school, including a larger, more bureaucratic system
with many more teachers, peers, and curricular choices (Dauber &
Epstein, 1989; Eccles & Harold, 1996; Hill & Chao, 2009). In the
context of such changes and development, adolescents’ aca-
demic performance often declines (Barber & Olsen, 2004;
Eccles, 2004; Gutman & Midgley, 2000), while at the same
time, the long-term implications of achievement for educational
and occupational attainment increase (Eccles & Harold, 1993).
The confluence of these developmental and contextual changes
at early adolescence increases the risk that students may not
reach their potential and heightens the need to identify sources
of support.
In promoting achievement across elementary and secondary
school levels, theories, research, and policies have identified the
significant role of families, family–school relations, and parental
involvement in education (Fan & Chen, 2001; Hill & Chao, 2009;
Seginer, 2006). Indeed, family–school relations and parental in-
volvement in education have been identified as a way to close
demographic gaps in achievement and maximize students’ poten-
tial (Dearing, Kreider, Simpkins, & Weiss, 2006; Hampton, Mum-
ford, & Bond, 1998; Hara, 1998). As such, federal policies like the
No Child Left Behind Act (NCLB; 2002) mandate parental in-
volvement in education and family–school relations across ele-
mentary and secondary school levels. Despite consensus about the
importance of families and schools working together across devel-
opmental stages, extant theories of parental involvement in edu-
cation have been based on elementary school students and elemen-
tary school contexts and do not account for the changes associated
with middle school and early adolescent development (Hill &
Taylor, 2004; Hill, Tyson, & Bromell, 2009). Indeed, some re-
search has demonstrated that the strength of the relation between
parental involvement and achievement declines between elemen-
tary and middle schools (e.g., Singh et al., 1995). Whereas some
aspects of parental involvement in education may decline in amount
or in effectiveness during middle school, like involvement at school
(Singh et al., 1995; Stevenson & Baker, 1987), other aspects of
involvement that are not accounted for in extant frameworks may
increase in significance (Chao, Kanatsu, Stanoff, Padmawidjaja, &
Nancy E. Hill and Diana F. Tyson, Department of Psychology and
Neuroscience, Duke University.
This study was funded in part by a seed grant from the National Institute
on Drug Abuse Transdisciplinary Research Center at the Center for Child
and Family Policy, Duke University. We thank Harris Cooper for his
consultation on meta-analytic strategies. In addition, we thank Lea Bro-
mell, Roxanne Flint, Joseph Bataille, and Andrea Malone-Reid for their
assistance with coding the reports for this meta-analysis.
Correspondence concerning this article should be addressed to Nancy E.
Hill, who is now at the Graduate School of Education, Harvard University,
Larsen 703, 14 Appian Way, Cambridge, MA 02138. E-mail:
hillna@gse.harvard.edu
Developmental Psychology © 2009 American Psychological Association
2009, Vol. 45, No. 3, 740 –763 0012-1649/09/$12.00 DOI: 10.1037/a0015362
740
Aque, 2009). Therefore, it is imperative to identify the extent to
which parental involvement in education is positively related to
achievement for middle school students and which types of in-
volvement are most effective.
In the last two decades, the amount of research on parental
involvement in education, especially for middle school, has in-
creased exponentially, but it has produced often competing find-
ings. For example, some research has demonstrated that parental
involvement in education is positively associated with adolescents’
academic outcomes throughout middle and high school (e.g., Cat-
sambis, 2001; Hill et al., 2004). However, other research found
that parental involvement is not related to achievement (e.g., Balli,
Wedman, & Demo, 1997; Bronstein, Ginsberg, & Herrera, 2005).
This growing but disaggregated body of research has used a
variety of methods, making it difficult to draw firm conclusions for
middle school. Although recent meta-analyses have focused on
parental involvement (e.g., Fan & Chen, 2001; Jeynes, 2003,
2005), these meta-analyses did not consider the developmental
stage of students as a unique factor in their analyses. These studies
collapsed across research from prekindergarten through high school.
Further, whereas Jeynes (2007) focused on secondary schools, this
meta-analysis collapsed across middle and high schools and was
limited to urban contexts. Unlike the high school years, when parents
have gained experience with supporting more autonomous adoles-
cents and larger, more bureaucratic schools, the middle school
years reflect a renegotiation for schools, families, and students.
Therefore, identifying the most effective strategies in middle
school will guide programs and policies so that they can promote
the most effective strategies (Hill et al., 2009). To this end, we
conducted a meta-analysis of the existing research on parental
involvement in middle school and situate our findings within
existing theories and frameworks and within the developmental
context of early adolescence. This meta-analysis addressed two
broad questions. First, what is the strength of the relation between
parental involvement in education and achievement during middle
school? Second, which types of involvement have the strongest
positive relation with achievement?
Parental Involvement in Education: Definitions and
Frameworks
Although there are numerous definitions of parental involve-
ment in education, we define it as “parents’ interactions with
schools and with their children to promote academic success”
(Hill et al., 2004, p. 1491). This is somewhat broader than the
definition articulated in the NCLB, which is “the participation
of parents in regular, two-way, and meaningful communication
involving student academic learning and other school activi-
ties” (No Child Left Behind Act, 2002, §9101). Consistently
included in the extant theories, frameworks, and assessments
are home-based and school-based involvement strategies (e.g.,
Kohl, Lengua, McMahon, & the Conduct Problems Prevention
Research Group, 2000; Seginer, 2006). Further, such a distinc-
tion is useful as it distinguishes policy-relevant realms—home
and school. The most widely cited among existing frameworks
is Epstein’s (1987; Conners & Epstein, 1995; Epstein & Sand-
ers, 2002), which includes school-based involvement strategies
(e.g., volunteering at school, communication between parents
and teachers, and involvement in school governance); home-
based involvement strategies, including engaging in educational
activities at home; school support for parenting (e.g., parent
training programs); and involvement between the school and
community agencies. Second, the framework undergirding
Comer’s (1995) School Development Program has also in-
formed research in this field. Comer’s framework also includes
school-based involvement—such as parent–teacher confer-
ences, volunteering and being present in the school, and par-
ticipation in school governance—and home-based involvement,
such as parental reinforcement of learning at home. Finally,
Grolnick and Slowiaczek (1994) articulated a three-pronged
framework: First, behavioral involvement includes both home-
based and school-based involvement strategies, such as active
connections and communication between home and school,
volunteering at school, and assisting with homework. Second,
cognitive–intellectual involvement reflects home-based in-
volvement and includes parental role in exposing their children
to educationally stimulating activities and experiences. Finally,
personal involvement includes attitudes and expectations about
school and education and conveying the enjoyment of learning,
which reflects parental socialization around the value and utility
of education.
Within an elementary school context, school-based involvement
is associated with children’s achievement, because such involve-
ment is likely to include visits to the classroom and interactions
with children’s teachers. Such interactions and exposure increase
parents’ knowledge about the curriculum, enhance social capital,
and increase the effectiveness of involvement at home (Comer,
1995; Epstein, 2001; Hill & Taylor, 2004). Further, interactions
between parents and teachers may increase mutual respect and
increase teachers’ perceptions about how much parents value ed-
ucation (Comer, 1995; Epstein, 2001). However, in middle school,
school-based involvement has been shown to change from assist-
ing in the classroom to attendance at school activities (Seginer,
2006). This latter type of school-based involvement is less likely to
provide middle school parents with information about pedagogy
and classroom content or the opportunity to create mutual respect
between parents and teachers. Therefore, its relation with aca-
demic outcomes may be weaker.
Home-based involvement has been advocated because it affirms
the knowledge and instruction received at school (Comer, 1995),
provides assistance and clarification with homework (Cooper,
1989), provides structure for free time and homework time (Fan &
Chen, 2001), includes visiting museums and other educational
venues (Reynolds & Gill, 1994), and enhances and encourages
motivations (Hoover-Dempsey & Sandler, 1995). In addition, as
part of home-based involvement, parents can supplement instruc-
tion through educationally based, cognitively stimulating activities
(Chao, 2000; Grolnick & Slowiaczek, 1994). However, by middle
school, many parents feel less able to assist with homework or
provide activities and experiences that increase their adolescents’
knowledge or achievement (Dauber & Epstein, 1993). Therefore,
the amount and type of home-based involvement that is effective
may be reduced during the middle school years (Seginer, 2006).
Yet another reason why parental involvement might change in
significance is that aspects of the middle school structure do not
support home- and school-based involvement strategies in the
same way as in elementary school.
741
PARENTAL INVOLVEMENT: A META-ANALYSIS
Middle School Context and Parental Involvement
The middle school context presents a number of challenges that
may undermine parents’ ability to be effectively involved in their
adolescents’ education and work productively with schools
(Dauber & Epstein, 1989; Hill & Chao, 2009; Sanders & Epstein,
2000). First, middle schools are large and complex, often making
it difficult for parents to figure out how to become effectively
involved. Second, middle school teachers instruct a large number
of students, making it difficult for teachers to develop and maintain
productive relations with the parents of each student. Further, the
departmentalization or specialization of instruction by academic
subject results in teachers having fewer interactions with individ-
ual students (Dornbusch & Glasgow, 1996; Eccles & Harold,
1996). Third, and in conjunction with the previous point, the
increase in the number of teachers each student has across subjects
makes it difficult for parents to know whom to contact to obtain
information about their adolescents’ progress. Fourth, the com-
plexity of curricular choices and the often obscured nature of
course tracking in middle school further complicate parental in-
volvement (Hill & Taylor, 2004). Not only does the middle school
context impact the types of involvement that matter, adolescents’
development itself impacts how parents can maintain involvement
and its effectiveness (Hill & Chao, 2009).
Early Adolescent Development and Parental Involvement
in Education
The types of parental involvement used and the nature of the
relation between parental involvement and achievement may be
influenced by characteristics of early adolescent development and
family dynamics during adolescence. As has been outlined exten-
sively elsewhere (Adams & Berzonsky, 2003; Lerner & Steinberg,
2004), adolescence is marked by dramatic cognitive development
and the development of conceptualizations of the self as an auton-
omous, efficacious individual. Cognitively, adolescents have an
increased ability to consider multiple dimensions of problems
simultaneously when making decisions (Keating, 2004). In addi-
tion, adolescents have an increased ability to anticipate the results
and consequences of their actions and decisions (Halpern-Felsher
& Cauffman, 2001), learn from their successes and failures and
apply that knowledge to future problem solving, and strategically
coordinate the pursuit of multiple goals (Byrnes, Miller, & Reyn-
olds, 1999).
Each of these abilities enables adolescents to play a more active
role in their education and educational decisions. These cognitive
changes may increase adolescents’ sense of efficacy, ability to
make decisions about course selection, and ability to understand
how courses and extracurricular activities are related to goals and
aspirations in the immediate time frame and for the future and
thereby decrease their need for direct parental involvement. That
is, more direct involvement strategies, such as school-based in-
volvement and direct homework assistance, may be needed less
and thus are less effective (Seginer, 2006). Indeed, students’ in-
creased sense of autonomy is associated with their desire to not
have their parents visit the school (Stevenson & Baker, 1987).
Often parents interpret students’ desire for autonomy as a cue to
reduce more direct forms of parental involvement, such as home-
and school-based involvement (Prescott, Pelton, & Dornbusch,
1986).
In addition to cognitive development, parent–adolescent rela-
tionships undergo a transformation and renegotiation during ado-
lescence as they become less hierarchical and are characterized by
increased bidirectional communication (Collins & Laursen, 2004;
Steinberg & Silk, 2002). Early adolescence is marked by the need
for a realignment of roles and expectations as adolescents question
their parents’ authority (Grolnick et al., 2007; Smetana et al.,
2004) and as parents attempt to set boundaries and communicate
expectations while promoting healthy independence. Parental in-
fluence often becomes more indirect. Parents’ beliefs about ado-
lescents’ abilities, skills, and potential shape adolescents’ own
beliefs, which influence their performance (Bleeker & Jacobs,
2004; Jones & Schneider, 2009).
As parental influence becomes more indirect and promotes the
use of adolescents’ developing decision-making skills, strategies
for involvement in education should change as well. For early
adolescence, parental involvement may entail communicating pa-
rental expectations for education and its value or utility, linking
schoolwork to current events, fostering educational and occupa-
tional aspirations, discussing learning strategies with children, and
making preparations and plans for the future—that is, academic
socialization. We hypothesize that involvement that scaffolds ad-
olescents’ burgeoning decision-making and problem solving skills
and elucidates linkages between their schoolwork and future goals
may be more strongly linked to achievement in middle school than
is home- or school-based involvement. Parental involvement in
education that reflects academic socialization allows parents to
maintain their involvement while also affirming adolescents’ au-
tonomy, independence, and advancing cognitive abilities.
In this meta-analysis, we examine the relative association be-
tween three types of parental involvement in education and aca-
demic achievement. Home-based involvement includes strategies
like communication between parents’ and children about school,
engagement with school work (e.g., homework help), taking chil-
dren to events and places that foster academic success (i.e., mu-
seums, libraries, etc.), and creating a learning environment at home
(e.g., making educational materials accessible, such as books,
newspapers, educational toys). School-based involvement includes
visits to school for school events (e.g., PTA meetings, open
houses, etc.), participation in school governance, volunteering at
school, and communication between parents and school personnel.
Finally, academic socialization includes communicating parental
expectations for education and its value or utility, linking school-
work to current events, fostering educational and occupational
aspirations, discussing learning strategies with children, and mak-
ing preparations and plans for the future.
Ethnic Variations in Parental Involvement in Education
In addition to outlining types of parental involvement strat-
egies, prior research has demonstrated ethnic differences in
mean levels of parental involvement strategies (Baker &
Stevenson, 1986; Kohl et al., 2000), parents’ beliefs about
involvement (Lareau, 1987; Lynch & Stein, 1987), and the
relations between parental involvement and academic outcomes
(e.g., Hill et al., 2004; Hill & Craft, 2003). African Americans,
in particular, have had a long and tumultuous history with
American schools (Cross, 2003; Spencer, Cross, Harpalani, &
Goss, 2003). Whereas African American cultural heritage has
742
HILL AND TYSON
placed an emphasis on the value and utility of education,
discrimination and bias experienced at school by many African
Americans has resulted in a mistrust of school and teachers by
many African American parents (Lareau, 1987; Ogbu, 1978).
These historical and contemporary experiences may influence
the nature of parental involvement and its influence. This is
heightened during adolescence, because it is a time when Af-
rican American students are grappling with their own ethnic
identity (Hughes et al., 2006).
Prior research on ethnic differences in parental involvement has
been mixed. Some research found that the relation between in-
volvement and achievement is stronger for African Americans
than European Americans (Hill et al., 2004), whereas others found
that the relation is weaker (e.g., Seyfried & Chung, 2002). In a
meta-analysis across prekindergarten to 12th grade, ethnicity had a
negligible effect (Fan & Chen, 2001). Another meta-analysis
found that the relation was positive for ethnic minorities (i.e.,
collapsing across African Americans, Latinos, and Asian Ameri-
cans), but the strength of the relation was not compared across
ethnicity (Jeynes, 2007). Although much has been written about
ethnic differences in levels and types of involvement, it is unclear
whether to expect the relation between involvement and achieve-
ment to vary across ethnicity.
The Meta-Analysis
The current investigation used meta-analytic techniques to
synthesize the results of the existing empirical literature to
determine the extent to which parental involvement is posi-
tively associated with achievement outcomes in middle school
and which types of involvement have the strongest relation. We
expect that involvement characterized as academic socialization
will have the strongest positive relation with achievement out-
comes as it empowers adolescents to act semiautonomously and
understand the consequences and purposes of their actions;
home-based and school-based involvement will have smaller
relations. To assess the empirical evidence of ethnic variations
in the relation between parental involvement and achievement,
we examined differences between African Americans and Eu-
ropean Americans in an exploratory manner.
Method
Extant Literature
To limit the potential cohort effects, we restricted our review
of the literature to those studies published between 1985 and
2006. The exhaustive search of the extant literature published
since 1985 produced 50 empirical reports (or articles). This set
of reports represents 127 correlations and 82 beta coefficients
for the relation between different types of parent involvement
and an array of achievement outcomes. These reports represent
three types of studies:
1. Naturalistic longitudinal and cross-sectional studies
that included correlations between parental involve-
ment and achievement (n ⫽ 27). These 27 articles and
unpublished data sets represented 32 different samples
and 92 separate correlations (see Table 1). Table 2
shows the studies that included partial correlations that
controlled for demographic or other variables (N ⫽ 1).
Table 3 shows those studies that were longitudinal in
design, which may provide some evidence of the di-
rection of effect (N ⫽ 2).
2. Studies that reported on the effects of interventions
designed to enhance parental involvement (N ⫽ 5; see
Table 4).
3. Articles that reported on data from public-access, na-
tionally representative datasets (e.g., the National Ed-
ucation Longitudinal Study 1988 [NELS-88], the Lon-
gitudinal Study of American Youth [LSAY], and the
National Longitudinal Survey of Youth [NLSY]; N ⫽
13; see Tables 5, 6, and 7).
When multiple articles use the same dataset, they pose the risk
of lack of independence and overrepresentation of the data in the
meta-analysis. When they provided enough information to calcu-
late an effect size, these effect sizes were averaged using the
shifting unit of analysis approach across all effect sizes and in-
cluded in the meta-analysis. In addition, there were 5 studies that
included middle school students and either older or younger stu-
dents and 1 study that used regression analyses but did not include
correlations (see Table 8).
Literature Search Procedures
To conduct a comprehensive review of the literature, we
searched the major databases that catalogue research abstracts.
These included PsychInfo, ERIC, Dissertation Abstracts Inter-
national, and Sociological Abstracts. Due to the paucity of
literature focusing solely on parental involvement in middle
school as a primary emphasis, extensive hand searches were
also conducted for the following journals to identify articles
that included parental involvement in education as a secondary
focus and thereby might have been missed in the database
search. The following journals were hand searched: American
Education Research Journal, Child Development, Developmen-
tal Psychology, Journal of Educational Psychology, Journal of
Family Psychology, Journal for Research on Adolescence, and
Review of Education Research.
The searches were conducted by combining terms reflecting mid-
dle school populations with terms reflecting parental involvement in
education. The following terms and phrases were used to reflect
parental involvement in education: parent involvement, parent–school
partnership, parent–school relationⴱ, family–school partnership,
family–school involvement, parental involvement in education, par-
enting involvement in school, family–school relationⴱ, family–school
involvement, and family involvement in education (the asterisk indi-
cates that all forms of that stem were included; e.g., relationⴱ includes
relations, relationship, and relationships). The search was later ex-
panded to include terms such as parent risk factors, as the initial
search demonstrated that many articles examining achievement out-
comes in middle school focused on parental influences on nonnor-
mative developmental trajectories for middle school students. To
identify studies of middle school samples, we used the following
search terms: middle schoolⴱ, middle school education, middle school
transition, junior high, junior high schoolⴱ, junior high students,
743
PARENTAL INVOLVEMENT: A META-ANALYSIS
Table 1
Studies Including Bivariate Correlations for the Relationship Between Parent Involvement and Achievement (25 studies, 32 Independent Samples)
Author (year) and
publication type Sample size Grade Type of parent involvement
Category for
meta-analysis
Outcome measure
Reading Math GPA
Baker & Stevenson (1986)
Journal article
40 8th Knowledge of child’s schooling (knows the names
of homeroom and other teachers, know how
child performance compares to previous year,
saw last report card)
Academic
socialization
⫹.24
Contact w/school (number of conferences
attended, contact teacher about specific school
problem)
School-based ⫹.10
Homework help (used a strategy to find out about
homework assignments, number of strategies
used)
Home-based ⫺.05
General academic strategies (used a tutor, used
sanctions in reaction to school grades)
General
involvement
⫺.35
High school course selection (selecting college
preparatory course for high school)
Academic
socialization
⫹.73
Balli et al. (1997)
Journal article
74 6th Homework help Home-based Reading
achievement:
⫺.24
Bandura et al. (1996)
Journal article
279 6th & 7th;
cross-
sectional
Communicating career aspirations Academic
socialization
⫹.60
Bronstein et al. (2005)
Journal article
77 7th Homework surveillance (“What percentage of the
time do you remind your child to do his/her
homework?”)
Home-based Total battery of
Stanford
achievement:
⫺.35
⫺.49
Homework surveillance (“What proportion of the
time do you insist that your child do his/her
homework?”)
Total battery of
Stanford
achievement:
⫺.20
⫺.27
Driessen et al. (2005)
Journal article
12,000 8th Homework help Home-based Language: ⫺.12 ⫺.21
Parents ask for information about school matters School-based Language: ⫺.01 ⫺.02
Leisure involvement (go to exhibitions, museums) Home-based Language: ⫹.13 ⫹.10
Garcia Bacete &
Ramirez (2001)
Journal article
150 7th (in Spain) Meetings with teacher School-based ⫹.39
Attending school events School-based ⫹.43
Grolnick & Slowiaczek (1994)
Journal article
302 6th, 7th, & 8th;
cross-
sectional
Mother behavioral involvement (attending parent-
teacher conferences, open houses, open school
nights, and other events)
School-based ⫹.30
Mother personal involvement (knowing what
happens at school, when the report card comes
out)
Academic
socialization
⫹.17
(table continues)
744
HILL AND TYSON
Table 1 (continued)
Author (year) and
publication type Sample size Grade Type of parent involvement
Category for
meta-analysis
Outcome measure
Reading Math GPA
Mother cognitive involvement (encouraging
reading newspapers and books, talking about
current events, taking child to the library or
museum)
Home-based ⫹.14
Father behavioral involvement School-based ⫹.19
Father personal involvement Academic
socialization
⫹.17
Father cognitive involvement Home-based ⫹.14
Grolnick et al. (2000)
Journal article
60 6th School involvement (includes participating in
events at child’s school)
School-based ⫹.33 ⫹.29
Cognitive involvement (intellectually stimulating
activities, going to the library, discussing
current events)
Home-based ⫹.42 ⫹.42
Personal involvement (parent knows classmates,
knows when report card comes out)
Academic
socialization
⫹.35 ⫹.27
Gutman & Eccles (1999)
Journal article
617 7th; two waves
(longitudinal)
Classroom volunteer School-based ⫹.14
Open house attendance School-based ⫹.25
PTA attendance School-based ⫹.10
Gutman & Midgley (2000)
Journal article
62 5th (students
pre- and post
transition to
middle
school)
General involvement (checking homework,
working as a school program supporter,
chaperoning field trips, discussing assignments)
General
involvement
⫹.07
Gutman et al. (2002)
Journal article
837 7th Parent school involvement (parent serves on
school staff, advocate, decision maker)
School-based Standardized test
scores: ⫹.07
⫹.10
Harris (1990)
Dissertation
76 8th Homework help (estimate of the extent of parental
awareness and support of homework, frequency
of assignments that requested parent help)
Home-based Test scores:
⫹.18
Language test
scores: ⫺.03
Test scores: ⫺.02
Communicating with teacher School-based Test scores:
⫺.01
Language test
scores: ⫺.28
Test scores: ⫺
.23
Hawes & Plourde (2005)
Journal article
48 6th General parent involvement (time spent helping
with homework or reading, time spent
conferencing with teachers)
General
involvement
Reading
comprehension
test: ⫹.13
Hughes et al. (2006)
Unpublished data
90 European
American
6th; correlations
provided by
ethnicity
School involvement School-based ⫹.19, ⫹.18
Home involvement Home-based ⫹.26, ⫹.01
Parents’ academic expectations Academic
socialization
⫹.31, ⫹.06
(table continues)
745
PARENTAL INVOLVEMENT: A META-ANALYSIS
Table 1 (continued)
Author (year) and
publication type Sample size Grade Type of parent involvement
Category for
meta-analysis
Outcome measure
Reading Math GPA
Hughes et al. (2006)
Unpublished data
183 African
American
School involvement School-based ⫹.16, ⫹.17
Home involvement Home-based ⫺.02, ⫹.12
Parents’ academic expectations Academic
socialization
⫹.39, ⫺.11
183 Chinese
American
School involvement School-based ⫺.07, ⫺.31
Home involvement Home-based ⫹.28, ⫹.26
Parents’ academic expectations Academic
socialization
⫹.23, ⫹.10
161 Hispanic School involvement School-based ⫹.27, ⫹.11
Home involvement Home-based ⫺.22, ⫹.12
Parents’ academic expectations Academic
socialization
⫹.19, ⫹.03
Jodl et al. (2001)
Journal article
444 7th Academic activity involvement (help with
homework, asking how the child did at school)
Home-based Academic ability (i.e.,
“How good is
your child in
math/how well
do you think
your child will
do in math?”)
Mother ⫹.09
Father ⫹.13
School involvement (PTA, class trips, etc.) School-based Mother ⫹.19
Father ⫹.06
Parents’ value of education (“What are the
chances your child will do well in junior high,
high school, and college?”)
Academic
socialization
Mother ⫹.59
Father ⫹.60
Educational expectations/aspirations for youth
(“How far would you like your child to go in
school?”)
Academic
socialization
Mother ⫹.35
Father ⫹.42
Juang & Silbereisen (2002)
Journal article
641 6th (German
sample)
Parent involvement (educational encouragement
and stimulation, interest in adolescent’s
schooling
Academic
socialization
⫹.52
Latendresse (2005)
Dissertation
309 6th Parent expectations (parents expect excellence and
have high expectations for children)
Academic
socialization
⫹.19
(table continues)
746
HILL AND TYSON
Table 1 (continued)
Author (year) and
publication type Sample size Grade Type of parent involvement
Category for
meta-analysis
Outcome measure
Reading Math GPA
Marchant et al. (2001)
Journal article
230 5th & 6th;
cross-
sectional
Parent values and attitudes about the importance
of academic success
Academic
socialization
⫹.25
School involvement (attending school events and
functions)
School-based ⫹.26
Reynolds & Gill (1994) 729 6th; longitudinal School involvement (participation in school
activities, talks to teacher about child progress,
PTA)
School-based Standardized test
scores: ⫹.12
Standardized test
scores: ⫹.11
Going to the museum, zoo Home-based ⫹.10 ⫹.10
Seidman et al. (2003)
Journal article
500 6th & 7th;
longitudinal
Leisure time involvement (participating in social,
cultural, and athletic events)
Home-based ⫹.07
Seyfried & Chung (2002)
Journal article
372 European
American
5th grade Expectations/career aspirations Academic
socialization
⫹.53
195 African
American
⫹.29
Sirin & Rogers-Sirin (2004)
Journal article
75 7th & 8th;
cross-
sectional
Communicating the value of education; obtained
through email
Academic
socialization
⫹.06
a
Useem (1992)
Journal article
88 6th Parent knowledge of the mathematics tracking
system
Academic
socialization
⫹.56
Parent integration (involvement in volunteer
activities at the school or informal school
networks)
School-based ⫹.57
Parent intervention (acting on behalf of child with
teachers to change or “customize” the
educational experience)
Academic
socialization
⫹.43
Xu & Corno (2003)
Journal article
118 6th Help with homework Home-based ⫺.03
a
Correlation obtained from S. R. Sirin via a separate analysis conducted for 7th and 8th grade students only (article collapses across middle and high school).
747
PARENTAL INVOLVEMENT: A META-ANALYSIS
junior high transitions, and early adolescence. We combined each
middle school term with a parental involvement term and then exam-
ined each study to determine whether an achievement outcome was
included. Achievement outcomes included grades, course or class
grades, grade point averages (GPA), test scores, and placement in
advanced courses.
In addition, descendant searches were conducted on major papers
in the field. Using the Social Sciences Citation Index, we located
articles that cited seminal articles, such as publications outlining the
major theories in the field (e.g., Comer, 1995; Conners & Epstein,
1995; Eccles & Harold, 1996; Epstein, 1987; Epstein & Sanders,
2002; Grolnick & Slowiaczek, 1994). In addition, we conducted
descendent searches on papers by authors who appeared in our search
at least twice. We conducted backward or ancestry searches by
examining the reference lists of all papers that resulted from our
search. Further, we identified key researchers in the field (defined as
having two or more papers in our database; n ⫽ 11), contacted them,
and requested relevant new work in press and unpublished findings.
We received responses with data from 4 of the 11 researchers con-
tacted. To maximize our sample size, we also contacted 4 authors of
recent papers that included middle school students combined with
other age groups to request correlations for just the middle-school-age
sample. We received one response with correlations for the middle-
school-age sample for the data reported in Sirin and Rogers-Sirin
(2004).
Criteria for inclusion. The criteria for inclusion of an article in
this meta-analysis were threefold. First, the report had to include a
measure of parental involvement and academic achievement. Because
the focus of this meta-analysis was the relation between parental
practices and academic outcomes, we limited the set of research
reports to those that measured academic outcomes. Second, the iden-
tified research reports needed to be based on middle school samples,
which are typically defined as Grades 6 through 8. However, in an
attempt to comprehensively account for middle school populations,
we used explicit identification as well. For example, when fifth or
ninth grade students were included in a study and the authors identi-
fied the population as middle school students, these reports were
included (e.g., Gutman & Midgley, 2000; Marchant, Paulson, &
Rothlisberg, 2001). Third, the report needed to include correlations
(Pearson’s r), d indexes, or sufficient information to calculate an
estimate of the effect size. This included means and standard devia-
tions to calculate the d index as outlined by Rosenthal (1991). Studies
that used a wide variety of statistical analyses were included, such as
studies that used structural equation modeling, hierarchical linear
modeling, regression, and other statistical techniques, provided that
the article also included information from which effect sizes could be
calculated. Reports were not included in the meta-analysis if they did
not include such information and it was not available from the author.
However, their findings and the directions of effects were coded to
Table 2
Studies Estimating the Effect of Parent Involvement Using Partial Correlations
Author (year) and
publication type Sample size Grade
Type of parent
involvement
Category for
meta-analysis
Outcome
measure Correlation
Kim (2002)
Journal article
245 Korean
adolescents
6th–12th General parent
involvement
General involvement GPA Partial correlations controlling for
age and parent level of
education: ⫹.16, ⫹.25
Note. GPA ⫽ grade point average.
Table 3
Studies Estimating the Long-Term Effect of Parent Involvement (Longitudinal Only)
Author (year)
and publication
type
Sample
size Grade
Type of parent
involvement
Category for meta-
analysis
Outcome measure
Reading Math GPA
Hill et al. (2004)
Journal article
463 7th grade students tested at
four waves
(7th grade parent
involvement correlated
with 9th grade scores)
General involvement General involvement
Mother report Grade: ⫹.13 Grade: ⫹.09
Standard test: ⫹.11 Standard test: ⫺.00
Student report Grade: ⫹.06 Grade: ⫹.11
Standard test: ⫹.04 Standard test: ⫺.01
Teacher report Grade: ⫹.08 Grade: ⫹.06
Standard test: ⫹.14 Standard test: ⫹.14
Melby & Conger
(1996)
Journal article
347 7th grade students tested at
four waves
General involvement
(setting standards
for behavior, such
as completing
schoolwork)
General involvement ⫹.22
8th grade parent
involvement predicting
11th grade GPA
⫹.27
Note. GPA ⫽ grade point average.
748
HILL AND TYSON
determine whether they provided additional support to the general
results of the meta-analysis.
Criteria for exclusion. We eliminated studies that used a broad
conceptualization of parental involvement, such as reports that in-
cluded assessments of general monitoring, parenting styles (i.e., au-
thoritative vs. authoritarian), or discipline in their conceptualization of
parental involvement in education, as this meta-analysis focused on
the specific strategies parents use to foster achievement on their own
and in collaboration with school. In addition, one central goal of this
meta-analysis was to determine whether parental involvement in
education, as defined by the prevailing theories of parental involve-
ment and family–school relations, was related to achievement among
middle school students. Studies were also eliminated if they combined
middle-school-age students with other age groups, as the focus of this
meta-analysis was on early adolescence and middle school as a unique
developmental stage and context. Finally, studies were eliminated that
demonstrated a lack of overall face validity (e.g., indicated that their
focus was on parental involvement but did not assess parental in-
volvement as defined by the prevailing theories). These criteria for
inclusion resulted in 50 articles, reflecting 127 separate correlations.
Coding and Information Retrieved From Research
Reports
Coding. The coding scheme was developed to extract infor-
mation about the characteristics of the sample, type of publi-
cation, theories used, conceptualization and measurement of
parental involvement and academic outcomes, and results (i.e.,
effect sizes, information to calculate effect sizes, and general
magnitude and direction of the relation between parental in-
volvement and academic outcomes). Each research report was
coded by two coders. The primary coder was a developmental
psychology graduate student with expertise in parental involve-
ment in education, academic motivation, and academic achieve-
ment. The secondary coder was an advanced undergraduate
research assistant who received extensive training on the theo-
ries and research related to parental involvement in education
and the goals of the meta-analysis. After retrieving abstracts
from the databases, each coder examined each abstract to de-
termine its relevance. If either coder deemed an abstract was
relevant, the full article was retrieved for further coding. Re-
search reports received through communications with key re-
searchers in the field were also included in this round of coding.
Both coders extracted information from each report selected for
inclusion. Any discrepancies between coders were discussed
until consensus was reached. If agreement could not be reached,
the disagreement was resolved by Nancy E. Hill. This method
of discussing discrepancies until consensus is reached, as a way
of assuring intercoder reliability, is consistent with the meta-
analysis methodology presented in Rosenthal (1991). This
meta-analysis focused on three types of involvement (home-
based involvement, school-based involvement, and academic
socialization). Studies were coded for these types of involve-
ment. In addition, several studies created measures of parent
involvement that combined several types of involvement into a
Table 4
Interventions
Author (year) and
publication type
Sample size
Grade Type of parent intervention
Category for
meta-analysis Outcome measure Pearson’s r
a
Treatment Control
Balli et. al (1998)
Journal article
22 25 6th Epstein’s TIPS intervention (Epstein
et al., 1995) used in math classes
Home-based 40-item mathematics
posttest
r ⫽ .29
Keisner (1997)
Dissertation
11 12 6th–8th Parents received instruction for
supervising math homework
Home-based Math achievement
test
r ⫽⫺.06
Cumulative math
achievement test
r ⫽⫺.26
Class grades r ⫽⫺.67
Rillero & Helgeson (1995)
Conference presentation
101 99 6th Quasi-experimental study: SPLASH
(Student-Parent Laboratories
Achieving Science at Home)
intervention where students were
given homework assignments that
required parent participation
Home-based Science homework None provided;
results
reported as
significant,
entered as
r ⫽ .20
b
Tamayo (1992)
Dissertation
16 15 7th Parents received instruction for
monitoring math homework;
control group parents received no
instruction
Home-based Math grade r ⫽⫺.07
Homework r ⫽⫺.06
Stanford
achievement test
r ⫽⫺.08
Van Voorhis (2003)
Journal article
146 107 6th & 8th TIPS intervention (Epstein et al.,
1995) used in science classes
Home-based Homework grades r ⫽ .15
Science class grades r ⫽ .15
a
For consistency of data presentation, the d indexes for the quasi-experimental studies have been converted to Pearson’s rs.
b
Study reported the effect
size as significant but did not provide an effect size. Therefore, we used the formula r ⫽ Z/公N, where Z ⫽ 1.96 and N is the sample size for the study.
749
PARENTAL INVOLVEMENT: A META-ANALYSIS
Table 5
Studies Using the National Education Longitudinal Study (NELS)
Author (year) and
publication type Sample characteristics
Modeling
technique Predictor variables
Outcome
variable
Regression
coefficient
Size and
significance
Desimone (1999)
a
Journal article
19,386 NELS 8th grade
students; 13,483
European American
b
Multiple
regression
Discussion with child about high school GPA ⫽.03 p ⫽ .050
Talk w/parents about post-high school
plans
⫽.05 p ⫽ .001
Volunteering or fundraising ⫽.11 p ⫽ .001
Rules about homework, GPA, and chores ⫽⫺.11 p ⫽ .001
PTO involvement ⫽.08 p ⫽ .001
Parent attends PTO meetings ⫽.00 ns
Rules about TV, friends, & chores ⫽.03 p ⫽ .001
Parents check homework ⫽⫺.06 p ⫽ .001
Contact school about academics ⫽⫺.19 p ⫽ .001
Discussion with parents about school ⫽.18 p ⫽ .001
Talk with father about planning high
school programs
⫽.02 ns
Social capital: knowing parents of child’s
friends
⫽.03 p ⫽ .001
2,334 African
American
Discussion with child about high school ⫽⫺.06 p ⫽ .057
Talk w/parents about post-high school
plans
⫽.05 p ⫽ .030
Volunteering or fundraising ⫽.03 ns
Rules about homework, GPA, and chores ⫽⫺.06 ns
PTO involvement ⫽.30 p ⫽
.001
Parent attends PTO meetings ⫽⫺.09 ns
Rules about TV, friends, and chores ⫽.05 p ⫽ .017
Parents check homework ⫽⫺.09 p ⫽ .001
Contact school about academics ⫽⫺.05 p ⫽ .005
Discussion with parents about school ⫽.08 p ⫽ .001
Talk with father about planning high
school programs
⫽.07 p ⫽ .002
Social capital: knowing parents of child’s
friends
⫽.02 p ⫽ .007
2,368 Hispanic Discussion with child about high school ⫽⫺.00 ns
Talk w/parents about post-high school
plans
⫽.02 ns
Volunteering or fundraising ⫽⫺.04 ns
Rules about homework, GPA, and chores ⫽.00 ns
PTO involvement ⫽.16 p ⫽ .051
Parent attends PTO meetings ⫽⫺.02 ns
Rules about TV, friends, and chores ⫽.05 p ⫽ .013
Parents check homework ⫽⫺.06 p ⫽ .001
Contact school about academics ⫽⫺.12 p ⫽ .001
Discussion with parents about school ⫽.19 p ⫽ .001
Talk with father about planning high
school programs
⫽.03 ns
Social capital: knowing parents of child’s
friends
⫽⫺.00 ns
1,201 Asian Discussion with child about high school ⫽⫺.06 ns
Talk w/parents about post-high school
plans
⫽⫺.04 ns
Volunteering or fundraising ⫽.05 ns
Rules about homework, GPA, and chores ⫽.12 p ⫽ .029
PTO involvement ⫽⫺.04 ns
Parent attends PTO meetings ⫽.10 ns
Rules about TV, friends, and chores ⫽.05 ns
Parents check homework ⫽⫺.09 p ⫽ .001
Contact school about academics ⫽⫺.08 p ⫽ .005
Discussion with parents about school ⫽.08 p ⫽ .010
Talk with father about planning high
school programs
⫽.09 p ⫽ .045
Social capital: knowing parents of child’s
friends
⫽⫺.03 p ⫽ .015
(table continues)
750
HILL AND TYSON
Table 5 (continued)
Author (year) and
publication type Sample characteristics
Modeling
technique Predictor variables
Outcome
variable
Regression
coefficient
Size and
significance
Hao & Bonstead-
Bruns (1998)
Journal article
25,000 NELS 8th grade
students
HLM Parent expectations for schooling HLM estimate
Math 3.92 p ⬍ .01
Reading 3.74 p ⬍ .01
GPA 0.33 p ⬍ .01
Kelly (2004)
Journal article
13,548 NELS (Parent
involvement in 8th
grade predicting
math track placement
in 9th grade)
Logit models Block One
Parental education
Parental income
Parental occupation
Math track
placement
Block Two
Skipped a grade 0.27
Held back a grade ⫺0.32 p ⬍ .001
Parent decides which courses student
takes
0.0074
Parent requested current math course 0.42 p ⬍ .05
Number of times parent contacted
school about academic program
⫺0.049
Neutral contact with school (volunteer,
fundraising)
⫺0.15 p ⬍ .01
Involvement in PTO activities 0.11 p ⬍ .005
Discussion about school 0.021
Talking about school 0.029
Parent knows parents of student’s
friends
⫺0.0037
McNeal (1999)
c
Journal article
11,401 NELS
(8th grade parent
involvement
predicting
achievement in 10th
grade)
Not specified (Controlling for GPA, hours worked,
hours of homework, having been
retained a grade in school, and gender)
Black
Hispanic
Asian
SES
Base-achievement
Single-headed household
Social capital
Parent–child discussion 0.150 p ⬍ .01
PTO involvement ⫺0.115 p ⬍ .01
Monitoring ⫺0.071 p ⬍ .10
Educational support strategies ⫺0.014
Muller (1995)
Journal article
13,881 NELS students
(analysis of 8th
grade test scores)
Hierarchical
multiple
regression
Block One 8th grade
mathematics
achievement
test score
Family income, parents’ highest
education, gender, single mother,
mother/stepfather, Asian American,
Hispanic, African American, Catholic,
other religious school, independent
private, urban, suburban, 8th grade
math grades, part-time, not in labor
force, number of children home
Block Two
Talk about school experiences
Talk about high school program w/father
Talk about high school program
w/mother
Parents check homework b ⫽⫺1.053 p ⬍ .001
Parents restrict TV
Child enrolled in extra music class
Time unsupervised after school
Number of friends’ parents know
PTO participation b ⫽⫺0.231 p ⬍ .01
Parents contact school b ⫽⫺0.653 p ⬍ .001
Parent volunteers at school b ⫽ 0.058 ns
(table continues)
751
PARENTAL INVOLVEMENT: A META-ANALYSIS
single construct. The code “general involvement” was used
when the report did not specify a specific type of involvement
or used a unidimensional construct of parental involvement. In
the meta-analysis, the construct general involvement was cre-
ated for all reports by combining across all indicators of in-
volvement.
Information Retrieved From Research Reports
The main information coded for each article included charac-
teristics of the publication, the independent variables, the sample
characteristics, and the outcomes measures of interest.
Characteristics of the publication. First, the authors of the
research report and the date of publication were recorded. Next,
information about the type of publication or report was recorded.
These included peer-reviewed journal articles, dissertations, pri-
vate reports (i.e., the author provided correlations not originally
included in a publication), government reports, conference papers,
and unpublished datasets. In addition, the type of study was coded;
that is, whether the study was a naturalistic correlational study,
controlled for a third variable, used advanced statistics and did not
include correlations (i.e., structural equation modeling, hierarchi-
cal linear modeling), was an intervention trial, or used public-
access datasets.
Characteristics of the sample. Demographic information
about samples was gathered and coded, including sample size,
ethnic or racial background of the sample, gender of the target
child, specific grade levels or ages of adolescents included in the
study, socioeconomic status of the families (including parental
education level), and any labels that were given to the samples
(e.g., “at risk,” “exceptional students”).
Characteristics of the independent variables. Studies were
coded and correlations gathered for each type of involvement,
along with the effect size for the overall relation between parental
involvement and achievement (i.e., general involvement). As men-
tioned previously, if the research report did not allow for distinc-
tions between the types of involvement, they were coded as having
an assessment of general parental involvement.
Characteristics of the dependent variables. The main outcome
of interest for this meta-analysis was academic achievement, con-
ceptualized as class grades, GPA, standardized test scores, track
placement, and other tests designed to measure achievement.
Information to calculate an effect size. Effect sizes were as-
certained from each research report for each relation between a
type of parental involvement and the outcomes. A single article
could provide more than one effect size if multiple dimensions of
parental involvement or multiple outcomes were included. If effect
sizes were not included in the research report, information that
could be used to calculate an effect size was gathered and input
into the Comprehensive Meta-Analysis program (CMA; Version
2.0; Biostat, Englewood, NJ) to calculate the appropriate effect
Table 5 (continued)
Author (year) and
publication type Sample characteristics
Modeling
technique Predictor variables
Outcome
variable
Regression
coefficient
Size and
significance
Peng & Wright
(1994)
Journal article
24,599 NELS (overlap:
includes 8th–10th
grade)
Multiple
regression
Asian vs. Hispanic Achievement
test scores
Asian vs. Black
Asian vs. White
Asian vs. Indian
School control
Parental education
Family income
Family composition
Homework
Television
Educational activity
Outside classes
Educational aspirations ⫽.25 p ⬍ .01
Discuss school ⫽⫺.03 p ⬍ .01
Assist in homework ⫽⫺.17 p ⬍ .01
Sui-Chi &
Willms(1996)
Journal article
25,000 NELS (8th
grade)
Hierarchical
linear
modeling
Adjusted school mean SES, family and
student background, parent
involvement factors
Achievement
test scores
Home discussion (talks with parents,
discusses school activities)
Reading ⫽.124 p ⬍ .01
Math ⫽.124 p ⬍ .01
Home supervision (monitor homework,
limits TV time, home after school)
Reading ⫽.009 ns
Math ⫽.033 p ⬍ .01
School communication (school contacts
parents, parents contact school)
Reading ⫽⫺.051 p ⬍ .01
Math ⫽⫺.056 p ⬍ .01
School participation (volunteer at school,
PTO)
Reading ⫽.030 p ⬍ .01
Math ⫽.026 p
⬍ .01
Note. GPA ⫽ grade point average; PTO ⫽ parent-teacher organization; HLM ⫽ hierarchical linear modeling; SES ⫽ socioeconomic status.
a
Actual report examines GPA, reading, and math scores as a function of ethnicity and SES in multiple regressions. For the purposes of this analysis, only
GPA is reported by ethnicity.
b
Results are presented by ethnicity and SES in the article.
c
Actual report also includes a breakdown by ethnicity.
However, for the purposes of this report, we used results that were collapsed across ethnic groups.
752
HILL AND TYSON
size (e.g., correlations and Cohen’s d index). In calculating the
effect size, we incorporated the shifting unit of analysis approach
to account for independence assumptions among the variables.
According to this approach, all study effect sizes are coded as
independent events (Cooper, 1998). Then, when the overall result
for the meta-analysis was generated, the effect sizes were weighted
in CMA so that each study contributed to the overall finding on the
basis of its sample size and other characteristics. The shifting unit
of analysis approach takes into account the fact that one study can
contribute multiple effect sizes (Cooper, 1998). For example,
Marchant et al. (2001) contributed two effect sizes from the same
sample: one for the relationship between academic socialization
and GPA and one for the relationship between school involvement
and GPA. As such, the shifting unit of analysis approach takes the
average of these correlations and contributes one effect size for the
purposes of examining the relation between general parental in-
volvement and achievement. However, when conducting modera-
tor analyses, the study effect sizes are only examined across the
separate categories of the moderator (Cooper, 1998). When anal-
yses of each type of involvement are considered separately, this
approach counts one effect size per category. Thus, in the case of
Marchant et al. (2001), when the impact of different forms of
parent involvement is examined, each of the two correlations
counts independently in the analysis.
Data Integration and Meta-Analysis Plan
We used meta-analytic techniques to calculate the relations
between parental involvement and achievement and the 95% con-
fidence interval. A random effects model was used, which extrap-
olates to the entire possible pool of studies that may potentially
examine the relation between parent involvement and achieve-
ment. Thus, random effects models make the current meta-analysis
generalizable to all possible studies. This is important because it
attempts to account for unpublished studies or studies not pub-
lished in peer-reviewed journals and book chapters. In addition to
determining whether the relation between parental involvement
and achievement was significantly different from zero, we exam-
ined the heterogeneity of the distribution of the effect sizes using
the Q statistic (Rosenthal, 1991). Heterogeneity may be due to the
inclusion of outliers, multiple underlying dimensions within the
distribution, or sampling error. We used the significance of the Q
w
to determine the appropriateness of conducting the moderator
analyses. We conducted moderator analyses using meta-analytic
strategies to compare the strength of the relations between parental
involvement and achievement across the three types of involve-
ment (i.e., school-based, home-based, and academic socialization),
which were our planned comparisons. The Q
B
statistic was used to
determine whether the groups of effect sizes for each type of
parental involvement differed from each other. Due to potential
violations of independence at the sample and item level, studies
that used public-access datasets were grouped together. These
studies did not include correlation matrices, likely due to the large
sample sizes. Rather, authors used a variety of modeling tech-
niques (i.e., multiple and hierarchical regression) to examine the
relation between parent involvement and achievement. We exam-
ined and reported the range of the betas for these studies, which
included an array of control variables in addition to parent involve-
ment items. Finally, we conducted exploratory meta-analyses to
examine variations in the relation between parental involvement
and achievement between African Americans and European Amer-
icans.
Results
Overall Relation Between General Parental Involvement
and Achievement
Overall, the meta-analysis of the correlational studies dem-
onstrated a positive relation between general parental involve-
ment and achievement in middle school. The correlations
ranged from ⫺.49 to .73; the average weighted correlation
across the 32 independent samples was r ⫽ .18, 95% confidence
interval (CI) ⫽ .12, .24, Q(31) ⫽ 1,581.10,p⬍ .0001. The
distribution of these is represented in the funnel plot in Figure
1. Because the confidence interval does not include zero, we
concluded that the relation between general parental involve-
ment and academic achievement is positive and significantly
different from zero. However, due to the size and the signifi-
cance of the Q statistic, which is an assessment of the hetero-
Table 6
Studies Using National Education Longitudinal Study (NELS) and Structural Equation Modeling
Author (year) and
publication type
Sample
characteristics
Modeling
technique Predictor variables
Outcome
variable
Regression
coefficient
Size and
significance
Keith et al. (1993)
Journal article
21,814 NELS 8th
grade students
SEM Parent involvement
composite
GPA ⫽.29 Not mentioned
Keith & Lichtman (1994)
Journal article
1,714 NELS
Mexican
American 8th
grade students
SEM Parent involvement
composite
Overall GPA ⫽.12 p ⬍ .05
Reading ⫽.12 p ⬍ .05
Math ⫽.08 p ⬍ .05
Science ⫽.10 p ⬍ .05
Social
studies
⫽.12 p ⬍ .05
Singh et al. (1995)
Journal article
25,000 NELS 8th
grade students
SEM Parent aspirations GPA ⫽.28 Not mentioned
Participation at
school
⫽⫺.02
Home environment ⫽⫺.10
Note. SEM ⫽ structural equation modeling; GPA ⫽ grade point average.
753
PARENTAL INVOLVEMENT: A META-ANALYSIS
Table 7
National Datasets Other Than the National Education Longitudinal Study (NELS)
Author (year) and
publication type
Sample
characteristics Modeling technique Predictor variables
a
Outcome variable Regression coefficient
Size and
significance
Eamon (2005)
Journal article
388 Latino youth
from the NLSY,
ages 10–14
Hierarchical multiple
regression
Block 1
Ethnicity, gender, age,
Reading achievement ⫽.09 p ⬍ .05
Block 2
Youth’s English language problem,
mother’s age at 1st birth,
mother’s years of education,
mother’s AFQT score, no
English language problem,
mother born in U.S., average
adult–child ratio, proportion of
youth’s lives poor
Math achievement ⫽.11 p ⬍ .05
Block Three
School environment, neighborhood
quality
Block Four
Cognitive stimulation, parent-youth
conflict, academic involvement
Ma (1999)
Journal article
3,116 students from
the LSAY
(longitudinal
analysis, 7th–
12th grade)
Logistic survival
analysis/event
history analysis
Parent expectations
Parent college planning
Volunteering
Participation in
advanced math
from Grades 8–12
b
Parent expectations and
college planning
linked to increased
achievement.
b
Students of parents
who volunteer were 9
times more likely to
take advanced
mathematics in
Grade 12.
Shumow & Miller
(2001)
Journal article
1,670 seventh and
8th grade
students from the
LSAY
c
Hierarchical multiple
regression
Block 1
Parent gender, child gender, past
school adjustment, parent
education level
Block 2
Parent involvement at home
(parental assistance with
homework including writing,
mathematics, and special
projects)
GPA Achievement
test scores
⫽⫺.10
⫽⫺.14
p ⬍ .001
p ⬍ .001
Parent involvement at school
(school visits, level of parent-
teacher organization
membership, attentiveness to
school issues)
GPA Achievement
test scores
⫽.05
⫽.02
p ⬍ .01 ns
Note. NLSY ⫽ National Longitudinal Survey of Youth; LSAY ⫽ Longitudinal Study of Adolescent Youth; GPA ⫽ grade point average; AFQT ⫽ Armed Forces Qualification Test Score.
a
Predictor variables in italics are the parent involvement variables; regression coefficients for each parent involvement predictor available are listed.
b
No effect size reported.
c
11.3% African
American; 3.9% Asian American; 10.3% Hispanic American; 67.7% Caucasian American.
754
HILL AND TYSON
Table 8
Studies Including Bivariate Correlations for the Relationship Between Parent Involvement and Achievement Where There Is an Age-Group Overlap
Author (year) and
publication type Sample size Grade
Type of parent
involvement
Category for meta-
analysis Outcome measure Correlation
Deutscher & Ibe (n.d.)
Web-based article
400 7th–11th Volunteering at school School-based California Standards Test
English ⫹.19
Math ⫹.23
Reading school newsletter Home-based Math ⫺.29
Checking student planner Home-based Math ⫹.28
Pelegrina et al. (2003)
Journal article
323 Adolescents, 11–15 years old Grolnick & Slowiaczek
(1994): Parent
involvement scale,
combined as general
involvement
General involvement GPA (math, language,
and science)
⫹.29
Salazar et al. (2001)
Journal article
400 7th–12th Parent involvement
(general)
General involvement Student involvement
(measure of effort,
persistence, and
interest in various
subjects
⫹.28
Shumow & Lomax (2002)
Journal article
929 students
(677 mothers, 322 fathers)
10- to 17-year-old students
from the Survey of Parents
and Children (1990)
Parent involvement
(attending events and
activities, talking to
teachers, attending PTA
meetings)
School-based GPA (overall) ⫹.21
GPA European
Americans
⫹.21
GPA African Americans ⫹.24
GPA Latinos ⫹.16
Stevenson & Baker (1987)
Journal article
179 5–17 years old:
5–8, 21.2%
9–11, 25.8%
12–14, 26.2%
15–17, 26.8%
Parental involvement in
school activities (teacher
report of the extent to
which parents participate
in PTO and conferences)
School-based Child performance (i.e.
how well is the student
performing relative to
his/her ability, rated on a
scale from 1 to 5)
⫹.34
Note. GPA ⫽ grade point average; PTO ⫽ parent–teacher organization.
755
PARENTAL INVOLVEMENT: A META-ANALYSIS
geneity of the distribution of correlations, there is likely more
than one underlying construct of parental involvement with
differing associations with academic outcomes.
When a meta-analysis is conducted, one common concern is
publication bias. That is, the field often has a bias against
publishing null results, which may render a meta-analysis based
on published studies biased in favor of statistical significance.
To prevent this, we attempted to obtain unpublished data from
key researchers in the field. However, this is not always pos-
sible. To obtain an estimate of the publication bias, the “trim
and fill” technique was used to impute potentially missing
studies (Duval, 2005; Taylor & Tweedie, 2000). The trim and
fill technique is based on the assumption that the full set of
possible studies on a topic will be distributed symmetrically
around a true mean. To estimate the number of plausibly
missing studies, the trim and fill method “trims” the outlying
studies that do not have a counterpart on the other side of the
mean. The mean effect is recalculated, often resulting in a more
conservative effect size; the outlying studies are returned, and
their counterparts are estimated based on the new mean level
effect size. Using this method, we estimated that 11 studies
were potentially missing. These imputed studies were each
below the mean and had a negative correlation between parental
involvement and achievement (see Figure 1).
Next, we conducted moderator analyses to determine whether
the strength of the relation between parental involvement and
achievement varied among the three types of involvement.
Are All Types of Involvement Equally Effective?
All but six samples provided separate correlations for specific
types of parental involvement (i.e., they only included assessments
of general forms of parental involvement or collapsed across the
different types of involvement). To determine the extent to which
each type of involvement was similarly related to achievement, we
first examined whether the simple relation between each type of
involvement and achievement was significantly different from
zero. Second, we compared the magnitudes of the relations across
type of involvement using the Q
B
statistic to determine whether
one type of involvement was more strongly related to achievement
than another.
Average weighted correlations between parental involvement
and achievement were positive and significantly different from
zero for school-based involvement and academic socialization
(See Table 9). For home-based involvement, the relation was not
significant. The 95% CI included zero, indicating that we could not
rule out that the relation between home-based involvement and
achievement was not significantly different from zero.
In comparing the strength of the relations across types of pa-
rental involvement, we found that the average weighted correlation
for each type of involvement and achievement was significantly
different, Q
B
(2) ⫽ 38.10, p ⬍ .0001. Three planned contrasts were
conducted to examine the differences between types of involve-
ment. These included a comparison between academic socializa-
tion and home-based involvement, between academic socialization
and school-based involvement, and between home- and school-
based involvement. Academic socialization was more strongly
related to achievement than was home-based involvement. The
average weighted correlation between academic socialization and
achievement and between home-based involvement and achieve-
ment were r ⫽ .39 and .03, respectively; Q
B
(1) ⫽ 36.68, p ⬍
.0001. For the comparison between academic socialization and
school-based involvement and their relation to achievement out-
comes, the relation was also stronger for academic socialization.
Whereas the average weighted correlation for academic socializa-
tion and achievement was .39, it was .19 for school-based involve-
ment and achievement, Q
B
(1) ⫽ 13.30, p ⬍ .0001. Finally, the
average weighted correlation between school-based involvement
and achievement was stronger than the average weighted correla-
-2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0
0.00
0.05
0.10
0.15
0.20
Standard Error
Fisher's Z
Funnel Plot of Standard Error by Fisher's Z
Figure 1. Funnel plot for the random effects model. The trim and fill technique imputes 11 studies to the left
of the mean. White circles represent the effect sizes of all samples from studies providing bivariate correlations.
Filled circles represent the imputed effect size.
756
HILL AND TYSON
tion between home-based involvement and achievement, Q
B
(1) ⫽
12.30, p ⬍ .0001. In summary, parental involvement in education
is positively associated with academic outcomes during middle
school. Further, among the types of parental involvement, aca-
demic socialization emerged as a critical component of parental
involvement in middle school that had the strongest positive rela-
tion with achievement.
As further evidence of the differences in the strength of the
relations between the types of involvement and achievement, the
range of the beta weights were examined for the studies using
the public-access dataset (e.g., NELS-88, NLSY, LSAY) that
could not be included in the meta-analysis. Indeed, the examina-
tion of the range suggests that the relation is stronger and more
positive for academic socialization (betas ranged from .00 to .42
for studies using the NELS-88 and .11 for the study using the
NLSY), compared to school-based involvement (betas ranged
from ⫺.06 to .11 for studies using the NELS-88 and ⫺.02 to .05
for studies using the LSAY) and home-based involvement (betas
ranged from ⫺.17 to .08 for studies using the NELS-88 and ⫺.14
to ⫺.11 for studies using the LSAY).
The relation between home-based involvement and achievement
was not significant, and it was weaker than the relation between
other types of involvement and achievement. The Q statistics for
home-based involvement suggest that there may be subtypes of home
involvement. Because prior research had suggested that home-based
involvement should be positively related to achievement, we at-
tempted to identify which types of home involvement were positively
related to achievement and whether some types of involvement had
a negative relation.
Are There Subtypes of Home Involvement?
To examine potential multidimensionality among studies of
home-based involvement, we examined the types of home-based
involvement that were assessed. Prior research suggested that
helping with homework is the most controversial type of home-
based involvement. Homework help has been shown to both ac-
celerate and interfere with achievement (Cooper, 1989, 2007;
Wolf, 1979). The negative relation may be due to parental inter-
ference with students’ autonomy, to excessive parental pressure, or
to differences between parents and schools in how they present the
material. Further, help with homework may be elicited by poor
school performance, also resulting in a negative relation between
homework help and achievement. Other types of home-based
involvement—such as providing educationally enriching activities
at home, making books and other educational materials available,
and taking children to museums, libraries, the zoo, and other
educational outlets— have been shown to have a more consistent
positive relation with achievement (Reynolds & Gill, 1994).
Therefore, we coded the studies into these two types of home-
based involvement and tested the relations with achievement using
meta-analytic techniques. There were five correlations represent-
ing involvement in activities at home and six correlations repre-
senting homework help.
Consistent with our post hoc hypothesis, help with homework
was negatively related to achievement, whereas other types of
involvement at home were significantly and positively related to
achievement. The average weighted correlation between activities
at home and achievement was .12 (95% CI ⫽ .05, .19), whereas
involvement in homework produced a significant but negative
average weighted correlation with achievement (r ⫽⫺.11; 95%
CI ⫽⫺.25, ⫺.04). These average weighted correlations were
significantly different from each other, Q
B
(1) ⫽ 7.61, p ⬍ .006.
Overall, among the types of home involvement, educationally
enriching activities were positively related to achievement, but
helping with homework was associated with lower levels of per-
formance.
In summary, parental involvement is positively related to
achievement in middle school. Further, parental involvement char-
acterized as academic socialization has the strongest and most
positive relation and helping with homework has the strongest
negative association with achievement. Other types of home-based
and school-based involvement demonstrated significant positive
relations with achievement. However, the strength of these rela-
tions was more moderate. Our final two questions were whether
Table 9
Moderator Analyses: Examining the Correlation Between Parent Involvement and Academic Achievement
Analysis kr
95% CI
Q
w
Low estimate High estimate
Overall 32 .04
ⴱⴱ
(.18)
.04 (.12) .05 (.24) 1,581.10
ⴱⴱ
Moderators Q
b
Type of parent involvement 1,206.92
ⴱⴱ
(38.10)
ⴱⴱ
School-based 21 .19
ⴱⴱ
.10 .21
Home-based 19 .03, ns ⫺.02 .11
Academic socialization 16 .39
ⴱⴱ
.26 .44
Type of home-based involvement 937.81
ⴱⴱ
(7.61)
ⴱ
Help with homework 6 ⫺.11
ⴱⴱ
⫺.04 ⫺.25
Activities at home 5 .12
ⴱⴱ
.05 .19
Ethnicity 32.67
ⴱⴱ
(1.80), ns
African American 7 .11
ⴱⴱ
.05 .17
European American 11 .19
ⴱⴱ
.09 .29
Note. Random effects Q values and point estimates are presented in parentheses. CI ⫽ confidence interval.
ⴱ
p ⬍ .005.
ⴱⴱ
p ⬍ .0001.
757
PARENTAL INVOLVEMENT: A META-ANALYSIS
the relations between parental involvement and achievement var-
ied across ethnicity and whether any evidence on the direction of
effect can be ascertained from the results of the five intervention
studies and the two longitudinal studies.
Ethnic differences in the relation between involvement and
achievement. Although most studies did not provide separate
correlations for each ethnic group, 15 studies did provide such
information for African Americans and European Americans. Six
studies and 7 samples provided data from African American par-
ticipants, and 9 studies with 11 samples provided data for Euro-
pean Americans. The overall weighted correlations suggested sim-
ilarities across ethnicities in the strength of the relation. For
African Americans, the average weighted correlation was .11
(95% CI ⫽ .05, .17); for European Americans, it was .19 (95%
CI ⫽ .09, .29). Whereas each was positive and significantly
different from zero, they were not significantly different from each
other, Q
B
(1) ⫽ 1.80, ns, suggesting that the strength of the relation
is similar between African Americans and European Americans.
1
Attempts at discerning directions of effect and causality from
longitudinal and intervention studies. Much debate in psycho-
logical research has focused on the ability to discern directions
of effects (e.g., Duncan, Magnusson, & Ludwig, 2004). Duncan
et al. suggested that research capitalize on natural experiments,
use longitudinal designs, and use quasi-experimental designs as
a way to attempt to establish causality and directions of effect.
Longitudinal and experimental studies are presented in Tables 3
and 4. As can be seen, longitudinal studies show a moderate
positive relation between parental involvement at Time 1 and
achievement at Time 2. Further, five studies employed an
experimental design that attempted to increase parental involve-
ment in education—specifically, involvement in homework—
and in turn, increase adolescents’ school performance.
The studies that used experimental designs to examine the
impact of parent training for homework were evaluated using
meta-analytic techniques to determine the nature of the relation.
The weighted mean d index was .21 and was not statistically
different from zero (95% CI ⫽⫺.54, .98). The weighed mean
correlations is .11 (95% CI ⫽⫺.26, .44). However, the test of
the distribution of d indexes was very large and significant,
Q
w
(3) ⫽ 15,074.48, p ⬍ .0001. Part of the heterogeneity may be
due to the fact that these studies were extremely different in
terms of design. Ideally, moderator analyses could be conducted
to determine whether there were subtypes of homework help
that were differently related to achievement, given the hetero-
geneity in the distribution of correlations; however, three of the
five intervention studies explicitly stated that parents were
given multiple types of instructions, precluding our ability to
examine subtypes. Based on the intervention studies, parental
involvement in homework shows a minor effect on achieve-
ment, according to the d index.
Overall Summary
Overall, parental involvement during middle school is positively
related to achievement. However, the types of involvement in which
parents engage matter. Among the types of involvement, parental
involvement that creates an understanding about the purposes, goals,
and meaning of academic performance; communicates expectations
about involvement; and provides strategies that students can effec-
tively use (i.e., academic socialization) has the strongest positive
relation with achievement. Involvement pertaining to homework as-
sistance and supervising or checking homework was the only type of
involvement that was not consistently related with achievement.
Whereas school-based involvement—including visiting the school,
volunteering at school, and attending school events—was moderately
positive in its association with achievement, our evidence suggests
that the most salient type of parental involvement is involvement that
relates to achievement, results in socialization around the goals and
purposes of education, and provides adolescents with useful strategies
that they can use in semiautonomous decision making.
Discussion
In the face of declines in academic achievement during middle
school and increased barriers associated with maintaining parental
involvement with adolescents (who are increasingly autonomous and
independent) and in middle schools (that are larger and more bureau-
cratic), the synthesis of the extant literature confirms that parental
involvement is positively associated with achievement. Moreover,
through this meta-analysis, we identified a specific type of involve-
ment, namely academic socialization, that has the strongest positive
relation with achievement during middle school. School-based in-
volvement was also positively related to achievement, but less
strongly so. Finally, the results for home-based involvement were
mixed. Involvement that entailed assisting with homework was not
consistently associated with achievement, whereas other types of
home-based involvement were positively related to achievement.
Academic socialization includes parents’ communication of
their expectations for achievement and value for education, fos-
tering educational and occupational aspirations in their adoles-
cents, discussing learning strategies with children, and making
preparations and plans for the future, including linking material
discussed in school with students’ interests and goals. An adoles-
cent’s ability to engage in logical and analytic thinking, problem
solving, planning, and decision making increase during adoles-
cence (Halpern-Felsher & Cauffman, 2001; Keating, 2004). Fur-
ther, it is during adolescence that goals, beliefs, and motivations
are internalized and such inner processes shape adolescents’ aca-
demic performance and course selection (Wigfield, Byrnes, &
Eccles, 2006). Academic socialization includes the types of strat-
egies that will scaffold adolescents’ burgeoning autonomy, inde-
pendence, and cognitive abilities. In addition, this type of involve-
ment represents developmentally appropriate strategies of
involvement, as it fosters and builds upon the development of
internalized motivation for achievement, focuses on future plans,
provides a link between school work and future goals and aspira-
tions, and is consistent with the needs of middle school students.
Further, it provides young adolescents with the tools to make
semiautonomous decisions about their academic pursuits.
1
Based on the fixed effects model, which does not generalize to the
broader literature but reflects the current set of studies, the average
weighted correlation was .07 (95% CI ⫽ .04, .11) for African Americans
and .20 (95% CI ⫽ .17, .23) for European Americans. These average
weighted correlations were significantly different from each other (Q (1) ⫽
32.67 p ⬍ .001), suggesting that the relation between parental involvement
and achievement is stronger for European Americans, albeit positive and
significant for both groups.
758
HILL AND TYSON
In addition to being developmentally appropriate for adolescents,
academic socialization strategies are developmentally appropriate for
middle school contexts. One of the largest challenges for middle
school teachers in their attempts to involve parents is the large number
of parents with whom they must develop relationships. Middle school
teachers instruct many more students than elementary school teachers.
Moreover, because students have multiple teachers, it is difficult for
parents to develop productive relationships with their adolescent’s
teachers (Hill & Chao, 2009). Academic socialization as a parental
involvement strategy is adaptive for middle school contexts because it
is not dependent on the development of deep, high-quality relation-
ships with each teacher—a goal that is often not feasible even for the
most motivated teacher. It is dependent on parents’ knowledge about
how to navigate the middle school context, which is information that
can more easily be provided to parents through communications
between the school and home and through electronic communications
(e.g., Bouffard, 2009), and builds upon the relationship between the
adolescent and the parent. This type of involvement can be more
easily solicited by adolescents as they assess their own needs and
direct their interests and trajectories. Further, students’ academic
promise may elicit this level of involvement and planning from
parents.
School-based involvement was also positively related to
achievement, although the relation was weaker than the relation for
academic socialization. Whereas prior research and theory have
demonstrated the positive effect of school-based involvement
(Comer, 1995; Epstein & Sanders, 2002; Grolnick & Slowiaczek,
1994; Hill, 2001; Hill & Craft, 2003; Lareau, 1987), it is possible
that the processes through which school-based involvement has its
effect (e.g., increasing social capital or knowledge) are more
difficult to realize in middle school. School-based involvement
during middle school is less likely to entail involvement directly in
one’s child’s classroom. It is more likely to entail assisting teach-
ers with preparation (e.g., bulletin boards, setting up classrooms),
fundraising, administrative duties in the office, or committee work.
Whereas this type of involvement is important for the functioning
of the school, it often does not directly provide parents with
knowledge about instructional styles and course content that will
facilitate their involvement with their students’ schoolwork. Fur-
ther, because students have multiple teachers in middle school,
parents would need to spend a considerable amount of time at
school to build relations with each teacher and spend time in each
classroom. Finally, as adolescents become more independent, they
do not want their parents to visit the school (Stevenson & Baker,
1987); they want to be trusted that they will manage their respon-
sibilities. That is, adolescents often indicate that they want their
parents’ help but do not want their parents to visit the school
(Collins & Laursen, 2004; Grotevant, 1998). Given adolescents’
increased sense of efficacy, autonomy, and problem-solving skills,
they may have a greater role in soliciting the type of involvement
they need from their parents, which would make active school-
based involvement less effective than other types of involvement.
Home-based involvement entails a range of activities from sup-
porting achievement by providing appropriate structure and intellec-
tually engaging materials in the home to monitoring and checking
homework. The provision of an educationally supportive home envi-
ronment consistently has been shown to be positively related to
achievement (Chao et al., 2009; Reynolds & Gill, 1994). In contrast,
helping with homework has been shown to both accelerate and
interfere with achievement (Cooper, 1989, 2007; Wolf, 1979). The
negative relation may be due to parental interference with students’
autonomy, to excessive parental pressure, or to differences in how
parents and schools present the material. On the contrary, supporting
a student who is having trouble completing or understanding home-
work can deepen and further the student’s understanding of the
material. The meta-analysis of the extant literature demonstrated that,
on the whole, parental assistance with homework is not consistently
associated with achievement. It is plausible that, rather than under-
mining achievement, parental engagement in homework is elicited by
poor school performance, which also results in a negative relation
between homework help and achievement.
Attempts at disentangling the direction of effect are futile with
correlational research. Longitudinal, natural, and experimental de-
signs provide the best context for social scientists to infer causality or
direction of effect (Duncan et al., 2004). The synthesis of interven-
tions designed to increase the amount and quality of parental involve-
ment in homework demonstrated only a weak association between
homework help and achievement, and in some cases a negative effect.
Whereas in some cases parents’ direct involvement in homework may
rescue a failing student, the provision of support and structure that
enable middle school students to function semiautonomously, under-
stand the value and utility of education for their future, and understand
how the knowledge gained at school links to their interests, talents,
and current events seems most significant.
In the context of these consistent findings showing that parental
involvement in education is positively associated with achieve-
ment during middle school (with the exception of homework help),
there are a number of limitations to the existing literature that give us
some pause in the confidence we have in our conclusions and provide
fruitful ground for future research. First, we have attempted to be
careful in our discussion of the findings to refrain from making causal
inferences. Whereas most theories suggest that parental involvement
improves achievement, there is also a growing body of literature that
points to the motivating effect of prior achievement in increasing or
decreasing levels of parental involvement (Eccles, 2007; Hoover-
Dempsey, Ice, & Whitaker, 2009). For example, the negative relation
between parental homework help and achievement may reflect par-
ents’ appropriate response for children who are not performing well,
rather than demonstrating that parental homework help undermines
achievement. Further, adolescents’ increased cognitive abilities, sense
of efficacy, and confidence may result in soliciting advice and in-
volvement from parents, which also impacts our understanding of the
nature of the dynamic relation between involvement and achievement.
Second, the studies included in this meta-analysis reflect incredible
heterogeneity in measurement and study design. Indeed, based on our
review of the literature, there is not a standard measure of involvement
that is used consistently in studies of middle school families. Rarely
does one see the same measure used across studies (Hill & Tyson,
2005). The most consistently used measure is from Steinberg et al.
(1992; five items). Three studies cited it; however, two studies mod-
ified it. Even when researchers used the same national datasets (e.g.,
NELS-88), different items were used to assess parent involvement
across studies. Although such heterogeneity might undermine our
ability to identify consistent patterns in the relation between parental
involvement and achievement, the meta-analysis still points to the
conclusion that parental involvement that reflects academic socializa-
tion has the strongest positive relation with achievement.
759
PARENTAL INVOLVEMENT: A META-ANALYSIS
Finally, the state of the extant literature did not permit a thor-
ough examination of ethnic and socioeconomic variations in in-
volvement and their relation with academic outcomes. The find-
ings suggested that there is no difference in the strength of the
relation between involvement and achievement for European
Americans compared to African Americans when considering the
findings from the random effects models, which extrapolate to the
broader literature (i.e., random-effects design). However, the fixed
effect models demonstrated that the relation was positive for both
African American and Euro-American families, but stronger for
European Americans. Some research suggests that parental in-
volvement has different meanings and motivations across ethnicity
(Hill & Craft, 2003; Lynch & Stein, 1987), and those from varying
economic background engage in parental involvement with differ-
ent levels of social capital (i.e., resources, knowledge; Hill et al.,
2004; Lareau, 2003; Lareau & Horvat, 1999). It is possible that
seemingly ethnic differences are ultimately the result of differ-
ences in economic resources. Thus, ethnic differences found in the
fixed effects model may be due to the potential confounding of
ethnicity and other contextual factors in the studies in this meta-
analysis. Supporting this contention, Jeynes’s (2005) meta-
analysis found no statistical differences in the strength of the
relation between studies reflecting “mostly ethnic minority” sam-
ples and Euro-American samples when socioeconomic indicators
were controlled. Similarly, Fan and Chen (2001) did not find
ethnicity to be a significant moderator in their meta-analysis. It is
also possible that ethnic differences in beliefs, practices, and
processes are not related to involvement as defined in this study.
For example, prior research has found that African American
parents’ involvement has entailed monitoring the school and teach-
ers rather than forming partnerships with them (Lareau & Horvat,
1999), and African American parents of high achievers have
indicated that they are involved at school, in part, to demonstrate
to school personnel their commitment to education.
Other than with African Americans, the body of literature on
parental involvement in middle school does not include sufficient
studies of other sizable ethnic groups, such as Latinos or Asian
Americans. Although there is evidence that Asian American stu-
dents have the highest average achievement levels, their parents
are the least involved in education as defined by the prevailing
theories (Chao, 2000). Given current demographic trends that
predict that Latinos will become the largest ethnic minority group
in the United States, it is imperative that psychologists conduct
research to understand how Latino families and schools work
together most productively. In addition, it is important to identify
the types of involvement strategies used by Asian American fam-
ilies. This is particularly important because academic socialization
as a parental involvement strategy is more dependent on parents’
knowledge and resources and schools’ ability to provide such
information to parents than are other types of involvement.
In the current policy climate— one that requires schools to
maintain policies and support parental involvement in educa-
tion—it is imperative that the scientific field identify developmen-
tally appropriate practical strategies for middle schools. Although
the NCLB Act (2002) requires parental involvement in education,
largely defined as accountability and communication between
families and schools, the results of this meta-analysis suggest that
programs and policies need to consider a broader range of involve-
ment strategies. In their mandates, policies such as the NCLB
should carefully consider the specific needs of middle school
students, including the provision of information about tracking and
placement as it effects college access, the ways in which curricu-
lum can be linked to students’ interests and current events, and
linkages between the middle school curriculum and students’ long
term goals. Lack of guidance was the primary reason that academ-
ically able students did not attend postsecondary institutions after
high school (Catsambis & Garland, 1997; Conners & Epstein,
1995; Jordan & Plank, 2000). In the current context of increased
demand for parental involvement in education (e.g., school choice,
tracking, course selection), without effective parental involvement,
adolescents’ opportunities are often foreclosed, leading to lost
potential, unrealized talent, diminished educational and vocational
attainment, and widening demographic gaps in achievement.
References
References marked with an asterisk indicate studies included in the
meta-analysis.
Adams, G. R., & Berzonsky, M. D. (Eds.). (2003). Blackwell handbook of
adolescence. Malden, MA: Blackwell.
*Baker, D. P., & Stevenson, D. L. (1986). Mothers’ strategies for chil-
dren’s school achievement: Managing the transition to high school.
Sociology of Education, 59, 156–166.
*Balli, S. J., Demo, D. H., & Wedman, J. F. (1998). Family involvement
with children’s homework: An intervention in the middle grades. Family
Relations, 47, 149–157.
*Balli, S. J., Wedman, J. F., & Demo, D. H. (1997). Family involvement
with middle-grades homework: Effects of differential prompting. Jour-
nal of Experimental Education, 66, 31–48.
*Bandura, A., Barbaranelli, C., Caprara, G. V., & Pastorelli, C. (1996).
Multifaceted impact of self-efficacy beliefs on academic functioning.
Child Development, 67, 1206–1222.
Barber, B. K., & Olsen, J. A. (2004). Assessing the transition to middle and
high school. Journal of Adolescent Research, 19, 3–30.
Bleeker, M. M., & Jacobs, J. E. (2004). Achievement in math and science:
Do mothers’ beliefs matter 12 years later? Journal of Educational
Psychology, 96, 97–109.
Bouffard, S. M. (2009). Tapping into technology: Using the Internet to
promote family–school communication. In N. E. Hill & R. K. Chao
(Eds.), Families, schools, and the adolescent: Connecting research,
policy, and practice (pp. 147–161). New York: Teachers College Press.
*Bronstein, P., Ginsburg, G. S., & Herrera, I. S. (2005). Parental predictors
of motivational orientation in early adolescence: A longitudinal study.
Journal of Youth and Adolescence, 34, 559–575.
Byrnes, J. P., Miller, D. C., & Reynolds, M. (1999). Learning to make good
decisions: A self-regulation perspective. Child Development, 70, 1121–1140.
Catsambis, S. (2001). Expanding knowledge of parental involvement in
children’s secondary education: Connections with high school seniors’
academic success. Social Psychology of Education, 5, 149–177.
Catsambis, S., & Garland, J. E. (1997). Parental involvement in students’
education during middle and high school (Report No. 18). Baltimore,
MD: Johns Hopkins University, Center for Research on the Education of
Students Placed at Risk.
Chao, R. K. (2000). Cultural explanations for the role of parenting in the
school success of Asian American children. In R. W. Taylor & M. C.
Wang (Eds.), Resilience across contexts: Family, work, culture, and
community (pp. 333–363). Mahwah, NJ: Erlbaum.
Chao, R. K., Kanatsu, A., Stanoff, N., Padmawidjaja, I., & Aque, C.
(2009). Diversities in meaning and practice: The parental involvement of
Asian immigrants. In N. E. Hill & R. K. Chao (Eds.), Families, schools,
and the adolescent: Connecting research, policy, and practice (pp.
110 –125). New York: Teachers College Press.
760
HILL AND TYSON
Collins, W. A., & Laursen, B. (2004). Parent–adolescent relationships and
influences. In R. M. Lerner & L. Steinberg (Eds.), Handbook of ado-
lescent psychology (2nd ed., pp. 331–362). Hoboken, NJ: Wiley.
Comer, J. P. (1995). School power: Implications of an intervention project.
New York: Free Press.
Conners, L. J., & Epstein, J. L. (1995). Parents and school partnerships. In
M. H. Bornstein (Ed.), Handbook of parenting (Vol. 4, pp. 437– 458).
Hillsdale, NJ: Erlbaum.
Cooper, H. (1989). Homework. New York: Longman.
Cooper, H. (1998). The data analysis stage. In H. Cooper (Ed.), Synthe-
sizing research (Vol. 2, pp. 104–156). Thousand Oaks, CA: Sage
Publications.
Cooper, H. (2007). The battle over homework: Common ground for ad-
ministrators, teachers, and parents. Thousand Oaks, CA: Corwin.
Cross, W. E. (2003). Tracing the historical origins of youth delinquency
and violence: Myths and realities about black culture. Journal of Social
Issues, 59, 67–82.
Dauber, S. L., & Epstein, J. L. (1989). Parents’ attitudes and practices of
involvement in inner-city elementary and middle schools (Report No.
33). Baltimore, MD: Johns Hopkins University, Center for Research on
Elementary and Middle Schools.
Dauber, S. L., & Epstein, J. L. (1993). Parents’ attitudes and practices of
involvement in inner-city elementary and middle schools. In N. F.
Chavkin (Ed.), Families and schools in a pluralistic soceity (pp. 53–71).
Albany, NY: State University of New York.
Dearing, E., Kreider, H., Simpkins, S., & Weiss, H. B. (2006). Family
involvement in school and low-income children’s literacy: Longitudinal
association between and within families. Journal of Educational Psy-
chology, 98, 653–664.
*Desimone, L. (1999). Linking parent involvement with student achieve-
ment: Do race and income matter? Journal of Educational Research, 93,
11–30.
*Deutscher, R. & Ibe, M. (n.d.). Relationships between parental involve-
ment and children’s motivation. Retrieved in May 2007 from the Lewis
Center for Educational Research website, http://www.lewiscenter.org/
research/relationships.pdf
Dornbusch, S. M., & Glasgow, K. L. (1996). The structural context of
family–school relations. In A. Booth & J. F. Dunn (Eds.), Family–school
links: How to affect educational outcomes (pp. 35– 44). Mahwah, NJ:
Erlbaum.
*Driessen, G., Smit, F., & Sleegers, P. (2005). Parental involvement and
educational achievement. British Educational Research Journal, 31,
509 –532.
Duncan, G. J., Magnuson, K. A., & Ludwig, J. (2004). The endogeneity
problem in developmental studies. Research in Human Development,
1(1 & 2), 59–80.
Duval, S. (2005). The trim and fill method. In H. R. Rothstein, A. J. Sutton,
& M. Borenstein (Eds.), Publication bias in meta-analysis: Prevention,
assessment and adjustments (pp. 127–144). Hoboken, NJ: Wiley.
*Eamon, M. K. (2005). Social-demographic, school, neighborhood, and
parenting influences on the academic achievement of Latino young
adolescents. Journal of Youth and Adolescence, 34, 163–174.
Eccles, J. S. (2004). Schools, academic motivation, and stage– environment
fit. In R. M. Lerner & L. D. Steinberg (Eds.), Handbook of adolescent
psychology (2nd ed., pp. 125–153). New York: Wiley.
Eccles, J. S. (2007). Families, schools, and development achievement-
related motivations and engagement. In J. E. Grusec & P. D. Hastings
(Eds.), Handbook of socialization: Theory and research (pp. 665– 691).
New York: Guilford.
Eccles, J. S., & Harold, R. D. (1993). Parent–school involvement during
the early adolescent years. Teachers College Record, 94, 568 –587.
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 outcomes (pp. 3–34). Mahwah, NJ:
Erlbaum.
Epstein, J. L. (1987). Toward a theory of family–school connections:
Teacher practices and parent involvement. In K. Hurrelman, F. X.
Kaufman, & F. Losel (Eds.), Social intervention: Potential and con-
straints (pp. 121–136). Berlin, Germany: de Gruyer.
Epstein, J. L. (2001). School, family, and community partnerships: Pre-
paring educators and improving schools. Boulder, CO: Westview.
Epstein, J. L., Salinas, K. C., & Jackson, V. E. (1995). Manual for teachers
and prototype activities: Teachers Involve Parents in Schoolwork (TIPS)
language arts, science/health, and mathematics interactive homework in
the middle grades (Rev. ed.). Baltimore: Johns Hopkins University,
Center on School, Family, and Community Partnerships.
Epstein, J. L., & Sanders, M. G. (2002). Family, school, and community
partnerships. In M. H. Bornstein (Ed.), Handbook of parenting: Vol. 5.
Practical issues in parenting (pp. 507–437). Mahwah, NJ: Erlbaum.
Fan, X., & Chen, M. (2001). Parental involvement and students’ academic
achievement: A meta-analysis. Educational Psychology Review, 13(1),
1–22.
*Garcia Bacete, F. J., & Ramirez, J. R. (2001). Family and personal
correlates of academic achievement. Psychological Reports, 88, 533–
547.
*Grolnick, W. S., Kurowski, C. O., Dunlap, K. G., & Hevey, C. (2000).
Parental resources and the transition to junior high. Journal of Research
on Adolescence, 10, 465–488.
Grolnick, W. S., Price, C. E., Beiswenger, K. L., & Sauck, C. C. (2007).
Evaluative pressure in mothers: Effects of situation, maternal, and child
characteristics on autonomy supportive versus controlling behavior.
Developmental Psychology, 43, 991–1002.
*Grolnick, W. S., & Slowiaczek, M. L. (1994). Parents’ involvement in
children’s schooling: A multidimensional conceptualization and moti-
vational model. Child Development, 65, 237–252.
Grotevant, H. S. (1998). Adolescent development in family contexts. In W.
Damon & N. Eisenberg (Eds.), Handbook of child psychology: Vol. 3.
Social, emotional, and personality development (5th ed., pp. 1097–
1150). New York: Wiley.
*Gutman, L. M., & Eccles, J. S. (1999). Financial strain, parenting behav-
iors, and adolescents’ achievement: Testing model equivalence between
African American and European American single- and two-parent fam-
ilies. Child Development, 70, 1464–1476.
*Gutman, L. M., & Midgley, C. (2000). The role of protective factors in
supporting the academic achievement of poor African American stu-
dents during the middle school transition. Journal of Youth and Adoles-
cence, 20, 223–248.
*Gutman, L. M., Sameroff, A. S., & Eccles, J. S. (2002). The academic
achievement of African American students during early adolescence: An
examination of risk, promotive, and protective factors. American Jour-
nal of Community Psychology, 30, 376–399.
Halpern-Felsher, B. L., & Cauffman, E. (2001). Costs and benefits of a
decision: Decision-making competence in adolescents and adults. Jour-
nal of Applied Developmental Psychology, 22, 257–276.
Hampton, F. M., Mumford, D. A., & Bond, L. (1998). Parental involve-
ment in inner city schools: The Project FAST extended family approach
to success. Urban Education, 33, 410–427.
*Hao, L., & Bonstead-Bruns, M. (1998). Parent– child differences in edu-
cational expectations and the academic achievement of immigrant and
native students. Sociology of Education, 71, 175–198.
Hara, S. R. (1998). Parent involvement: The key to improved student
achievement. School Community Journal, 8(2), 9–19.
*Harris, A. (1990). Students’ perceptions and attitudes toward parent
involvement in academic homework and its relationship to academic
achievement. (Doctoral dissertation, Florida State University). Disser-
tation Abstracts International, 51, p. 4018.
*Hawes, C. A., & Plourde, L. A. (2005). Parental involvement and its
761
PARENTAL INVOLVEMENT: A META-ANALYSIS
influence on the reading achievement of 6th grade students. Reading
Improvement, 42, 47–57.
Hill, N. E. (2001). Parenting and academic socialization as they relate to
school readiness: The role of ethnicity and family income. Journal of
Educational Psychology, 93, 686–697.
*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 aspirations: Demographic
variations across adolescence. Child Development, 75, 1491–1509.
Hill, N. E. & Chao, R. K. (2009). Families, schools, and the adolescent:
Connecting research, policy, and practice. New York: Teachers College
Press.
Hill, N. E., & Craft, S. A. (2003). Parent–school involvement and school
performance: Mediated pathways among socioeconomically comparable
African American and Euro-American families. Journal of Educational
Psychology, 95, 74–83.
Hill, N. E., & Taylor, L. C. (2004). Parental school involvement and
children’s academic achievement: Pragmatics and issues. Current Di-
rections in Psychological Science, 13, 161–164.
Hill, N. E., & Tyson, D. F. (2005, February). Family–school involvement:
Developmental and demographic variations during the transition to
middle school. Paper presented at the Carolina Consortium on Human
Development: Series on School Transitions, Chapel Hill, NC.
Hill, N. E., Tyson, D. F., & Bromell, L. (2009). Parental involvement in
middle school: Developmentally appropriate strategies across SES and
ethnicity. In N. E. Hill & R. K. Chao (Eds.), Families, schools, and the
adolescent: Connecting research, policy, and practice (pp. 53–72). New
York: Teachers College Press.
Hoover-Dempsey, K. V., Ice, C. L., & Whitaker, M. C. (2009). “We’re
way past reading together”: Why and how parental involvement in
adolescence makes sense. In N. E. Hill & R. K. Chao (Eds.), Families,
schools, and the adolescent: Connecting research, policy, and practice
(pp. 19–36). New York: Teachers College Press.
Hoover-Dempsey, K. V., & Sandler, H. M. (1995). Parent involvement in
children’s education: Why does it make a difference. Teachers’ College
Record, 97, 310–331.
Hughes, D., Rodriguez, J., Smith, E. P., Johnson, D. J., Stevenson, H. C.,
& Spicer, P. (2006). Parents’ ethnic–racial socialization practices: A
review of research and directions for future study. Developmental Psy-
chology, 42, 747–770.
*Hughes, D. (2004). [Data from the Center on Research, Culture, Devel-
opment and Education on a multiethnic sample of middle school stu-
dents]. Unpublished raw data.
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 involve-
ment to urban elementary school student academic achievement. Urban
Education, 40, 237–269.
Jeynes, W. H. (2007). The relationship between parental involvement and
urban secondary school student academic achievement: A meta-analysis.
Urban Education, 42, 82–110.
*Jodl, K. M., Michael, A., Malanchuk, O., Eccles, J. S., & Sameroff, A.
(2001). Parents’ roles in shaping early adolescents’ occupational aspi-
rations. Child Development, 72, 1247–1265.
Jones, N. & Schneider, B. (2009). Rethinking the role of parenting:
Promoting adolescent academic success and emotional well-being. In
N. E. Hill & R. K. Chao (Eds.), Families, schools, and the adolescent:
Connecting research, policy, and practice (pp. 73–90). New York:
Teachers College Press.
Jordan, W. J., & Plank, S. B. (2000). Talent loss among high-achieving
poor students. In M. G. Sanders (Ed.), Schooling students placed at risk:
Research policy and practice in the education of poor and minority
adolescents (pp. 83–108). Mahwah, NJ: Erlbaum.
*Juang, L. P., & Silbereisen, R. K. (2002). The relationship between
adolescent academic capability beliefs, parenting and school grades.
Journal of Adolescence, 25, 3–18.
Keating, D. P. (2004). Cognitive and brain development. In R. M. Lerner
& L. Steinberg (Eds.), Handbook of adolescent psychology (2nd ed., pp.
45– 84). Hoboken, NJ: Wiley.
*Keisner, J. W. (1997). The effects of a parental homework monitoring
intervention on school engagement of high risk middle school students.
(Doctoral dissertation, University of Oregon, 1997). Dissertation Ab-
stracts International, 58, 0980.
*Keith, P. B., & Lichtman, M. V. (1994). Does parental involvement
influence the academic achievement of Mexican-American eighth grad-
ers? Results from the National Education Longitudinal Study. School
Psychology Quarterly, 9, 256–273.
*Keith, T. Z., Keith, P. B., Troutman, G. C., & Bickley, P. G. (1993). Does
parental involvement affect eighth-grade student achievement? Struc-
tural analysis of national data. School Psychology Review, 22, 474 – 496.
*Kelly, S. (2004). Do increased levels of parental involvement account for
social class differences in track placement? Social Science Research, 33,
626 – 659.
*Kim, E. (2002). The relationship between parental involvement and
children’s educational achievement in the Korean immigrant family.
Journal of Comparative Family Studies, 33, 529–563.
*Kim, K., & Rohner, R. P. (2002). Parental warmth, control, and involve-
ment in schooling: Predicting academic achievement among Korean
American adolescents. Journal of Cross-Cultural Psychology, 33, 127–
140.
Kohl, G. O., Lengua, L. J., McMahon, R. J., & the Conduct Problems
Prevention Research Group. (2000). Parent involvement in school: Con-
ceptualizing multiple dimensions and their relations with family and
demographic risk factors. Journal of School Psychology, 38, 501–538.
Lareau, A. (1987). Social class differences in family–school relationships:
The importance of cultural capital. Sociology of Education, 60, 73–85.
Lareau, A. (2003). Unequal childhoods: Class, race and family life. Berke-
ley, CA: University of California.
Lareau, A., & Horvat, E. M. (1999). Moments of social inclusion and
exclusion: Race, class, and cultural capital in family–school relation-
ships. Sociology of Education, 72, 37–53.
*Latendresse, S. (2005). Perceptions of parenting among affluent youth:
Exploring antecedents of middle school adjustment trajectories. [Doc-
toral dissertation]. Dissertation Abstracts International, Session B: The
Sciences and Engineering,, 65, 3746.
Lerner, R. M., & Steinberg, L. D. (Eds.). (2004). Handbook of adolescent
psychology (2nd ed.). New York: Wiley.
Lynch, E. W., & Stein, R. C. (1987). Parent participation by ethnicity: A
comparison of Hispanic, Black and Anglo families. Exceptional Chil-
dren, 54(2), 105–111.
*Ma, X. (1999). Dropping out of advanced mathematics: The effects of
parental involvement. Teachers College Record, 101, 60–81.
*Marchant, G. J., Paulson, S. E., & Rothlisberg, B. A. (2001). Relations of
middle school students’ perceptions of family and school contexts with
academic achievement. Psychology in the Schools, 38, 505–519.
*McNeal, R. B. (1999). Parental involvement as social capital: Differential
effectiveness on science achievement, truancy, and dropping out. Social
Forces, 78, 117–144.
*Melby, J. N., & Conger, R. D. (1996). Parental behaviors and adolescent
academic performance: A longitudinal analysis. Journal of Research on
Adolescence, 6, 113–137.
Miller, J. D., Kimmel, L., Hoffer, T., & Nelson, C. (2001). Longitudinal
study of American youth: User’s manual. Chicago: International Center
for the Advancement of Scientific Literacy at Northwestern University.
*Muller, C. (1995). Maternal employment, parent involvement, and math-
ematics achievement among adolescents. Journal of Marriage & the
Family, 57, 85–100.
762
HILL AND TYSON
National Education Longitudinal Study of 1988 [Data file]. Washington,
DC: National Center for Education Statistics.
National Longitudinal Survey of Youth, 1997 [Data file]. Washington, DC:
Bureau of Labor Statistics.
No Child Left Behind Act of 2001, Pub. L. No. 107–110, 114 Stat. 1425
(2002). Retrieved February 26, 2009, from http://www.ed.gov/policy/
elsec/leg/esea02/107–110.pdf
Ogbu, J. U. (1978). Minority education and caste: The American system in
cross-cultural perspective. New York: Academic Press.
*Pelegrina, S., Garcı´a-Linares, M., & Casanovam, P. F. (2003). Adoles-
cents and their parents’ perceptions about parenting characteristics. Who
can better predict the adolescent’s academic competence? Journal of
Adolescence, 26, 651–665.
*Peng, S. S., & Wright, D. (1994). Explanation of academic achievement
of Asian American students. Journal of Educational Research, 87,
346 –352.
Prescott, B. L., Pelton, C. L., & Dornbusch, S. M. (1986). Teacher
perceptions of parent–school communication: A collaborative analysis.
Teacher Education Quarterly, 13, 67–83.
*Reynolds, A. J., & Gill, S. (1994). The role of parental perspectives in the
school adjustment of inner-city Black children. Journal of Youth and
Adolescence, 23, 671–695.
*Rillero, P., & Helgeson, S. L. (1995, April). An evaluation of the use of
hands-on science homework assignments by sixth grade students and
their parents. Paper presented at the annual meeting of the National
Association for Research in Science Teaching, San Francisco, CA.
Rosenthal, R. (1991). Meta-analytic procedures for social research (Rev.
ed.). Newbury Park, CA: Sage.
*Salazar, L. P., Schuldermann, S. M., Schuldermann, E. H., & Huynh,
C. M. (2000). Filipino adolescents’ parental socialization for academic
achievement in the United States. Journal of Adolescent Research, 15,
564 –586.
Sanders, M. G., & Epstein, J. L. (2000). Building school–family–
community partnerships in middle and high schools. In M. G. Sanders
(Ed.), Schooling students placed at risk: Research, policy, and practice
in the education of poor and minority adolescents (pp. 339 –361).
Mahwah, NJ: Erlbaum.
Seginer, R. (2006). Parents’ educational involvement: A developmental
ecological perspective. Parenting: Science and Practice, 6, 1–48.
*Seidman, E., Allen, L., Aber, J. L., Mitchell, C., & Feinman, J. (1994).
The impact of school transitions in early adolescence on the self-system
and perceived social context of poor urban youth. Child Development,
65, 507–522.
*Seidman, E., Lambert, L. E., Allen, L., & Aber, J. L. (2003). Urban
adolescents’ transition to junior high school and protective family trans-
actions. Journal of Early Adolescence, 23, 166–193.
*Seyfried, S. F., & Chung, I. J. (2002). Parent involvement as parental
monitoring of student motivation and parent expectations predicting
later achievement among African American and European American
middle school age students. Journal of Ethnic & Cultural Diversity in
Social Work, 11, 109–131.
*Shumow, L., & Lomax, R. (2002). Parental efficacy: Predictor of parent-
ing behavior and adolescent outcomes. Parenting: Science and Practice,
2, 127–150.
*Shumow, L., & Miller, J. D. (2001). Parents’ at-home and at-school
academic involvement with young adolescents. Journal of Early Ado-
lescence, 21, 68–91.
*Singh, K., Bickley, P, Trivette, P., Keith, T. Z., Keith, P. B., & Anderson,
E. (1995). The effects of four components of parental involvement on
eighth grade student achievement: Structural analysis of NELS-88 data.
School Psychology Review, 24, 299–317.
*Sirin, S. R., & Rogers-Sirin, L. (2004). Exploring school engagement of
middle-class African American adolescents. Youth &