Evidence-Base d Paren t Involvemen t Interventions wit h...
32
School Psychology Quarterly, Vol. 20, No. 4, 2005, pp. 371-402 Evidence-Based Parent Involvement Interventions with School-Aged Children Maria Fishel and Lucila Ramirez University of Texas at Austin This paper reviewed 24 studies of parent involvement for school-aged children conducted between 1980 and 2002 and evaluated them according to the criteria developed by the Task Force on Evidence-Based Interventions in School Psychology. The parent involve- ment component of all studies had parents helping children learn at home, with most tar- geting a change in academic performance, including reading skills, mathematics skills, spelling, and homework completion. Results yielded a wide range of treatment effective- ness. The strongest evidence for parent involvement was provided for programs that im- plemented parent tutoring in the home and targeted a single academic problem of the ele- mentary school-aged child, primarily reading and mathematics skills. Despite promising evidence for the effectiveness of parent home tutoring, it was concluded that the evidence base for the effectiveness of parent involvement as an intervention for children's aca- demic problems is inconclusive due to methodological weaknesses in the studies re- viewed. Recommendations for future empirical research are provided. This article reviewed and evaluated parent involvement interventions with school-aged children according to the set of comprehensive criteria proposed as best practices by the American Psychological Association's Division 16 Task Force on Evidence-Based Interventions in School Psychology (hereafter referred to as Task Force) (Division 16 and Society for the Study of School Psychology Task Force, 2003). Parent involvement1 generally refers to the participation of significant caregivers (including parents, grandparents, stepparents, foster par- ents, etc.) in the educational process of their children in order to promote their academic and social well-being (Wolfendale, 1983). For most of the 20th cen- tury, American schools were considered solely responsible for children's educa- 1 Parent involvement and parent participation are used interchangeably throughout this paper. A version of this paper was presented at the annual meetings of the American Psychological Associ- ation in San Francisco, August, 2001 and Honolulu, August 2004. Address correspondence to Maria Fishel, Department of Educational Psychology, 1 University Station, Mail Station D5800, The University of Texas at Austin, Austin, TX 78712-0383; E-mail: [email protected]. 371
Transcript of Evidence-Base d Paren t Involvemen t Interventions wit h...
School Psychology Quarterly, Vol. 20, No. 4, 2005, pp. 371-402
Evidence-Based Parent Involvement Interventionswith School-Aged Children
Maria Fishel and Lucila RamirezUniversity of Texas at Austin
This paper reviewed 24 studies of parent involvement for school-aged children conductedbetween 1980 and 2002 and evaluated them according to the criteria developed by theTask Force on Evidence-Based Interventions in School Psychology. The parent involve-ment component of all studies had parents helping children learn at home, with most tar-geting a change in academic performance, including reading skills, mathematics skills,spelling, and homework completion. Results yielded a wide range of treatment effective-ness. The strongest evidence for parent involvement was provided for programs that im-plemented parent tutoring in the home and targeted a single academic problem of the ele-mentary school-aged child, primarily reading and mathematics skills. Despite promisingevidence for the effectiveness of parent home tutoring, it was concluded that the evidencebase for the effectiveness of parent involvement as an intervention for children's aca-demic problems is inconclusive due to methodological weaknesses in the studies re-viewed. Recommendations for future empirical research are provided.
This article reviewed and evaluated parent involvement interventions withschool-aged children according to the set of comprehensive criteria proposed asbest practices by the American Psychological Association's Division 16 TaskForce on Evidence-Based Interventions in School Psychology (hereafter referredto as Task Force) (Division 16 and Society for the Study of School PsychologyTask Force, 2003). Parent involvement1 generally refers to the participation ofsignificant caregivers (including parents, grandparents, stepparents, foster par-ents, etc.) in the educational process of their children in order to promote theiracademic and social well-being (Wolfendale, 1983). For most of the 20th cen-tury, American schools were considered solely responsible for children's educa-
1 Parent involvement and parent participation are used interchangeably throughout this paper.
A version of this paper was presented at the annual meetings of the American Psychological Associ-ation in San Francisco, August, 2001 and Honolulu, August 2004.
Address correspondence to Maria Fishel, Department of Educational Psychology, 1 UniversityStation, Mail Station D5800, The University of Texas at Austin, Austin, TX 78712-0383; E-mail:[email protected].
371
372 FISHEL AND RAMIREZ
tion, and parent involvement was ignored or downplayed by educators and re-searchers (Zellman & Waterman, 1998). Reforms to increase academic achieve-ment that focused exclusively on the school or classroom, however, have hadlimited success (Christenson, Hurley, Sheridan, & Fenstermacher, 1997). De-clines in the educational outcomes of students, in combination with significantchanges in the social demographics of the family, raised the possibility that edu-cational deficits were related to factors in the home environment. This perspec-tive has made parent involvement a priority in current national educational andsocial policy (Zellman & Waterman, 1998).
Advocacy for parent involvement in education is intrinsic to numerous federalinitiatives, beginning in the 1960s with Head Start, and reflected today in the NoChild Left Behind Act of 2001 (NCLB). Head Start provided educational inter-ventions during the preschool years for economically disadvantaged childrenthat included a broad parent component. Other federal projects promoting parentparticipation followed, including Title I of the Elementary and Secondary Edu-cation Act in 1965 and Project Follow Through in 1968 (Doernberger & Zigler,1993). Title I broadened parental roles by mandating increased consultation andcollaboration with parents (Arroyo & Zigler, 1993). Project Follow Through waseffective at increasing parent participation in tutoring, volunteering, school gov-ernance, and parent education, but funding cuts undermined its initial success(Zigler & Styfco, 1993). Judicial support for the involvement of parents in theeducation of their children came in the 1970s and 1980s with the passage of thefederal statute Public Law (PL) 94-142 (also known as Individuals with Disabil-ities Education Act, or IDEA) and the Education of the Handicapped Amend-ments of 1986 (PL 99-457). More recently, we have witnessed a consensus inpolicies on the local, state, and federal levels regarding the benefits of parentparticipation in education (Chrispeels, 1996; Mattingly, Prislin, McKenzie, Ro-driguez, & Kayzar, 2002). The reauthorization of Title I by Congress in 1994makes it clear that parent involvement at the state, district, and school levels isnow viewed as crucial to student success. Increasing parent involvement in pro-moting children's academic, social, and emotional development was also recog-nized as one of the objectives included in Goals 2000: Educate America Act(U.S. Department of Education, 1996). Most recently, Section 1118 of theNCLB Act of 2001 (U.S. Department of Education, 2002) requires each schooldistrict that receives Title I funds to implement programs, activities, and proce-dures for the involvement of parents with participating children, including thosewith limited English proficiency, disabilities, and migrant children. In sum, nu-merous federal legislative initiatives, based on the assumption that parents are animportant contributor to children's academic success and social well-being atschool, have mandated the implementation of parent involvement programs andprocedures (Christenson, Rounds, & Gorney, 1992; Wolfendale, 1983).Notwithstanding the considerable research that confirms the important roleplayed by parents in the school-related success of children, the question remains:Are parent involvement programs effective in changing parents' behavior such
PARENT INVOLVEMENT 373
that children's performance at school is positively affected? The purpose of thisreview is to answer this question.
DEFINING PARENT INVOLVEMENT
The definition of parent involvement has changed throughout the years from anexclusive focus on specific activities and roles played by caregivers to an inclu-sive emphasis on a wide range of parent activities that support children's learn-ing. Specific activities defined as parent involvement in early studies includedsupport with homework, school-home notes, school-based parent workshopswith few ties to curriculum, as well as encouragement of parents to "join thePTA, provide merchandise for the bake sale, and show up at times specified bythe school" (Chrispeels, 1996; Zellman & Waterman, 1998, p. 370). The mostwidely cited contemporary definition of parent involvement is one based on a ty-pology proposed by Joyce Epstein and her colleagues (Epstein, 1987; Epstein,1995). This classification consists of six categories, including (1) parenting (i.e.,parents' responsibility to provide for children's basic needs of food, shelter,emotional support, etc., throughout their developmental years), (2) communicat-ing (i.e., parents and school staying in contact), (3) learning at home (i.e., prac-tices occurring at home in which parents interact, monitor, or assist their childrenin educationally related activities), (4) volunteering and/or attending (i.e., all ac-tivities in which the parents come to the school setting to either help or support),(5) decision making (i.e., parents participating in parent-teacher organizationsand school advisory or governance), and (6) community connections (parentscollaborating with community and other outside agencies to facilitate students'education). Epstein's typology owes its popularity to the ease with which ittranslates into the range of parent activities that can be implemented in theschools (Bauch, 1994).
Although many research studies continue to use Epstein's activity-based cate-gories, rival perspectives on the construct of parent involvement have emerged.One such conceptualization views parent involvement systemically, as a home-school-community partnership (e.g., Chrispeels, 1996; Comer & Haynes, 1991;Smith, Connelly, Sizer, & Norman, 1997) that implies reciprocal interactions be-tween the individual, family, and community. Christenson (1995), in contrast,views parent involvement and home-school partnership as distinct. Whereas thegoals of the parents and the schools are mutually agreed upon and responsibili-ties are shared in home-school partnership, in parent involvement, schools andparents are often unequal partners working toward a common goal because par-ent participation is initiated or directed by the school. Thus, according to Chris-tenson (1995), parent involvement is a one-way flow of information. Yet anotherchallenge to Epstein's typology-based definition of parent involvement has beenposed by those who argue that parent involvement is a multidimensional variable(e.g., Grolnick & Slowiaczek, 1994), which includes a varying number of behav-ioral, personal, and intellectual components. These components could have a di-
374 FISHEL AND RAMIREZ
rect or moderating effect on the student outcomes. Although the definition ofparent involvement continues to evolve, the majority of research studies on theeffects of parent involvement programs either use or are consistent with the ac-tivity-based typological definition proposed by Epstein (1987, 1995). Thus, thisreview uses Epstein's typology to define parent involvement. Consistent withthe distinction between parent involvement and home-school partnership/collab-oration made by Christenson (1995), this review is limited to parent involvementprograms. Studies involving home-school collaboration were examined else-where in this issue (see article by D. Cox).
PARENT INVOLVEMENT: BRIEF REVIEW OF EVIDENCE
The benefits of parent involvement in education have been the focus of researchfor several decades (Christenson et al., 1997; Grolnick & Slowiaczek, 1994).The importance of parental involvement in education was underscored in the re-search of Stevenson and Stigler (1992) who found that differences between theachievement of Asian and U.S. students were related to the more active maternalinvolvement in education of the former (Zellman & Waterman, 1998). Parent in-volvement studies target primarily a change in academic achievement, and edu-cational researchers tend to focus on a single specific parent involvement activ-ity at a time (e.g., helping children with homework, frequency of family-schoolcontacts, or participation in school activities and functions) (Grolnick &Slowiaczek, 1994).
Similarly, most reviews of parent involvement have focused on a subset ofparent involvement behaviors. Toomey (1993) reviewed over 40 mostly Britishand Australasian studies, with and without a control group, in which parents lis-tened to their children read at home. He concluded that studies with an explicit"parent training" component (where parents not only received explanation andmodeled appropriate behaviors, but also were monitored and received guidedpractice) were more successful than studies without parent training. Miller andKelley (1991), when examining the body of research on parent involvement inhomework, found no consistent support for a positive association between parentparticipation in homework and academic achievement. In contrast, a more recentreview of this literature by Hoover-Dempsey and colleagues (2001) concludedthat parent involvement in homework was positively related to student achieve-ment, although the authors noted that the influence may be mostly indirect, viamoderating variables. Bempechat's (1992) review of literature examined de-scriptive and correlational studies in several areas of parent involvement, includ-ing socialization practices, parent education, and parent involvement programs.She concluded that parent involvement is positively associated with children'sacademic performance.
Overall, the effects of parent involvement on children's academic perform-ance have been inconclusive, with some research studies yielding results sup-porting the beneficial role of parent involvement (e.g., Christenson et al., 1992;
PARENT INVOLVEMENT 375
Epstein, 1991; Keith, et al., 1993; Shaver & Walls, 1998; Zellman & Waterman,1998), whereas the results of other studies are less promising (e.g., Keith,Reimers, Fehrman, Pottebaum, & Aubey, 1986; Natriello & McDill, 1986). Un-fortunately, the parent involvement literature is characterized by a prevalence ofdescriptive and nonexperimental studies, many with archival data, which haveused correlational analytic methods (e.g., Epstein, 1991; Falbo, Lein, & Amador,2001; Keith et al., 1993; Zellman & Waterman, 1998). Despite methodologicalweaknesses in the literature, most researchers and reviewers of research tend toconcur that parent involvement is associated with achievement gains for students(e.g., Bempechat, 1992; Zellman & Waterman, 1998).
The most comprehensive evaluation of the parent involvement literature todate was recently completed by Mattingly and colleagues (2002). The authorsanalyzed the effectiveness of 41 parent involvement programs that included anevaluation of study characteristics, research design, significant outcomes, anddata analytic methods. Mattingly et al. found insufficient empirical evidence forthe positive effect of parent involvement on either the academic or social well-being of children. Because the Mattingly et al. review of the parent involvementliterature bears a close resemblance to the current review, a clarification of thedistinctions is relevant. The differences lie primarily in the inclusion criteria,goals, and evaluation methods. First, the Mattingly et al. criteria were more in-clusive. Mattingly et al. used a broad definition of parent involvement, which al-lowed for the inclusion of home-school collaboration programs. These reviewersalso included multicomponent programs that did not isolate the parent involve-ment component, programs without control groups, programs with post-test dataonly, and studies that used qualitative interview and survey data. In terms ofgoals, Mattingly et al.'s review aimed at pinpointing how the effectiveness ofprograms differed based on the quality of methodology. They did not evaluatethe effectiveness of each program separately; therefore, the reader could notdraw clear conclusions about which assessment, design, or methodology flawsmight have influenced the effectiveness of the interventions. In comparison withthe current review, Mattingly et al. used an effectiveness ratio and not an effectsize. Thus, Mattingly et al. provide a methodological review of the parent in-volvement literature but provide the practitioner with little guidance regardingevidence-based parent involvement interventions.
In summary, reviews of the evidence supporting the effectiveness of parentinvolvement in enhancing the academic performance of children are inconclu-sive. Clarifying the evidence that supports the beneficial impact of parent in-volvement programs on children's school outcomes is paramount given its fi-nancial and social importance in education (Mattingly et al., 2002). This reviewused a standardized coding system to compare the effectiveness of parent in-volvement programs. To clarify the unique effect of parent involvement onchildren's performance in school, the review was limited to studies in whichthe parent involvement component was identifiable and child outcomes weremeasured.
376 FISHEL AND RAMIREZ
METHODS
Review Strategy
The present review encompassed empirical studies of parent involvement pub-lished between 1980 and early 2003. Only studies with a target population ofschool-aged children and adolescents (K-12) were included. To find appropriatestudies, the authors conducted a thorough search of relevant databases. Searchterms included but were not limited to the general terms of "parent involve-ment," "parent participation" and more specific activities involving parents, suchas "parent tutoring," "parent volunteering," and "parenting." To narrow thewealth of studies, the terms above were crossed with outcome-related terms suchas "academic achievement," "education," "behavior," and/or "school." In addi-tion, a manual search was conducted by tracking the relevant references in arti-cles and books on parent involvement. This process yielded hundreds of studiesof varying quality and design, requiring refinement in the selection of appropri-ate studies for this review.
Several exclusion criteria were applied to the initial pool of studies. Descrip-tive studies, case studies, and correlational studies were excluded. Studies wereincluded only if they measured a behavioral outcome, included a control group(applied only to group design studies), and used pre- and post-test results. Tofurther narrow the field of studies to a manageable number for coding, studieswere excluded in which the primary student outcomes were health-related, suchas drug and alcohol use or food consumption. Due to differences in the educationsystems, which may have limited generalization of conclusions, studies con-ducted outside of North America were excluded from analysis.2
Coding
After the studies were identified as meeting the selection criteria, they werecoded by the authors using the coding manual developed by the Division 16Task Force on Evidence-Based Interventions in School Psychology {Proceduraland Coding Manual for Review of Evidence-Based Interventions, March 21,2003 version). Each author coded half of the articles. Interrater reliability wasestablished on a sample of nine articles. If during coding an interrater reliabilitycoefficient was lower than .80, and/or systematic differences were discovered inhow a specific rating was assigned, the coders reached consensus and adjustedthe ratings accordingly. The final interrater reliability coefficient, based on per-cent agreement on the Summary Key Evidence ratings, was .85.
Effect sizes were calculated according to the procedures suggested by theManual. For group designs, the Cohen d method was used. For single-participant
2A list of parent involvement studies conducted outside of the United States is available from the firstauthor.
PARENT INVOLVEMENT 377
designs, similar to the effect size for the group design, the baseline was sub-tracted from the treatment mean and divided by the baseline standard deviation.This latter method was outlined as Method 1 of effect size calculation in theManual (Division 16 and Society for the Study of School Psychology TaskForce, 2003). Whether effect sizes were considered large, medium, or small var-ied, depending on which statistical procedure was used. The same procedureswere used to calculate and evaluate effect sizes with and without covariates (i.e.,ANOVA and ANCOVA). Where multiple outcomes were listed, effect sizeswere listed as ranges and included effect sizes for both main effects and interac-tions, provided that they lend themselves to calculation.
RESULTS
Descriptive Analysis
Twenty-four studies from 22 articles were selected and coded. The majority ofstudies (n = 14) utilized a between-subject group design; however, a substantialnumber (n = 8) used single-participant or mixed designs (n = 2). Descriptivecharacteristics of the studies appear on Tables 1 and 2 for between-group andsingle-participant/mixed designs, respectively. Most studies were conducted inthe 1980s and 1990s in the United States, with only two studies published in the2000s. More than half of all studies (58%) involved treatment of children withongoing school problems, while the remainder were selective or targeted preven-tion, and one study involved both prevention and intervention. Few studies col-lected follow-up data. Typical intervention duration was 10 or more weeks, al-though duration varied widely.
Most studies utilized a single type of parent involvement, learning at home, inwhich the parents worked directly with children at home assisting them in learn-ing school-relevant skills. Activities included parent-implemented tutoring, par-ent reinforcement/encouragement, and parents reading to their children. Fewstudies compared parent involvement with another treatment. In those that did,parent involvement was compared to another intervention (typically, peer orparaprofessional tutoring) (e.g., Fantuzzo, Davis, & Ginsburg, 1995; Heller &Fantuzzo, 1993) or two types of academic interventions by parents were com-pared (Powell-Smith, Stoner, Shinn, & Good, 2000). The primary outcomes inmost reviewed studies were students' academic performance/achievement, in-cluding reading or pre-reading skills and mathematics skills. Other outcomes in-cluded spelling and appropriate behaviors. Few studies focused on auxiliary(secondary) goals, such as self-concept.
Participating children represented a wide range of demographic characteris-tics, including academic, sensory, and cognitive delays/deficits, identified dis-abilities, and varying grade placements (kindergarten-seventh grade); however,no studies examined the efficacy of parent involvement with high school popula-tions. Group design studies typically demonstrated gender-balanced samples,
TABL
E 1
. Des
crip
tive
Cha
ract
eris
tics
of P
aren
t In
volv
emen
t In
terv
entio
n O
utco
me
Stu
dies
with
Gro
up D
esig
n
Typ
e of
Pro
gram
/ JV
/Par
ticip
ant
Setti
ng, L
ocal
e,St
udy
# R
efer
ence
D
escr
iptio
n C
hara
cter
istic
s T
arge
t B
ehav
ior
& L
engt
h Fi
ndin
gs
1 Fa
ires
, Nic
hols
, &
In
terv
entio
n/T
reat
- N
=8;
fir
st g
rade
rs;
Rea
ding
pro
fici
ency
U
.S. [
publ
ic]
scho
ol
Inte
rven
tion
succ
essf
ul:
Exp
erim
enta
lR
icke
lman
(20
00)
men
t; Pa
rent
tra
in-
thre
e C
auca
sian
pr
ogra
m &
hom
e gr
oup
mad
e ga
ins
from
pre
- to
pos
t-te
st i
nin
g an
d in
volv
e-
boys
; one
ESL
gir
l; tr
eatm
ent;
read
ing
prof
icie
ncy,
whe
reas
con
trol
gro
upm
ent:
In-h
ome
one
Afr
ican
Am
eri-
5
wee
ks, 2
0-30
di
d no
t.pa
rent
-im
plem
ente
d ca
n bo
y, &
thre
e m
inut
es, t
hree
tim
esre
adin
g le
sson
s A
fric
an A
mer
ican
pe
r w
eek
base
d on
Rea
ding
gi
rls;
38%
low
SE
S,R
ecov
ery
mod
el
37%
low
mid
dle
orm
iddl
e SE
S, 1
2%up
per
SES,
& 1
2%no
t sp
ecif
ied;
age
not
spec
ifie
d
2Fa
ntuz
zo, D
avis
, &
Tar
gete
d Pr
even
- N
= 7
2; fo
urth
and
M
athe
mat
ics
U.S
. pub
lic s
choo
l In
terv
entio
n pa
rtia
lly s
ucce
ssfu
l: A
cade
-G
insb
urg
(199
5)
tion;
Par
ent
and
fift
h gr
ader
s;
achi
evem
ent
and
prog
ram
& h
ome
mic
res
ults
not
sig
nifi
cant
for
PI
alon
epe
er i
nvol
vem
ent:
Afr
ican
Am
eric
an
self
-con
cept
tr
eatm
ent;
Peer
tu-
(im
prov
ed m
ath
com
puta
-tio
nal
skill
s an
dPa
rent
-im
plem
ente
d st
uden
ts w
ith d
iffi-
ta
ring
: 10
wee
ks, 4
5 se
lf-p
erce
ptio
n of
soc
ial
conf
iden
ce f
or P
Itu
tori
ng a
lone
(PI
) cu
lties
in
com
puta
- rn
in.,
two
times
per
+
RPT
com
pare
d to
con
trol
s an
d PT
alo
ne),
and
in c
ombi
natio
n tio
nal
skill
s; m
ean
wee
k, P
I: in
tens
ity
Impr
oved
per
cept
ion
of s
chol
astic
com
pe-
with
rec
ipro
cal
age
10 y
ears
; 54
%
vari
ed
tenc
e an
d be
havi
oral
con
duct
for
PI
+ R
PTpe
er-i
mpl
emen
ted
girl
s; S
ES
not
spec
i-
and
for
PI a
lone
com
pare
d to
con
trol
s,tu
tori
ng (
PI +
RPT
) fi
ed
378
3 H
elle
r &
Fan
tuzz
o T
arge
ted
Prev
en-
N =
84;
fou
rth-
and
M
athe
mat
ics
U.S
. pub
lic s
choo
l R
PT +
PI
grou
p di
d be
tter
than
RPT
gro
up(1
993)
tio
n; P
aren
t an
d fi
fth-
grad
ers;
ac
hiev
emen
t an
d pr
ogra
m &
hom
e an
d co
ntro
l gr
oup
on b
oth
mea
sure
s of
peer
inv
olve
men
t: A
fric
an A
mer
ican
se
lf-c
once
pt
trea
tmen
t; 8
mon
ths
mat
h ac
hiev
emen
t. In
terv
entio
n su
cces
sful
Rec
ipro
cal
peer
-im
- st
uden
ts a
t ris
k fo
r fo
r bo
th P
I an
d fo
r se
cond
ary
goal
s of
pos
itive
lea
rnin
gpl
emen
ted
tuto
ring
m
ath
achi
evem
ent,
RPT
, 45
min
., 2
skill
s, a
sser
tiven
ess,
and
task
ori
enta
tion,
alon
e (R
PT)
and
in
mea
n ag
e 9
year
s 9
times
per
wee
k M
ost
mea
sure
s of
sch
ool
adju
stm
ent
not
com
bina
tion
with
m
onth
s; 5
0% g
irls
; si
gnif
ican
t,pa
rent
-im
plem
ente
d lo
wer
to m
iddl
etu
tori
ng (
RPT
+ P
I)
SES
4 Ja
son,
Kur
asak
i, Se
lect
ive
Prev
en-
N=
55;
thir
d A
cade
mic
ach
ieve
- U
.S. p
aroc
hial
In
terv
entio
n im
prov
es m
athe
mat
ics
Neu
son,
& G
arci
a,
tion;
Par
apro
fes-
th
roug
h fi
fth
men
t/ pe
rfor
man
ce
scho
ol (
para
prof
es-
achi
evem
ent/p
erfo
rman
ce.
No
sign
ific
ance
(199
3)—
pilo
t fo
r si
onal
and
par
ent
in-
grad
ers;
49%
gir
ls;
in m
athe
mat
ics,
si
onal
trea
tmen
t) &
fo
r sp
ellin
g an
d re
adin
g. I
nter
vent
ion
not
Jaso
n, W
eine
et a
l. vo
lvem
ent:
51 %
Afr
ican
Am
er-
spel
ling,
and
rea
d-
hom
e (p
aren
t tr
eat-
su
cces
sful
for
sec
onda
ry g
oals
of
soci
al(1
993)
. Als
o, s
ee
Ori
enta
tion
prog
ram
ic
an, 3
6% H
ispa
nic,
in
g m
ent)
; Par
apro
fes-
co
ncep
t, be
havi
or, t
utor
com
pete
nce,
and
Jaso
n, W
eine
, et
al.
plus
par
apro
fes-
11
% C
auca
sian
; si
onal
tuto
ring
: ap
- so
cial
asp
ects
of
clas
sroo
m e
xper
ienc
e.(1
993)
bel
ow-Y
ear
sion
al t
utor
ing
at
child
ren
at r
isk
prox
. one
aca
dem
ic2
scho
ol a
lone
and
in
base
d on
low
er
year
, fo
r on
e ho
ur,
com
bina
tion
with
SE
S, li
fe e
vent
tw
o tim
es p
er w
eek,
pare
nt-i
mpl
emen
ted
stre
ssor
s, s
pelli
ng,
Pare
nt t
rain
ing:
one
tuto
ring
at h
ome
read
ing,
and
mat
he-
one-
hour
ses
sion
;m
atic
s ac
hiev
emen
t; tr
eatm
ent
dura
tion
age
not
spec
ifie
d an
d in
tens
ity n
otsp
ecif
ied
379
TABL
E 1
. Des
crip
tive
Cha
ract
eris
tics
of P
aren
t In
volv
emen
t In
terv
entio
n O
utco
me
Stu
dies
with
Gro
up D
esig
n (c
ontin
ued)
Typ
e of
Pro
gram
/ T
V/P
artic
ipan
t Se
tting
, Loc
ale,
Stud
y #
Ref
eren
ce
Des
crip
tion
Cha
ract
eris
tics
Tar
get
Beh
avio
r &
Len
gth
Find
ings
Jaso
n, W
eine
, et a
l.,
Sele
ctiv
e Pr
even
- N
=
150-
180;
thi
rd
Aca
dem
ic a
chie
ve-
U.S
. par
ochi
al
Sign
ific
ant
gain
s in
rea
ding
ach
ieve
men
t5
(199
3)-Y
ear3
tio
n; P
arap
rofe
s-
thro
ugh
fift
h m
erit/
per
form
ance
sc
hool
(pa
rapr
ofes
- an
d sp
ellin
g pe
rfor
man
ce.
Tre
atm
ent
ofsi
onal
and
par
ent
in-
grad
ers;
50%
gir
ls,
in m
athe
mat
ics,
si
onal
tre
atm
ent)
&
seco
ndar
y go
als
is m
ostly
uns
ucce
ssfu
l:vo
lvem
ent:
dive
rse
sam
ple,
sp
ellin
g, a
nd r
ead-
ho
me
(par
ent
trea
t-
sign
ific
ant
for
soci
al w
ithdr
awal
and
ina
t-O
rien
tatio
n pr
ogra
m
mos
tly A
fric
an
ing
men
t); A
ppro
xi-
tent
ion
but n
ot f
or s
ocia
l co
ncep
t, be
hav-
plus
par
apro
fes-
A
mer
ican
and
His
- m
atel
y on
e ac
a-
ior,
tuto
r co
mpe
tenc
e, a
nd s
ocia
l as
pect
s of
sion
al tu
tori
ng a
t pa
nic;
chi
ldre
n at
de
mic
yea
r, e
ach
clas
sroo
m e
xper
ienc
e,sc
hool
alo
ne a
nd in
ri
sk b
ased
on
low
er
com
pone
nt f
or 4
0-co
mbi
natio
n w
ith
SES,
life
eve
nt
60 m
inut
es, t
wo
pare
nt-i
mpl
emen
ted
stre
ssor
s, a
nd
times
per
wee
ktu
tori
ng a
t hom
e gr
ades
in
spel
ling,
read
ing,
and
mat
he-
mat
ics;
age
not
spec
ifie
d
6 M
ehra
n &
Whi
te
Sele
ctiv
e Pr
even
- N
= 7
6; K
inde
r-
Rea
ding
com
pe-
U.S
. edu
catio
nal
Tre
atm
ent
grou
p sh
owed
gai
ns o
ver
con-
(198
8)
tion;
Par
ent
trai
ning
ga
rten
ers
havi
ng
tenc
e/ac
quis
itio
n of
ag
ency
(tr
aini
ng),
tr
ol g
roup
on
mos
t mea
sure
s lo
adin
g on
and
invo
lvem
ent:
diff
icul
ty l
earn
ing
spec
ific
rea
ding
ho
me
(tre
atm
ent)
; ke
y va
riab
les.
Pare
nts
wer
e tr
aine
d to
rea
d; 3
9% fe
- sk
ills
eigh
t mon
ths,
15
in a
nd i
mpl
emen
ted
mal
es; m
ean
age
7 m
inut
es p
er s
essi
on,
tuto
ring
in r
eadi
ng
year
s, 4
mon
ths;
va
ried
num
ber
ofas
a s
uppl
emen
t to
lo
w to
low
mid
dle
sess
ions
per
wee
kco
mpe
nsat
ory
edu-
SE
S; e
thni
city
not
catio
n sp
ecif
ied
380
1 M
iller
& K
ra-
Inte
rven
tion/
Tre
at-
N=
52;
sec
ond,
R
eadi
ng a
ccur
acy,
U
.S. p
ublic
sch
ool
Tre
atm
ent
grou
p di
d no
t im
prov
e on
ove
r-to
chw
ill (
1996
) m
ent;
Pare
nt tr
ain-
th
ird,
and
fou
rth
rate
and
com
pre-
(t
rain
ing)
, hom
e al
l rea
ding
sco
res
com
pare
d to
con
trol
ing
and
invo
lve-
gr
ader
s; m
ostly
he
nsio
n (t
reat
men
t);
10
grou
p,m
ent:
Pare
nt-
Cau
casi
an;
havi
ng
wee
ks,
10-t
o-15
-im
plem
ente
d tu
tor-
tr
oubl
e in
rea
ding
, m
inut
e se
ssio
ns,
ing
usin
g th
e Pa
ired
re
ceiv
ing
Cha
pter
1
five
tim
es p
er w
eek
Rea
ding
met
hod,
se
rvic
es;
scho
olfo
cusi
ng o
n m
odel
- SE
S lo
w;
gend
er,
ing,
pra
ctic
e, p
rais
e,
age,
sam
ple
SES
and
posi
tive
atm
os-
not
spec
ifie
dph
ere.
Pow
ell-
Smith
, In
terv
entio
n/T
reat
- N
= 3
6; s
econ
d R
eadi
ng a
chie
ve-
U.S
. pub
lic s
choo
l R
eadi
ng p
rofi
cien
cy d
id n
ot i
mpr
ove;
Ston
er,
Shin
n, G
ood
men
t; Pa
rent
tra
in-
grad
ers;
mea
n ag
e 7
men
t (t
rain
ing)
, hom
e ho
wev
er,
indi
vidu
al a
naly
ses
show
ed i
m-
III,
& R
olan
d in
g an
d in
volv
e-
year
s 11
mon
ths;
(t
reat
men
t);
15
prov
emen
ts, p
artic
ular
ly o
f th
e st
uden
ts(2
000)
m
ent;
Com
pari
ng
39%
gir
ls; 7
7% o
f w
eeks
ove
rall
(5
with
poo
rest
ini
tial
read
ing
achi
evem
ent,
two
pare
nt-i
mpl
e-
pare
nts
Cau
casi
an,
wee
ks o
f tr
eatm
ent)
,m
ente
d tu
tori
ng
12%
His
pani
c, 1
2%
20 m
inut
es,
four
prog
ram
s: U
sing
lit-
N
ativ
e A
mer
ican
; tim
es p
er w
eek
erat
ure
and
usin
g 50
% p
aren
ts w
ithcu
rric
ulum
mat
eri-
po
st-h
igh
scho
ol e
d-al
s uc
atio
n; m
ost
iden
-tif
ied
as lo
w r
eade
rsbu
t not
in S
peci
alE
duca
tion;
mos
t re-
ceiv
ing
Cha
pter
1re
adin
g se
rvic
es.
8
381
TABL
E 1
. Des
crip
tive
Cha
ract
eris
tics
of P
aren
t In
volv
emen
t In
terv
entio
n O
utco
me
Stu
dies
with
Gro
up D
esig
n (c
ontin
ued)
Typ
e of
Pro
gram
/ M
Part
icip
ant
Setti
ng, L
ocal
e,St
udy
# R
efer
ence
D
escr
iptio
n C
hara
cter
istic
s T
arge
t B
ehav
ior
& L
engt
h Fi
ndin
gs
Sear
ls, L
ewis
, &
Sele
ctiv
e Pr
even
- N
= 5
0; f
irst
M
athe
mat
ics
and
U.S
. [pu
blic
] sc
hool
R
eadi
ng a
chie
vem
ent,
wor
d an
alys
is s
kills
,9
Mor
row
(19
82)-
tio
n; P
aren
t tra
inin
g gr
ader
s; A
ge 5
re
adin
g ac
hiev
e-
(tra
inin
g), h
ome
and
over
all
mat
h ac
hiev
emen
t im
prov
edSt
udy
1 an
d in
volv
emen
t: ye
ars
7 m
onth
s to
7
men
t (t
reat
men
t); 2
0 fo
r th
e tr
eatm
ent
grou
p. C
linic
al S
igni
fi-
Dai
ly p
aren
t tu
tor-
ye
ars
9 m
onth
s,
wee
ks: 2
0-30
min
- ca
nce:
Chi
ldre
n's
attit
ude
tow
ard
read
ing,
ing
to r
einf
orce
52
% g
irls
; mos
t ute
s, o
nce
daily
pa
rent
atti
tude
tow
ard
scho
ol i
mpr
oved
,re
adin
g/la
ngua
ge
Cau
casi
an;
24%
of
self
-con
cept
and
mot
ivat
ion
impr
oved
,ar
ts a
nd m
ath
skill
s lo
w S
ES;
low
tohi
gh r
eadi
ng r
eadi
-ne
ss l
evel
s
10
Sear
ls, L
ewis
, &
Sele
ctiv
e Pr
even
- N
-52;
se
cond
M
athe
mat
ics
and
U.S
. [pu
blic
] sc
hool
N
o si
gnif
ican
t di
ffer
ence
on
the
post
test
Mor
row
(19
82)-
tio
n; P
aren
t tr
aini
ng
grad
ers;
Age
6
read
ing
achi
eve-
(t
rain
ing)
, hom
e m
easu
res
of m
athe
mat
ics
and
read
ing.
Stud
y 2
and
invo
lvem
ent:
year
s 9
mon
ths
to 7
m
ent
(tre
atm
ent)
; 20
Dai
ly p
aren
t tut
or-
year
s 11
mon
ths,
w
eeks
: 20-
30 m
in-
ing
to r
einf
orce
80
% g
irls
; mos
t ute
s, o
nce
daily
read
ing/
lang
uage
C
auca
sian
; no
ne o
far
ts a
nd m
athe
mat
- lo
w S
ES;
low
toic
s sk
ills
high
rea
ding
rea
di-
ness
lev
els
11Sh
uck,
Uls
h, &
In
terv
entio
n/T
reat
- N
= 1
50;
thir
d-fi
fth
Rea
ding
ach
ieve
- U
.S. p
ublic
sch
ool,
Tre
atm
ent
grou
p ga
ins
over
con
trol
gro
upPl
att
(198
3)
men
t; Pa
rent
tra
in-
grad
ers
with
rea
d-
men
t ho
me
(tre
atm
ent)
; in
rea
ding
ach
ieve
men
t,in
g an
d in
volv
e-
ing
dela
ys;
gend
er,
One
aca
dem
ic y
ear,
men
t: Pa
rent
s SE
S, e
thni
city
, age
in
tens
ity u
nkno
wn
enco
urag
ed p
upils
no
t sp
ecif
ied
to r
ead
(PE
P pr
oj-
ect)
382
12
Tro
vato
& B
uche
r In
terv
entio
n/T
reat
- N
= 6
9; 1
1 se
cond
R
eadi
ng a
bilit
y C
anad
ian
publ
ic
Posi
tive
resu
lts
for
oral
rea
ding
, com
pre-
(198
0)
men
t; Pe
er a
nd p
ar-
grad
ers,
48
thir
d sc
hool
(tr
eatm
ent-
1 -
hens
ion,
and
ove
rall
read
ing,
with
Hom
e-en
t in
volv
emen
t: gr
ader
s, a
nd 1
0 pe
er t
utor
ing)
; B
ased
gro
up o
utpe
rfor
min
g Pe
er T
utor
ing
Ope
rant
-bas
ed c
or-
four
th g
rade
rs w
ith
hom
e (t
reat
men
t-2-
gr
oup,
and
the
latte
r ou
tper
form
ing
the
rect
ive
peer
-im
ple-
re
adin
g de
lays
fro
m
pare
nt tu
tori
ng);
co
ntro
ls,
men
ted
tuto
ring
se
ven
scho
ols;
Pe
er tu
tori
ng:
Ave
r-w
ith o
r w
ithou
t m
ean
age
9 ye
ars;
ag
e of
15
wee
ks; 4
6ho
me-
base
d pa
rent
- ge
nder
not
spe
ci-
sess
ions
; 30
min
-im
plem
ente
d tu
tor-
fi
ed;
SES
vari
ed,
utes
ing
rein
forc
emen
t m
ostly
mid
dle
clas
s; e
thni
city
not
spec
ifie
d
13
Vin
ogra
d-B
ause
ll &
In
terv
entio
n/T
reat
- AT
= 1
95;
firs
t W
ord
reco
gniti
on
U.S
. pub
lic s
choo
l E
xper
imen
tal
grou
p re
cogn
ized
sig
nifi
-B
ause
ll (1
987)
m
ent;
Pare
nt t
rain
- gr
ader
s, m
ost
en-
(tra
inin
g), h
ome
cant
ly m
ore
wor
ds th
an t
he c
ontr
ols.
ing
and
invo
lve-
ro
lled
in r
egul
ar e
d-
(tre
atm
ent)
; 2m
ent:
Pare
nt tu
tor
ucat
ion
clas
ses;
w
eeks
, with
an
av-
wor
d re
cogn
ition
ge
nder
, SE
S, e
th-
erag
e of
8.3
day
s,sk
ills
nici
ty, a
nd a
ge n
ot
17 m
inut
es a
day
spec
ifie
d
Wal
ters
& G
unde
r-
Inte
rven
tion/
Tre
at-
N =
39;
fou
rth g
rade
E
nglis
h la
ngua
ge
Can
adia
n pu
blic
A
ll th
ree
grou
ps m
ade
sign
ific
ant
gain
s,so
n (1
985)
m
ent;
Pare
nt i
n-
ESL
stu
dent
s (C
an-
read
ing
skill
s sc
hool
(tr
eatm
ent)
; w
ith n
o si
gnif
ican
t di
ffer
ence
s be
twee
nvo
lvem
ent:
Pare
nt
tone
se s
peak
ers)
; 40
min
utes
, tw
o gr
oups
.vo
lunt
eers
rea
d ge
nder
, SE
S, a
nd
times
per
wee
k,bo
oks
in C
anto
nese
ag
e no
t sp
ecif
ied
leng
th n
ot r
epor
ted
or E
nglis
h to
ESL
stud
ents
Not
e: B
rack
ets
wer
e us
ed w
hen
a ty
pe o
f sch
ool (
e.g.
, [pu
blic
]) w
as in
ferre
d.
14
383
TABL
E 2
. D
escr
iptiv
e C
hara
cter
istic
s of
Par
ent
Invo
lvem
ent
Inte
rven
tion
Out
com
e S
tudi
es w
ith S
ingl
e-P
aren
t an
d M
ixed
Des
igns
Typ
e of
Pro
gram
/ T
V/P
artic
ipan
t Se
tting
, Loc
ale,
Stud
y #
Ref
eren
ce
Des
crip
tion
Cha
ract
eris
tics
Tar
get
Beh
avio
r &
Len
gth
Find
ings
Cal
laha
n,
Tar
gete
d Pr
even
- N
= 2
6; s
ixth
and
H
omew
ork
per-
U
.S. p
ublic
sch
ool
Mat
hem
atic
s ac
hiev
emen
t, ho
mew
ork
Rad
emac
her,
& H
il-
tion;
Par
ent
trai
ning
se
vent
h gr
ader
s fo
rman
ce i
n m
athe
- (t
rain
ing)
, hom
e co
mpl
etio
n, a
nd h
omew
ork
qual
ity i
m-
dret
h (1
998)
) an
d in
volv
emen
t: fr
om t
he a
t ris
k fo
r m
atic
s an
d ac
a-
(tre
atm
ent)
; Ins
uffi
- pr
oved
; rea
ding
and
spe
lling
ach
ieve
men
tIn
-hom
e, p
aren
t-fa
- al
coho
l an
d dr
ug
dem
ic a
chie
vem
ent
cien
t in
form
atio
n di
d no
t im
prov
e; th
ose
stud
ents
who
se p
ar-
cilit
ated
sel
f-m
an-
use
yout
h pr
ogra
ms;
(m
ath,
rea
ding
, pr
ovid
ed
ents
wer
e m
ore
invo
lved
im
prov
ed m
ore.
agem
ent
and
rein
- ge
nder
, SE
S, e
th-
spel
ling)
; Mix
edfo
rcem
ent
prog
ram
ni
city
, age
not
spe
c-ifi
ed
Coa
tes
&
Inte
rven
tion/
Tre
at-
N=
\;l
year
old
R
eadi
ng a
chie
ve-
Can
adia
n [p
ublic
] R
eadi
ng s
peed
and
acc
urac
y di
d no
t in
-2
McL
augh
lin (
1992
) m
ent;
Pare
nt t
rain
- fi
rst
grad
e bo
y w
ith
men
t; Si
ngle
-Par
tic-
scho
ol, h
ome
(tre
at-
crea
se; h
owev
er, t
here
was
clin
ical
im
-in
g an
d in
volv
e-
visu
al-m
otor
and
ip
ant
men
t); 6
0 sc
hool
pr
ovem
ent
(rel
atio
ns w
ith p
aren
ts, a
ttitu
dem
ent:
In-h
ome
audi
tory
def
icits
re-
da
ys, t
reat
men
t fo
r to
war
d re
adin
g, p
eer
rela
tions
),pa
rent
-im
plem
ente
d pe
atin
g fir
st g
rade
; 24
sch
ool
days
, fo
rtu
tori
ng o
f re
adin
g SE
S &
eth
nici
ty n
ot
15-3
0 m
inut
es p
erus
ing
flas
h ca
rds
spec
ifie
d se
ssio
n; n
umbe
r of
sess
ions
not
re-
port
ed
3D
uval
l, D
elqu
adri
, In
terv
entio
n/T
reat
- N
= 4
; Tw
o se
cond
O
ral r
eadi
ng s
kills
U
.S. r
ural
[pu
blic
] A
ll fo
ur s
tude
nts
dem
onst
rate
d ex
pect
edE
lliot
t, &
Hal
l m
ent
for
thre
e pa
r-
grad
ers
(mea
n ag
e 7
(cor
rect
rat
es a
nd
scho
ol, h
ome
(som
e in
crea
ses
in c
orre
ct r
eadi
ng r
ates
and
(199
2)
ticip
ants
; T
arge
ted
year
s 6
mon
ths)
, er
ror
rate
s) a
nd
data
col
lect
ion
and
thre
e-qu
arte
rs s
how
ed c
onsi
sten
tly l
owPr
even
tion
for
one
one
thir
d gr
ader
(8
read
ing
achi
eve-
tr
eatm
ent)
; Dur
ing
erro
r ra
tes.
Pre
-pos
t re
adin
g ga
ins
are
re-
part
icip
ant;
Pare
nt
year
s 1
mon
th),
and
m
ent;
Sing
le-P
artic
- th
e su
mm
er b
reak
po
rted
on
a te
st o
f re
adin
g ac
hiev
emen
t,tr
aini
ng a
nd in
- on
e fif
th g
rade
r (9
ip
ant
(end
of
spri
ng s
e-
Res
ults
gen
eral
ized
for
thr
ee o
f fo
ur c
hil-
volv
emen
t: In
-hom
e ye
ars
10 m
onth
s);
mes
ter
- st
art
of f
all
dren
acr
oss
time
and
setti
ngs
(hom
e to
pare
nt-i
mpl
emen
ted
two
girl
s; tw
o w
ith
sem
este
r), n
umbe
r sc
hool
),tu
tori
ng o
f re
adin
g ab
ove
avg.
IQ
&
of s
essi
ons
vari
edtw
o w
ith a
vg. I
Q;
26-4
3, 1
0 m
inut
esth
ree
iden
tifie
d as
da
ily
1
384
LD
, rea
ding
diff
i-cu
lties
; all
mid
dle
SES;
eth
nici
ty n
otsp
ecif
ied
4 G
ang&
Poc
he
Inte
rven
tion/
Tre
at-
N=
3; E
nd o
f th
ird
Chi
ldre
n's
acqu
isi-
U
.S. h
ome
(tra
inin
g A
ll 3
child
ren
reac
hed
100%
acc
urac
y in
(198
2)
men
t; Pa
rent
tra
in-
grad
e, 2
yea
rs b
e-
tion
of r
eadi
ng
and
trea
tmen
t); 7
al
l se
ven
cate
gori
es o
f re
adin
g sk
ills
and
ing
and
invo
lve-
hi
nd in
rea
ding
; al
l sk
ills
and
pare
nts'
w
eeks
, 25
min
utes
, av
erag
ed r
eadi
ng g
ains
3X
the
expe
cted
men
t: In
-hom
e bo
ys;
SES,
eth
nic-
tu
tori
ng b
ehav
iors
fo
ur
times
per
wee
k ga
ins.
Gai
ns m
aint
aine
d at
11-
wee
k fo
l-pa
rent
-im
plem
ente
d ity
, and
age
not
lo
w-u
p,tu
tori
ng o
f rea
ding
sp
ecif
ied
5 H
ook&
DuP
aul
Inte
rven
tion/
Tre
at-
N =
4; T
hree
7 y
ear
Rea
ding
per
form
- U
.S. p
ublic
sch
ool
At-
hom
e or
al r
eadi
ng r
ate
mea
ns a
nd l
ev-
(199
9)
men
t; Pa
rent
tra
in-
olds
in s
econ
d gr
ade
ance
at h
ome
and
(« =
3),
U.S
. pri
vate
el
s im
prov
ed f
or a
ll. G
ains
mai
ntai
ned
for
ing
and
invo
lve-
an
d on
e 8
year
old
sc
hool
; Si
ngle
-Par
- sc
hool
(n
= 1
) &
on
ly o
ne. A
t-sc
hool
rea
ding
per
form
ance
men
t: In
-hom
e in
thir
d gr
ade
with
tic
ipan
t ho
me.
Var
ied
for
impr
oved
for
all;
two
of f
our
gain
s at
fol
-pa
rent
-im
plem
ente
d A
DH
D; t
hree
boy
s,
each
par
ticip
ant;
10
low
-up.
No
chan
ge i
n te
ache
r-ra
ted
over
all
tuto
ring
of
read
ing
one
girl
; SE
S, e
th-
min
utes
, tw
o-th
ree
read
ing
perf
orm
ance
,ni
city
not
spe
cifi
ed
times
per
wee
k fo
r45
, 37
, 37
, and
26
wee
ks
Law
& K
rato
chw
ill
Inte
rven
tion/
Tre
at-
N=
13
; Fiv
e co
m-
Rea
ding
ach
ieve
- U
.S. p
ublic
sch
ool
Rea
ding
acc
urac
y an
d re
adin
g fl
uenc
y di
d(1
993)
m
ent;
Pare
nt t
rain
- pl
eted
fir
st g
rade
; m
ent;
Mix
ed
(tra
inin
g), h
ome
not
incr
ease
.in
g an
d in
volv
e-
five
com
plet
ed s
ec-
(tre
atm
ent)
; 6m
ent:
In-h
ome
ond
grad
e, &
thre
e w
eeks
, 10
min
utes
pair
ed r
eadi
ng (
chil-
co
mpl
eted
thi
rd
per
sess
ion,
fiv
edr
en-p
aren
ts)
pro-
gr
ade;
gen
der,
SE
S,
times
per
wee
kgr
am
ethn
icity
, ag
e no
tsp
ecif
ied
6
385
TABL
E 2
. Des
crip
tive
Cha
ract
eris
tics
of P
aren
t In
volv
emen
t In
terv
entio
n O
utco
me
Stu
dies
with
Sin
gle-
Par
ent
and
Mix
ed D
esig
ns (
cont
inue
d)
Typ
e of
Pro
gram
/ M
Part
icip
ant
Setti
ng, L
ocal
e,St
udy
# R
efer
ence
D
escr
iptio
n C
hara
cter
istic
s T
arge
t B
ehav
ior
& L
engt
h Fi
ndin
gs
7 L
opez
& C
ole
Tar
gete
d Pr
even
- N
= 5
; Thr
ee 5
- A
cade
mic
rea
dine
ss
Puer
to R
ican
pub
lic
Mea
n nu
mbe
r of
kno
wn
lette
rs i
mpr
oved
(199
9)
tion;
Par
ent
trai
ning
ye
ar-o
lds
and
two
(pre
-rea
ding
ski
lls);
sc
hool
& h
ome;
fo
r al
l fiv
e ch
ildre
n, b
ut s
lope
cha
nged
for
and
invo
lvem
ent:
6-ye
ar-o
lds
in
Sing
le P
artic
ipan
t fo
ur
sess
ions
per
on
ly tw
o of
fiv
e; M
ean
lette
r-na
min
g ra
teIn
-hom
e pa
rent
-im
- K
inde
rgar
ten;
fou
r w
eek,
unk
now
n an
d sl
ope
impr
oved
for
all
five
chi
ldre
n,pl
emen
ted
tuto
ring
gi
rls,
one
boy
; SE
S,
leng
th o
f tim
e pe
rof
pre
-rea
ding
ski
lls
ethn
icity
not
spe
ci-
sess
ion,
for
a to
tal
usin
g fl
ash
card
s fie
d of
15
or 2
5 se
ssio
ns
8 T
hurs
ton
& D
asta
In
terv
entio
n/T
reat
- N
= 3
; Tw
o 9
year
M
athe
mat
ics
per-
U
.S. p
riva
te s
choo
l M
athe
mat
ics
achi
evem
ent
scor
es, a
t-(1
990)
-sec
ond
stud
y m
ent;
Pare
nt t
rain
- ol
ds a
nd o
ne 1
0 fo
rman
ce a
t hom
e (s
ome
data
col
lec-
sc
hool
kno
wle
dge
of m
ath
fact
s, a
vg. a
t-in
g an
d in
volv
e-
year
old
; al
l fo
urth
an
d sc
hool
; Si
ngle
- ti
on),
hom
e (s
ome
hom
e m
ath
prob
lem
sol
ving
acc
urac
y, a
ndm
ent:
In-h
ome
grad
ers
faili
ng i
n Pa
rtic
ipan
t da
ta c
olle
ctio
n an
d ov
eral
l gr
ades
in m
ath
impr
oved
for
all
pare
nt-i
mpl
emen
ted
mat
h; a
ll gi
rls;
SE
S,
trea
tmen
t);
5 or
8
stud
ents
.tu
tori
ng o
f bas
ic
ethn
icity
, ag
e no
t w
eeks
, 10
-min
ute
mat
h us
ing
flash
sp
ecif
ied
sess
ions
dai
lyca
rds
386
9 T
hurs
ton
& D
asta
In
terv
entio
n/T
reat
- N
= 1
; fou
rth
grad
e Sp
ellin
g pe
rfor
m-
U.S
. uni
vers
ity-
af-
m-s
choo
l m
ean
spel
ling
perf
orm
ance
im
-(1
990)
-thi
rd s
tudy
m
ent;
Pare
nt tr
ain-
gi
rl;
SES,
eth
nici
ty,
ance
and
acc
urac
y fi
liate
d ce
nter
pr
oved
; A
t-ho
me
spel
ling
accu
racy
im
-in
g an
d in
volv
e-
age
not
spec
ifie
d at
sch
ool;
Sing
le-
(tra
inin
g), s
choo
l pr
oved
,m
ent:
In-h
ome,
Pa
rtic
ipan
t (d
ata
colle
ctio
n),
pare
nt-i
mpl
emen
ted
hom
e (t
reat
men
t);
tuto
ring
of
spel
ling
10-m
inut
e se
ssio
nsus
ing
flas
h ca
rds
daily
for
10
wee
ks
10
Witt
, Han
nafi
n, &
In
terv
entio
n/T
reat
- N
= 3
; fou
rth
Aca
dem
ic p
erfo
rm-
U.S
. sch
ool
(dat
a A
ll su
bjec
ts'm
eans
and
lev
els
of a
cade
mic
Mar
tens
(19
83)
men
t; Pa
rent
tra
in-
grad
ers
with
poo
r an
ce in
lang
uage
co
llect
ion)
, hom
e pe
rfor
man
ce (
% c
orre
ct o
n as
sign
men
ts)
ing
and
invo
lve-
ac
adem
ic p
erfo
rm-
arts
and
inap
prop
ri-
(tre
atm
ent)
; 9,
14,
im
prov
ed;
tren
d ch
ange
d fo
r on
e su
bjec
tm
ent:
In-h
ome
ance
and
ina
ppro
- at
e be
havi
ors
at
or 2
0 da
ys, i
nten
sity
on
ly; A
ll su
bjec
ts'm
eans
and
leve
ls o
f in
-pa
rent
-im
plem
ente
d pr
iate
beh
avio
r; a
ll sc
hool
; Si
ngle
-Par
- no
t sp
ecif
ied
appr
opri
ate
beha
vior
s im
prov
ed;
tren
ds d
idac
adem
ic p
erfo
rm-
boys
; SE
S, e
thni
c-
ticip
ant
not c
hang
e,an
ce r
einf
orce
men
t ity
, age
not
spe
ci-
prog
ram
fle
d
Not
e: B
rack
ets
wer
e us
ed w
hen
a ty
pe o
f sc
hool
(e.
g., [
publ
ic])
was
inf
erre
d.
387
388 FISHEL AND RAMIREZ
students of lower-middle to middle socioeconomic status (SES) and inclusion ofseveral minority groups, including African Americans, Latinos, and English as aSecond Language (ESL) students. In single-participant design studies, few au-thors indicated participants' ethnicity or SES.
Methodological Quality
Methodological features of the studies included in this review are presented inTables 3 and 4, and a Summary of Evidence for Key Methodological Featuresappears in Tables 5 and 6.
Group design studies. As shown in Table 3, methodological characteristics ofgroup design parent involvement studies varied greatly and overall yieldedstrengths as well as weaknesses. Group design studies demonstrated consistentuse of such methodological features as appropriate unit of analysis (all studies),documentation of program components (93%), equivalent mortality with low at-trition (93%, although in many low attrition was not reported but inferred),group equivalence (86%), manualization (79%), and randomization (64%). Lessconsistent features, found in many but not the majority of studies, were assess-ment of educational/clinical significance (50%), use of multiple methods to col-lect data (50%), sufficiently large number of participants (50%), reporting nullfindings (46%), controlling for Type I error (46%), linking identifiable compo-nents to primary outcomes (36%), and using multiple sources for data collection(25%). Several group design studies utilized standardized tests without referenceto their validity and reliability with the population under study, or studies did notreport the validity and reliability of outcome measures, resulting in loweredevaluative ratings. Group design studies showed pronounced methodologicalweaknesses in the counterbalancing of change agents (only one study by Heller& Fantuzzo, 1993 addressed the issue) and the reporting of effect sizes (n = 2). Aserious shortcoming across group design studies was failure to report essentialdata. Specifically, the number of participants in each group, means, standard de-viations, F-ratios, and p-values were frequently missing from published articles,making it impossible to calculate effect sizes. A summary of the methodologicalfeatures necessary for strong evidence in group design studies appears on Table5. Methodological strengths include the quality of comparison group, school-based implementation of the intervention, and implementation fidelity. Method-ological weaknesses include lack of replication studies, lack of significant keyoutcomes, and failure to assess educationally significant outcomes, or differenti-ate components in multicomponent designs.
Single-participant/mixed design studies. When looking at methodologicalfeatures found in the single-participant and mixed designs (see Table 4), the au-thors must note that several of the features evaluated in group design studies(i.e., control of Type I error, sufficiently large N, randomization, counterbalanc-ing, and appropriate unit of analysis) were applicable only to the mixed designstudies. With this exclusion in mind, when evaluating total methodological fea-
TABL
E 3
. Met
hodo
logi
cal
Feat
ures
of
Par
ent
Invo
lvem
ent
Stu
dies
with
Gro
up D
esig
n
Fea
ture
S
tudy
Nu
mb
er
SI
S2
S3
S4
S5
S6
S7
S8
S9
S10
S
ll
S12
S
13
S14
Ran
do
miz
atio
n 0
1 1
01
11
10
01
01
1
App
ropr
iate
uni
t of
ana
lysi
s 1
1 1
1 1
1 1
1 1
1 1
(1
)1
1
Fam
ily-
wis
e er
ror
rate
con
trol
led
01
11
01
01
00
11
10
Suf
fici
entl
y la
rge
Mg
rou
p 0
01
01
11
00
11
1 1
0
Rel
iabl
e o
utc
om
e m
easu
res
11
00
00
0 1
00
00
1 0
Mul
tipl
e as
sess
men
t m
eth
od
s 0
11
11
10
10
00
10
0
Mea
sure
s ob
tain
ed f
rom
mul
tipl
e so
urce
s —
0
01
11
00
00
00
—
0
Val
idit
y of
mea
sure
s re
port
ed
10
10
00
0 1
00
00
1 0
Con
trol
or
com
pari
son
gro
up
* 1
1 1
1 1
11
11
11
1 1
1
Cou
nter
bala
ncin
g of
cha
nge
agen
ts
00
10
00
00
00
00
00
Gro
up
equi
vale
nce
esta
blis
hed
0 1
11
0 1
1 1
11
11
1 1
Equ
ival
ent
mor
tali
ty w
ith
low
att
riti
on
(1
)1
1 1
(1)
1 1
1 (1
) (1
) (1
) (1
) 0
(1)
Eff
ect
size
rep
orte
d 1*
* 0
00
01
00
00
00
00
Nu
ll f
indi
ngs
repo
rted
1
01
00
11
10
0-
01
0
Edu
cati
onal
/cli
nica
l si
gnif
ican
ce o
f ch
ange
ass
esse
d 0
0 1
11
00
11
10
10
0
Pro
gram
co
mp
on
ents
do
cum
ente
d 1
1 1
1(
1)
1 1
1 1
1 1
1 1
0
Iden
tifi
able
co
mp
on
ents
lin
ked
to p
rim
ary
ou
tco
mes
0
0 1
1 0
10
00
01
1 1
0
Inte
rven
tion
s w
ere
man
ual
ized
1
1 1
0(
1)
1 1
1 1
1 1
1 0
0
Tot
al (
Nu
mb
er o
f fe
atur
es o
ut o
fto
tal
appl
icab
le f
eatu
res)
9/
17
11/1
8 15
/18
10/1
8 9/
18
14/1
8 9/
18
13/1
8 8/
18
9/18
10
/17
11/1
8 11
/17
5/18
Not
e: S
= s
tudy
; 1
= m
etho
dolo
gica
l fe
atur
e w
as p
rese
nt;
0 =
met
hodo
logi
cal
feat
ure
was
abs
ent,
unkn
own,
or
unco
dabl
e; —
= m
etho
dolo
gica
l fe
atur
e w
as n
ot a
pplic
able
; ()
= m
etho
dolo
gica
l fe
a-tu
re w
as n
ot r
epor
ted,
but
inf
erre
d by
rev
iew
ers.
* St
udie
s w
ere
sele
cted
for
rev
iew
if
they
had
a c
ontr
ol g
roup
.**
eff
ect
size
(s)
calc
ulat
ed b
ut a
re n
ot i
nter
pret
able
.
389
TA
BL
E 4
. M
eth
od
olo
gic
al
Fea
ture
s of
Par
ent
Inv
olv
emen
t S
tudi
es w
ith
Sin
gle-
Par
tici
pant
an
d M
ixed
D
esig
n
Feat
ure
Stud
y N
umbe
r
SI
S2
S3
S4
S5
S6
S7
S8
S9
S10
Rel
iabl
e ou
tcom
e m
easu
res
0 1
1 1
1 1
11
10
Mul
tiple
ass
essm
ent
met
hods
—
—
1
1 —
—
—
1
1 0
Mea
sure
s ob
tain
ed f
rom
mul
tiple
sou
rces
0
0 1
01
00
11
0V
alid
ity o
f m
easu
res
repo
rted
0
1 1
11
11
11
0V
isua
l an
alys
is f
indi
ngs
pres
ente
d 0
0 1
11
11
11
1E
ffec
t si
ze r
epor
ted
00
0 0
10
00
01
Edu
catio
nal/c
linic
al s
igni
fica
nce
of c
hang
e as
sess
ed
11
1 0
1 1
11
0 0
Prog
ram
com
pone
nts
docu
men
ted
10
1 1
11
11
11
Iden
tifia
ble
com
pone
nts
linke
d to
pri
mar
y ou
tcom
es
0 0
10
10
10
11
Inte
rven
tions
man
ualiz
ed
10
1 1
11
11
11
Tot
al (
Num
ber
of f
eatu
res
out
of to
tal
appl
icab
le f
eatu
res)
3/
10
3/10
9/
10
6/10
9/
9 6/
9 7/
9 8/
10
8/10
5/
10
Not
e: S
= s
tudy
; 1
= m
etho
dolo
gica
l fe
atur
e w
as p
rese
nt;
0 =
met
hodo
logi
cal
feat
ure
was
abs
ent,
unkn
own,
or
unco
dabl
e; —
= m
etho
dolo
gica
l fe
atur
e w
as n
ot a
pplic
able
; ()
= m
etho
dolo
gica
l fe
a-tu
re w
as n
ot r
epor
ted,
but
inf
erre
d by
rev
iew
ers.
390
TA
BL
E 5
. S
um
mar
y of
Evi
denc
e fo
r K
ey M
eth
od
olo
gic
al
Fea
ture
s of
Par
ent
Inv
olv
emen
t S
tud
ies
wit
h G
rou
p D
esig
n
Fea
ture
St
udy
Num
ber
SI
S2
S3
S4
S5
S6
S7
S8
S9
S10
S
ll
S12
S
13
S14
Mea
sure
men
t 0
20
00
00
00
00
03
0C
ontr
ol/C
ompa
riso
n G
roup
1
3 3
2 (2
) 2
2 2
(2)
(2)
2 (2
) 1
(3)
Mea
sure
s Su
ppor
t Pr
imar
y O
utco
mes
0
13
11
20
00
03
20
0E
duca
tiona
l/Clin
ical
Sig
nifi
canc
e 1
12
12
10
21
12
11
0Id
entif
iabl
e co
mpo
nent
s 0
13
11
20
00
03
20
0Im
plem
enta
tion
fide
lity
22
3 1
2 2
3 2
22
2 (2)
0
0R
eplic
atio
n 0
00
02
03
00
00
00
0Si
te o
f im
plem
enta
tion
2 2
2 2
1 1
(2)
2 (2
) (2
) 3
2 2
(2)
Tot
al m
etho
dolo
gy
6 12
16
8
11
10
10
8 8
8 15
11
7
5
Not
e. S
= s
tudy
; 0
= n
o ev
iden
ce/n
ot r
epor
ted;
1 =
wea
k ev
iden
ce; 2
= p
rom
isin
g ev
iden
ce;
3 =
str
ong
evid
ence
. Rat
ings
ran
ge f
rom
0 to
3. N
umbe
rs i
n pa
rent
hese
s in
-di
cate
rat
ing
was
inf
erre
d.
391
TA
BL
E 6
. S
um
mar
y of
Evi
denc
e fo
r K
ey M
eth
od
olo
gic
al
Fea
ture
s of
Par
ent
Inv
olv
emen
t S
tud
ies
wit
h S
ingl
e P
arti
cip
ant
and
Mix
ed D
esig
ns
Feat
ure
Stud
y N
umbe
r
SI
S2
S3
S4
S5
S6
S7
S8
S9
S10
Mea
sure
men
t 0
1 3
02
10
22
0Q
ualit
y of
Bas
elin
e/C
ompa
riso
n G
roup
0
03
23
13
01
3M
easu
res
Supp
ort
Prim
ary
Out
com
es a
cros
s Pa
rtic
ipan
ts
00
3 1
1 1
21
22
Edu
catio
nal/C
linic
al S
igni
fica
nce
11
2 1
2 1
22
11
Iden
tifia
ble
Com
pone
nts
0 0
3 0
1 1
2 1
2 2
Impl
emen
tatio
n fi
delit
y 2
1 3
33
32
2 1
2R
eplic
atio
n 0
00
11
00
00
0Si
te o
f im
plem
enta
tion
2 1
(1)
1 3
3 2
2 1
3T
otal
met
hodo
logy
5
4 18
9
16
11
14
10
10
13
Not
e. S
= s
tudy
; 0
= n
o ev
iden
ce/n
ot r
epor
ted;
1 =
wea
k ev
iden
ce; 2
= p
rom
isin
g ev
iden
ce;
3 =
str
ong
evid
ence
. Rat
ings
ran
ge f
rom
0 to
3. N
umbe
rs i
n pa
rent
hese
s in
dica
te r
atin
g w
as i
nfer
red.
392
PARENT INVOLVEMENT 393
tures, present, single-participant, and mixed design studies demonstrated bettermethodology compared with group design studies. Most demonstrated adequatedocumentation of program components (90%), manualized interventions (90%),reported validity and used reliable measures (80%), used multiple assessmentmethods (80% of applicable studies) and visual analysis (80%), and assessed ed-ucational/clinical significance (70%). Additionally, half of single-participant/mixed design studies linked the intervention components to the outcomes, andfour of ten studies obtained measures from multiple sources. Few studies re-ported effect sizes (n = 2). Table 6 presents ratings of key methodological fea-tures for single-participant and mixed design studies. As a group, these studiesevidenced strengths in the categories of treatment fidelity and site of implemen-tation, that is, school-based. All remaining categories of methodology related tothe determination of evidence were strong in fewer than half of the studies.
In sum, across all reviewed studies, methodological strengths were present indocumentation of the program components and manualization or adequate de-scription of program procedures. Consistent methodological weaknesses werefailure to report effect sizes and failure to clearly link the parent involvement in-terventions to the key outcomes.
Effect Sizes
Effect sizes are reported in Tables 7 and 8. A majority of the effect sizes forgroup design studies were large but showed variability indicating a wide rangeof treatment effectiveness. In contrast to group designs, single-participant/mixedstudies had less variation in their effect sizes, with all effect sizes that could becalculated being large (from 1.45-19.04). Several single-participant studies pro-vided visual analysis but not the actual data tables; therefore, many of single-participant effect sizes were calculated using estimates from visually presenteddata and should be interpreted with caution (for a more detailed discussion onthe topic of effect sizes in single-participant designs, see article by L. Guli in thisissue).
Effectiveness of Parent Involvement Interventions: WhatWorks for Whom
Across designs, it appears that there is insufficient evidence to conclude that par-ent involvement, as a method of intervention, is effective. In general, studieswith effective methodology failed to demonstrate significant change in child out-comes, and studies with large effect sizes had flawed methodology. There are afew studies that were judged to be promising based on the combination of highmethodological ratings, significant student outcomes, and large effect sizes. Pro-grams identified as promising used parent tutoring or parent encouragement athome to prevent or change a single academic problem (mathematics or reading)of elementary school-aged children in public schools.
The most promising intervention improved mathematics achievement and
Tab
le 7
. E
ffect
Siz
es f
or
Prim
ary
Ou
tco
me
s fo
r G
rou
p D
esig
n S
tudi
es
Stud
y #
Out
com
e St
atis
tical
pro
cedu
re u
sed
Eff
ect
Size
Val
ue
51
Rea
ding
lev
el
?-te
st
2.76
C
52
Mat
h co
mpu
tatio
n O
ne-w
ay A
NC
OV
A
Ran
ge:
.69C
-.73C
52
Self
-con
cept
of
scho
ol c
ompe
tenc
y, b
ehav
iora
l co
nduc
t, an
d so
cial
acc
epta
nce
One
-way
AN
CO
VA
R
ange
: 1.
00c-
1.05
c
53
Mat
h co
mpu
tatio
n (c
urri
culu
m-b
ased
and
sta
ndar
dize
d)
One
-way
AN
OV
A
Ran
ge: .
86C-1
.63C
54
Aca
dem
ic a
chie
vem
ent
and
perf
orm
ance
(m
ath)
O
ne-w
ay A
NO
VA
R
ange
: .7
4C-
.85C
55
Aca
dem
ic a
chie
vem
ent
and
acad
emic
per
form
ance
(m
ath,
rea
ding
, spe
lling
) N
/A
Insu
ffic
ient
dat
a56
R
eadi
ng c
ompe
tenc
e; s
tude
nt a
cqui
sitio
n of
spe
cifi
c re
adin
g sk
ills
Tw
o in
depe
nden
t sa
mpl
e M
est
Ran
ge:
.10a
-.74
b
57
Rea
ding
T
wo
inde
pend
ent
sam
ple
f-te
st
.23a
58
Rea
ding
ach
ieve
men
t O
ne-w
ay A
NO
VA
R
ange
: . 1
9a—
5.17
C
59
Rea
ding
and
mat
h ac
hiev
emen
t; w
ord
anal
ysis
N
/A
Insu
ffic
ient
dat
a51
0 R
eadi
ng a
nd m
ath
achi
evem
ent;
wor
d an
alys
is
N/A
In
suff
icie
nt d
ata
511
Rea
ding
O
ne-w
ay A
NO
VA
.9
2°51
2 R
eadi
ng p
erfo
rman
ce (
oral
and
com
preh
ensi
on);
rea
ding
ach
ieve
men
t; re
adin
g N
/A
Insu
ffic
ient
dat
aab
ility
; er
ror
rate
in o
ral r
eadi
ng a
nd c
ompr
ehen
sion
513
Wor
d re
cogn
ition
T
wo-
way
AN
OV
A
.92°
514
Eng
lish
lang
uage
rea
ding
N
/A
Insu
ffic
ient
dat
a
Not
e.
aind
icat
es a
sm
all
effe
ct s
ize,
bin
dica
tes
a m
ediu
m e
ffec
t si
ze, a
nd 'i
ndic
ates
a la
rge
effe
ct s
ize
(Div
isio
n 16
and
Soc
iety
for
the
Stud
y of
Sch
ool P
sych
olog
y T
ask
Forc
e, 2
003)
. Eff
ect
size
s fo
rno
nsig
nifi
cant
out
com
es w
ere
not
calc
ulat
ed.
394
PARENT INVOLVEMENT 395
Table 8. Effect Sizes for Primary Outcomes for Single Participant and Mixed DesignStudies
Note. All effect sizes calculated represent large effects. All effect sizes calculated by the authors were calculatedusing Effect SizeMethod 1 (Division 16 and Society for the Study of School Psychology Task Force, 2003).
self-concept in African American fourth and fifth- graders, at risk for mathemat-ics problems, by comparing peer tutoring alone to the combination of peer andparent tutoring (Heller & Fantuzzo, 1993). The combined intervention that in-cluded parent tutoring was more effective. Effect sizes were large, ranging from.86 to 1.63, for changing mathematics achievement. Intervention duration was 8months, with two sessions per week. This study was rated as having mostmethodological features present and strong, or promising evidence in most cate-gories relevant to the determination of evidence except measurement and repli-cation. The study did not use reliable measures or collected data from multiplesources.
In the single-participant category, two promising interventions were identi-fied: Duvall, Delquadri, Elliott, and Hall (1992) and Hook and DuPaul (1999).Both studies examined the effectiveness of parent tutoring in improving readingproblems. Duvall et al. (1992) used in-home parent tutoring of reading with ele-mentary school children ranging from second to fifth grade, most with readingdifficulties, and Hook and DuPaul (1999) evaluated the effects of in-home par-ent tutoring of reading with second- and third-grade children with attention-deficit hyperactivity disorder (ADHD). For both studies, intervention was suc-cessful for all participants and effect sizes were large (1.45-12.98). Gains weremaintained over time and across settings for three of four participants in thestudy by Duvall and colleagues. In Hook and DuPaul's study, maintenance of
Study # Participants Outcome Effect Size
51 hi = 26 Homework performance in math; academic Insufficient dataachievement (math, reading, spelling)
52 7V= 1 Reading achievement Insufficient data53 N = 4 Words read correctly Range: 1.47-4.8253 N = 4 Error rates Insufficient data54 TV = 3 Sounds pronounced correctly Range: 3.00-3.5354 N=3 Reading responses Range: 1.62-2.2855 TV = 4 Words correct (through parent checks) Range: 2.29-7.2955 N=4 Words correct (CBM probes) Range: 1.45-12.9856 TV = 13 Reading achievement N/A - Mixed design57 N=5 Number of known letters Range: 3.53-11.2757 N=5 Letter-naming rate Range: 3.28-19.0458 N= 3 Math performance (home and school) Insufficient data59 N = 1 Spelling performance and accuracy at school Insufficient dataS10 7V=3 Inappropriate behaviors in class Range:-1.92—3.92S10 7V=3 Correct responses in language arts class Range: 1.79-3.49
396 FISHEL AND RAMIREZ
treatment was shown across time for one participant and across settings for twoparticipants.
DISCUSSION
This review critically evaluated the research design, methodological quality, andeffectiveness of 24 studies of parent involvement aimed at improving children'sschool-related learning and behavior. The results of the current review indicatethat there is no conclusive evidence that parent involvement, as a broadly de-fined intervention strategy, is effective in improving academic achievement andbehavior. There are, however, several methodologically sound studies, usingboth single and group designs, that yield promising evidence that one componentof parent involvement, parent home tutoring, improves academic performanceamong elementary school-aged children. Specifically, parent tutoring improvedexisting problems in reading and, in combination with peer tutoring, preventedfurther difficulties in students at risk for mathematics achievement. Unfortu-nately, the lack of methodologically rigorous programs of research on parent in-volvement, broadly defined, seriously compromises the determination of astronger evidence base.
The overall lack of strong evidence to support the effectiveness of parent in-volvement interventions, which results primarily from methodological problemsinherent in the body of literature, is consistent with the conclusions reached byMattingly et al. (2002) in their critical analysis of parent involvement programs.Although the reviewers used somewhat different inclusion criteria for studies,both reviews noted multiple methodological weaknesses in parent involvementstudies, which stand out as the most critical challenge to the determination of anevidence base for the effectiveness of parent involvement interventions. Al-though existing studies are to be commended for their use of manuals or ade-quate description of program procedures, on average, studies failed to demon-strate that significant outcomes were produced by parent involvement activities,to account for family-wise error and unequal groups, to use active controlgroups, and to report follow-up data. Weaknesses in measurement were alsocommon among studies, with many not obtaining information from multiplesources or using valid and reliable instruments. Finally, except for Fantuzzo etal. (1995), replications were absent from the literature.
Another serious challenge to the determination of evidence for the effective-ness of parent involvement stems from insufficient information and data re-ported in studies (also noted by Mattingly et al.), especially in the areas of partic-ipant description, procedures, and results. For example, although most reviewedstudies reported age/grade, disability status, and functional descriptors, manyfewer reported other descriptions of samples and control/treatment groups (i.e.,participants' gender, ethnicity, parent education, or socioeconomic status). Im-portant details were frequently omitted as well in the areas of measurement andstatistical analyses. References to the reliability and validity of measures were
PARENT INVOLVEMENT 397
seldom provided, resulting in lowered ratings of methodological quality for stud-ies with missing values. When reporting results, it is crucial for parent involve-ment studies to include all relevant values and descriptions, such as effect sizes,statistical tests performed, ways to control Type I error, and the rationale for thestatistical procedure employed. Effect sizes are particularly relevant for single-participant studies, where one generally relies on visual analysis to examine thedata. Inclusion of these data would allow for meaningful interpretations and cal-culations as well as cross-study comparisons. Finally, reporting detailed proce-dural descriptions would ensure the ease of future replications.
Determination of the evidence base for parent involvement is also compro-mised by the complexity of the construct and the lack of theory-based researchdesigns that appropriately measure this complexity. Parent involvement has beenvariously and broadly defined, yet studies included in this review tend to meas-ure a single parent involvement activity in a single-component intervention. Amulticomponent design that compares the effectiveness of different parent activ-ities on specific child outcomes and/or compares interventions across differenttreatment delivery agents (e.g., parents and peers) or settings (e.g., community,school, and home) would strengthen the internal validity of conclusions. For sin-gle-participant studies, internal validity would be enhanced by the use of multi-ple baselines to control for within-subject variance in alternating-treatment andsimultaneous-treatment designs. Furthermore, we concur with Mattingly et al.that parent involvement research should be theory driven.
Another important challenge to educators who wish to examine and/or utilizeevidence-based parent involvement programs is the limited scope of existing re-search. Studies in this review used almost exclusively parent-implemented,home-based tutoring treatment interventions to address children's academicproblems. The limited focus of parent involvement research is surprising consid-ering the variety of possible parent involvement activities implemented in con-temporary American schools (including those outlined by Epstein and col-leagues), many of which have been found in descriptive studies to have apositive association with improved academic outcomes. To date we have nomethodologically sound studies that inform educators on which types of parentinvolvement activities have the greatest impact on which school-related behav-iors and achievement, although several researchers have underscored the impor-tance of the issue (Keith et al., 1993; Powell-Smith et al., 2000). Another short-coming of the parent involvement literature is the failure to include high schoolpopulations and treatment implementers other than mothers. It is important to re-search how parent involvement affects secondary school populations (Keith etal., 1993; Falbo et al., 2001). Broadening the participants in parent tutoring stud-ies to include fathers, grandparents, or older siblings would be consistent withthe diversity that characterizes families today.
In light of the aforementioned findings, the authors provide recommendationsfor researchers in the area of parent involvement. It is suggested that researchersapproach their investigation of parent involvement in schools with increased sci-
398 FISHEL AND RAMIREZ
entific rigor. Researchers are further encouraged to increase the scope and com-plexity of studies, while providing theoretical and design links to the specifiedoutcomes of child behaviors and testing for generalizability of results across dif-ferent populations, developmental stages, and school settings. Implementationfidelity should be monitored more closely because occasional integrity checksoften used in the present sample of studies do not always guarantee the satisfac-tory adherence to the procedure. Researchers must carefully select their instru-ments, bearing in mind their validity and reliability with specific populationsunder study. Standardized tests commonly used to measure academic outcomesin group designs may not be well-suited for the specific population under studyor specific variable of interest. Finally, we encourage parent involvement re-searchers to consider parents and practitioners when conducting research. Practi-tioners will benefit, for example, from better procedural descriptions and cost-benefit analyses; parents should be included in program design anddevelopment. Most importantly, despite such apparent technical obstacles as co-ordination/scheduling difficulties and possible high dropout rates, studies initi-ated by school districts rather than researchers are more desirable because suchprojects are likely to become institutionalized.
Several limitations of this review deserve comment. First, specific inclusion/exclusion criteria limited the parent involvement studies that were reviewed. Re-sults may differ with broader inclusion criteria. Second, although the number ofadequately designed studies for inclusion in this review was limited, the parentinvolvement literature base is large and located across multiple databases; de-spite the best efforts of the authors, it is possible that some relevant studies wereoverlooked. Additionally, studies published after March 2003 and those found innon-peer-reviewed sources were not included in the review. Updates to this re-view are encouraged, as it is hoped that more recent studies will demonstratestronger research designs.
Limitations associated with the coding procedure also deserve comment. De-spite numerous revisions, there continues to be a degree of subjectivity in certaincoding decisions required by the Manual that impacts inter-coder reliability.There are also technical ambiguities in the coding procedure, such as in the areasof effect size evaluation (particularly noteworthy when analyzing single-partici-pant and mixed designs), identifiable components, differentiation between cod-ing rubrics, and coding of descriptive features. Of greater concern are notablediscrepancies in the ratings of the effectiveness of interventions for different de-signs in certain rubrics (e.g., measurement and effect size), likely resulting in in-flated ratings of single-participants designs when compared to group designs.Although the Task Force is to be commended for the inclusion of single-subjectcase designs in the determination of evidence for school-based research, and thedevelopment of a parallel coding system, the comparability of ratings across sin-gle-subject, group, and mixed designs is a concern that suggests that our resultsbe considered with caution.
Given the current government policy and its financial ramifications, establish-
PARENT INVOLVEMENT 399
ing a solid evidence base for parent involvement, as it affects school-related out-comes, is of paramount importance for those in education. As there is a seriouslack of evidence for the effectiveness of parent involvement, the need formethodologically rigorous, theory-based investigation of the causal mechanismsthat link the various parent involvement activities with specific child, parent, andschool outcomes remains. The field needs greater numbers of empirical parentinvolvement studies initiated by school districts, representing a variety of de-signs, with solid methodology and diverse target behaviors published in peer-re-viewed journals. It is the authors' hope that this article will spark further re-search and discussion. Researchers are encouraged to build on the present studywith an update and meta-analysis.
References
Arroyo, C, & Zigler, E. (1993). America's Title I/Chapter 1 programs: Why the promise has notbeen met. In E. Zigler & S. Styfco (Eds.), Head Start and beyond: A national plan for ex-tended childhood intervention (pp. 73-96). New Haven, CT: Yale University Press.
Bauch, J. P. (1994). Categories of parent involvement. The School Community Journal, 4, 53-60.Bempechat, J. (1992). The role of parent involvement in children's academic achievement. The
School Community Journal, 2, 31—41.Callahan, K., Rademacher, J. A., & Hildreth, B. L. (1998). The effect of parent participation in
strategies to improve the homework performance of students who are at risk. Remedial andSpecial Education, 19, 131-141.
Chrispeels, J. (1996). Effective schools and home-school-community partnership roles: A frame-work for parent involvement. School Effectiveness and School Improvement, 7, 297-323.
Christenson, S. L. (1995). Supporting home-school collaboration. In A. Thomas & J. Grimes (Eds.),Best practices in school psychology, III (pp. 253-267). Washington, DC: The National Asso-ciation of School Psychologists.
Christenson, S. L., Hurley, C. M., Sheridan, S. M., & Fenstermacher, K. (1997). Parents' and schoolpsychologists' perspectives on parent involvement activities. School Psychology Review, 26,111-130.
Christenson, S. L., Rounds, T., & Gorney, D. (1992). Family factors and student achivement: An av-enue to increase students' success. School Psychology Quarterly, 7, 178-206.
Coates, K. S., & McLaughlin, T. F. (1992). The effect of parent tutoring on oral reading rate withmeasures of clinical significance. British Columbia Journal of Special Education, 16,241-248.
Comer, J. P., & Haynes, N. M. (1991). Parent involvement in school: An ecological approach. TheElementary School Journal, 91, 271-277.
Division 16 and Society for the Study of School Psychology Task Force. Procedural and CodingManual for Evidence Based Interventions. Retrieved June 2, 2003, from www.sp-ebi.org.
Doernberger, C, & Zigler, E. (1993). Project Follow Through: Intent and reality. In E. Zigler & S.Styfco (Eds.), Head Start and beyond: A national plan for extended childhood intervention(pp. 43-72). New Haven, CT: Yale University Press.
Duvall, S. F., Delquadri, J. C , Elliott, M, & Hall, R. V. (1992). Parent-tutoring procedures: Experi-mental analysis and validation of generalization in oral reading across passages, settings, andtime. Journal of Behavioral Education, 2, 281—303.
Epstein, J. L. (1987). Toward a theory of family-school connections: Teacher practices and parent in-volvement. In K. Kurrelmann, F. Kaufmann, & F. Lasel (Eds.), Social intervention: Potentialand constraints (pp. 121-136). New York: De Gruyter.
Epstein, J. L. (1991). Effects on student achievement of teachers' practices of parent involvement. In
400 FISHEL AND RAMIREZ
S. Silvera (Ed.), Advances in reading/language research (Vol. 5, pp. 261-276). Greenwich,CT: JAI Press
Epstein, J. L. (1995). School/family/community partnerships. Phi Delta Kappan, 76, 701-712.Faires, J., Nichols, W. D., & Rickelman, R. J. (2000). Effects of parental involvement in developing
competent readers in first grade. Reading Psychology, 21, 195-215.Falbo, T., Lein, L, & Amador, N. A. (2001). Parental involvement during the transition to high
school. Journal of Adolescent Research, 16, 511—529.Fantuzzo, J. W., Davis, G. Y., & Ginsburg, M. D. (1995). Effects of parent involvement in isolation
or in combination with peer tutoring on student self-concept and mathematics achievement.Journal of Educational Psychology, 87, 272-281.
Gang, D., & Poche, C. E. (1982). An effective program to train parents as reading tutors for theirchildren. Education and Treatment of Children, 5, 211-232.
Grolnick, W. S., & Slowiaczek, M. L. (1994). Parents' involvement in children's schooling: A multi-dimensional conceptualization and motivational model. Child Development, 65, 237-252.
Heller, L. R., & Fantuzzo, J. W. (1993). Reciprocal peer tutoring and parent partnership: Does parentinvolvement make a difference? School Psychology Review, 22, 517-535.
Hook, C. L., & DuPaul, G. J. (1999). Parent tutoring for students with attention-deficit/hyperactivitydisorder: Effects on reading performance at home and school. School Psychology Review, 28,60-75.
Hoover-Dempsey, K. V., Battiato, A. C, Walker, J. M. T., Reed, R. P., DeJong, J. M., & Jones, K. P.(2001). Parental involvement in homework. Educational Psychologist, 36, 195-209.
Jason, L. A., Kurasaki, K. S., Neuson, L., & Garcia, C. (1993). Training parents in a preventive inter-vention for transfer children. Journal of Primary Prevention, 13, 213-227.
Jason, L. A., Weine, A. M., Johnson, J. H., Danner, K. E., Kurasaki, K. S., & Warren-Sohlberg, L.(1993). The school transitions Project: A comprehensive preventive intervention. Journal ofEmotional and Behavioral Disorders, 1, 65-70.
Keith, T. Z., Keith, P. B., Troutman, G. C, Bickley, P. G., Trivette, P. S., & Singh, K. (1993). Doesparental involvement affect eighth-grade student achievement? Structural analysis of nationaldata. School Psychology Review, 22(3), 474^96.
Keith, T. Z., Reimers, T. M, Fehrman, P. G., Pottebaum, S. M., & Aubey, L. W. (1986). Parent in-volvement, homework, and TV time: Direct and indirect effects on high school achievement.Journal of Educational Psychology, 78(5), 373—380.
Law, M, & Kratochwill, T. R. (1993). Paired reading: An evaluation of parent tutorial program.School Psychology International, 14, 119-147.
Lopez, A., & Cole, C. L. (1999). Effects of a parent-implemented intervention on the academicreadiness skills of five Puerto-Rican kindergarten students in an urban school. School Psy-chology Review, 2, 439^47.
Mattingly, D. J., Prislin, R., McKenzie, T. L., Rodriguez, J. L., & Kayzar, B. (2002). Evaluatingevaluations: The case of parent involvement programs. Review of Educational Research, 72,549-576.
Mehran, M., & White, K. R. (1988). Parent tutoring as a supplement to compensatory education forfirst-grade children. Remedial and Special Education, 9, 35^41.
Miller, B. V., & Kratochwill, T. R. (1996). An evaluation of the Paired Reading program using com-petency-based training. School Psychology International, 17(3), 269-291.
Miller, D. L., & Kelley, M. L. (1991). Interventions for improving homework performance: A criti-cal review. School Psychology Quarterly, 6, 174-185.
Natriello, G., & McDill, E. L. (1986). Performance standards, student effort on homework, and aca-demic achievement. Sociology of Education, 59, 18-31.
Powell-Smith, K. A., Stoner, G., Shinn, M. R., & Good, R. H., III. (2000). Parent tutoring in readingusing literature and curriculum materials: Impact on student reading achievement. SchoolPsychology Review, 29, 5-27.
Searls, E. F., Lewis, M. B., & Morrow, Y. B. (1982). Parents as tutors—it works! Reading Psychol-ogy: An International Quarterly, 3, 117-129.
PARENT INVOLVEMENT 401
Shaver, A. V., & Walls, R. T. (1998). Effect of Title I parent involvement on student reading andmathematics achievement. Journal of Research and Development in Education, 31, 90—97.
Shuck, A., Ulsh, F., & Platt, J. S. (1983). Parents encourage pupils (PEP): An innercity parent in-volvement reading project. The Reading Teacher, 36, 524—528.
Smith, E. P., Connell, C. M., Wright, G., Sizer, M., & Norman, J. M. (1997). An ecological model ofhome, school, and community partnerships: Implications for research and practice. Journal ofEducational and Psychological Consultation, 8, 339—360.
Stevenson, H., & Stigler, J. (1992). The learning gap. New York: Summit Books.Thurston, L. P., & Dasta, K. (1990). An analysis of in-home parent tutoring procedures: Effects on
children's academic behavior at home and in school and on parents' tutoring behaviors.RASE: Remedial & Special Education, 11, 41-52.
Toomey, D. (1993). Parents hearing their children read: A review. Rethinking the lessons of theHaringey Project. Educational Research, 35, 223-236.
Trovato, J., & Bucher, B. (1980). Peer tutoring with or without home-based reinforcement for read-ing remediation. Journal of Applied Behavior Analysis, 13, 129—141.
U.S. Department of Education (1996). Goals 2000: Educate America Act. Retrieved May 6, 2004,from www.ed.gov/legislation/GOALS2000/TheAct/index.html.
U.S. Department of Education (2002). No Child Left Behind Act of 2001 (NCLB). Retrieved May 5,2004, from www.ed.gov/policy/elsec/leg/esea02/index.html.
Vinograd-Bausell, C. R., & Bausell, R. B. (1987). Home teaching of word recognition skills. Journalof Research and Development in Education, 20, 57-65.
Walters, K., & Gunderson, L. (1985). Effects of parent volunteers reading first language (LI) booksto ESL students. Reading Teacher, 39, 66-69.
Witt, J. C, Hannafin, M. J., & Martens, B. K. (1983). Home-based reinforcement: Behavioral co-variation between academic performance and inappropriate behavior. Journal of School Psy-chology, 21, 337-348.
Wolfendale, S. (1983). Parental participation in children's development and education. New York:Gordon and Breach.
Zellman, G. L., & Waterman, J. M. (1998). Understanding the impact of parent school involvementon children's educational outcomes. The Journal of Educational Research, 91, 370-380.
Zigler, E., & Styfco, S. (1993). Strength in unity: Consolidating federal education programs foryoung children. In E. Zigler & S. Styfco (Eds.), Head Start and beyond: A national plan forextended childhood intervention (pp. 111-145). New Haven, CT: Yale University Press.
Action Editor: Cindy Carlson
Maria Fishel, M.A., is a School Psychologist at Round Rock Independent School Dis-trict and a doctoral candidate at the School Psychology Program, Department of Educa-tional Psychology of the University of Texas at Austin. Her research interests includeanxiety and depression in children and young adults. Ms. Fishel received her M.A. fromthe University of Texas at Austin. The School Psychology Program at the University ofTexas at Austin is accredited by the APA.
Lucila Ramirez, M.A., is a gradute student in the doctoral School Psychology programat the University of Texas at Austin. Her research interests include treatment of anxiety
402 FISHEL AND RAMIREZ
and depressive disorders in preadolescent girls, as well as Latino parents' attitudes towardtheir children's mental health problems. Ms. Ramirez received her M.A. in Academic Ed-ucational Psychology from the University of Texas at Austin. The School PsychologyProgram at the University of Texas at Austin is accredited by the APA.
