Evaluating the effect of educational media exposure on aggression in early childhood ☆

Abstract

Preschool-aged children (M = 42.44 months-old, SD = 8.02) participated in a short-term longitudinal study investigating the effect of educational media exposure on social development (i.e., aggression and prosocial behavior) using multiple informants and methods. As predicted, educational media exposure significantly predicted increases in both observed and teacher reported relational aggression across time. Follow-up anal-yses showed that educational media exposure also significantly predicted increases in parent reported rela-tional aggression across more than a two year period. Results replicate and extend prior research that has demonstrated links between educational media exposure and relational aggression, but not physical aggres-sion, during early childhood. Over the past fifty years, hundreds of empirical studies have demon-strated that exposure to media influences children's beliefs, attitudes, and behavior (see Gentile, 2003; Roberts & Foehr, 2004). Exposure to vi-olent media during early childhood is considered especially harmful, with results from a meta-analysis (Paik & Comstock, 1994) indicating that individuals of all ages can be influenced by media exposure, though preschoolers showed the largest effect size. There are several possible explanations for this developmental difference (see Gentile & Sesma, 2003). First, learning during this developmental period is especially critical, as younger children are not likely to have incorporated social norms against aggressive behavior (e.g., Huesmann, 1998). Second, younger children have problems differentiating reality from fantasy be-tween two-to five-years of age (e.g., Richert & Smith, 2011). As a result, they are increasingly likely to imitate even the most unrealistic behav-ior patterns. Third, media exposure during early childhood may be an especially salient influence on social relationships because social de-velopment is likely more malleable than in later childhood or adoles-cence and younger children have less control over the activities they engage in compared to older children (Huston, Wright, Marquis, & Green, 1999). Taken together, various cognitive and social factors at this developmental period may make young children more susceptible to effects of media.

Full text

Page 1

This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright

Page 2

Author's personal copy
Evaluating the effect of educational media exposure on aggression in
early childhood☆
Jamie M. Ostrova,⁎, Douglas A. Gentileb, Adam D. Mullinsa
aUniversity at Buffalo, The State University of New York, USA
bIowa State University, USA
a b s t r a c ta r t i c l ei n f o
Article history:
Received 25 March 2012
Received in revised form 5 September 2012
Accepted 23 September 2012
Available online 9 November 2012
Keywords:
Educational media exposure
Relational aggression
Early childhood
Preschool
Preschool-aged children (M = 42.44 months-old, SD = 8.02) participated in a short-term longitudinal study
investigating the effect of educational media exposure on social development (i.e., aggression and prosocial
behavior) using multiple informants and methods. As predicted, educational media exposure significantly
predicted increases in both observed and teacher reported relational aggression across time. Follow-up anal-
yses showed that educational media exposure also significantly predicted increases in parent reported rela-
tional aggression across more than a two year period. Results replicate and extend prior research that has
demonstrated links between educational media exposure and relational aggression, but not physical aggres-
sion, during early childhood.
© 2012 Elsevier Inc. All rights reserved.
Overthepastfiftyyears,hundredsofempiricalstudies havedemon-
strated that exposure to media influences children's beliefs, attitudes,
andbehavior(seeGentile,2003;Roberts&Foehr,2004).Exposuretovi-
olent media during early childhood is considered especially harmful,
with results from a meta-analysis (Paik & Comstock, 1994) indicating
thatindividualsofallagescanbeinfluencedbymediaexposure,though
preschoolers showed the largest effect size. There are several possible
explanations for this developmental difference (see Gentile & Sesma,
2003). First, learning during this developmental period is especially
critical, as younger children are not likely to have incorporated social
norms against aggressive behavior (e.g., Huesmann, 1998). Second,
younger children haveproblemsdifferentiatingrealityfrom fantasybe-
tweentwo- tofive-years of age (e.g.,Richert & Smith, 2011).As a result,
they are increasingly likely to imitate even the most unrealistic behav-
ior patterns. Third, media exposure during early childhood may be an
especially salient influence on social relationships because social de-
velopment is likely more malleable than in later childhood or adoles-
cence and younger children have less control over the activities they
engage in compared to older children (Huston, Wright, Marquis, &
Green, 1999). Taken together, various cognitive and social factors at
this developmental period may make young children more susceptible
to effects of media.
Media exposure
Theories regarding media effects on children and adolescents often
fallintotwogeneralcategories:those associatedwiththeamountof ex-
posure and those associated with the content of the programming (see
Anderson, Huston, Schmitt, Linebarger, & Wright, 2001).
In contrast to amount effects, in which time spent with media dis-
places alternative activities (e.g., physical activities, reading, or sociali-
zation), content effects refer to changes (behavioral, physiological,
social, etc.)due to the messages and behavioral models in the program-
ming of the media being consumed (Anderson et al., 2001). Amount-
and content-specific theories of media effects are valuable for ex-
plaininggeneralmediaeffects;however,researchershaveanintegrated
model of media effects and aggression: the general aggression model.
Andersonandcolleagues(e.g.,Anderson,1997;Anderson,Anderson,
& Deuser, 1996; Anderson & Bushman, 2002) proposed an integrated
model of human aggression, the general aggression model (GAM:
see Carnagey & Anderson, 2003), to describe and predict short- and
long-term increases or decreases in aggressive behavior (Gentile &
Stone, 2005). According to the GAM, input variables, the individual's
present internal state, and outcome variables reciprocally interact to
produce aggressive behavior by priming aggression-related cognitions,
increasing anger-related affective state, and/or increasing arousal
(Lindsay & Anderson, 2000). The GAM can be used to interpret and pre-
dict the effects of most experiences or situations to which an individual
isexposed thatcould resultinaggression.Exposuretoviolentmedia,for
example, has been demonstrated to increase each of the three posited
Journal of Applied Developmental Psychology 34 (2013) 38–44
☆ The contribution of the entire UB Social Development Laboratory staff is greatly
appreciated. Special recognition is given to Jamie L. Guzzo and Christa M. Bishop for
their contributions and assistance with the coordination of this project. We thank the
families, teachers and directors of participating schools. We also thank Dr. Leonard J.
Simms for comments on an earlier version of this manuscript.
⁎ Corresponding author at: 227 Park Hall, Department of Psychology, University at
Buffalo, SUNY, Buffalo, NY 14260-4110, USA. Tel.: +1 716 645 3680.
E-mail address: jostrov@buffalo.edu (J.M. Ostrov).
0193-3973/$ – see front matter © 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.appdev.2012.09.005
Contents lists available at SciVerse ScienceDirect
Journal of Applied Developmental Psychology

Page 3

Author's personal copy
internal state routes — aggressive thoughts, feelings, and affect
(Anderson & Bushman, 2001).
The association between violent media consumption and physically
aggressive and prosocial behavior has been thoroughly investigated in
the literature (see Gentile & Stone, 2005; Huesmann, Moise-Titus,
Podolski, & Eron, 2003). These effects have also been successfully ex-
tended to the study of relational (and indirect) aggression in several
studies (e.g., Coyne & Archer, 2005; Coyne et al., 2008; Gentile,
Mathieson, & Crick, 2011; Möller & Krahé, 2009). Comprehensive
meta-analyses (e.g., Anderson & Bushman, 2001; Anderson et al.,
2010; Paik & Comstock, 1994) have documented short- and long-term
negative effects of violent media exposure, across media types, experi-
mental methodologies and samples. Not surprisingly, the majority of
studies on links between media exposure and aggressive behavior in
children focus on violent media content. In contrast, exposure to pro-
social media (e.g., TV) has documented effects on increasing prosocial
behavior, positive social interactions, and tolerance for others (see
Gentile et al., 2009; Mares& Woodard,2005;Wilson,2008).Education-
al programming is defined as media products that have an explicit in-
tent to education children in a school-related skill such as literacy,
numeracy, as well as social–emotional and character development
domains (Vandewater & Bickham, 2004). Furthermore, educational
media (which are often believed to have prosocial themes) have docu-
mented value for educational outcomes (e.g., Fisch & Truglio, 2001). A
recent meta-analysis documented large effects for links between
prosocial content and altruistic/prosocial behavior (Mares & Woodard,
2005). Interestingly, Mares and Woodard (2005) make the distinction
between purely prosocial or aggressive content from aggressive
prosocial content, in which positive messages follow aggressive actions
orviolentcontent.MaresandWoodard(2005)furtherindicatethatthis
specific type of media content included cases in their meta-analysis in
whichphysical or verbal aggression wasresolved in a prosocial manner
and they found that this category of content was problematic with re-
gard to aggression and other outcomes of interest. The meta-analysis
did not explicitly mention relational aggression, it is likely that similar
processes occur for this behavior as well.
There is also evidence that viewingmedia violencecanleadto forms
of aggression among viewers different from the form viewed, a phe-
nomenon known as the “crossover” effect (see Coyne et al., 2008). Spe-
cifically, several studies have demonstrated that televised physical
aggression can lead to relational aggression in viewers. Huesmann et
al. (2003) found that girls who viewed physical violence on television
as children engaged in more indirect aggression as adults. Ostrov,
Gentile and Crick (2006) also found that young children who viewed
high amounts of physical violence on TV were more relationally aggres-
sive.Inanexperimentaldesign,Coyne,Archer,andEslea(2004)demon-
strated that adolescents who viewed physical aggression were in some
cases subsequently more indirectly aggressive. Linder and Gentile
(2009) also found that exposure to televised physical aggression was
positively associated with teacher reports of indirect aggression in a
sample of fifth grade girls. Gentile, Coyne and Walsh (2011) found
thatchildren'sconsumptionofmediaviolenceearlyinaschoolyearpre-
dicted higher verbal, relational, and physical aggression, as well as less
prosocial behavior later in the school year. Although the GAM model
as originally described by Anderson and his colleagues does not specifi-
cally predict crossover effects, it fits within the model. Media violence
does not have its effects solely through modeling, but also by increasing
aggressivefeelings,arousal,andthoughts.Thesearenotnecessarilyspe-
cific to what was modeled, and can be generalized to new situations. If
children become more willing to aggress, how they aggress will be de-
terminedbymultiplefactors,suchassex(i.e.,boystendtobemorewill-
ing to hit, whereas girls tend to use relational aggression more) or
opportunities (e.g., people tend to be aware of the greater likelihood
of being caught and punished for physical aggression).
Only one study to date, however, has investigated educational
media exposure (EME) and subtypes of aggression (i.e., physical and
relational) in preschoolers (Ostrov et al., 2006). Using a sample of
78 preschoolers, Ostrov et al. (2006) evaluated the effects of media
exposure on aggressive and prosocial behaviors. Interestingly, EME
was associated with future relational aggression. After reviewing
the most frequently reported programs the authors speculated that
many of the educational programs model relationally aggressive be-
havior as a way to frame friendship conflicts (Ostrov et al., 2006).
The reconciliation among the friends occurs at the end of the pro-
grams and given research that suggests that young children have dif-
ficulty understanding plots and connecting content across a program
(Bryant & Anderson, 1983) and given that young children often focus
and retell specific components of the story rather than the overall
general principle (Goldman, Reyes & Varnhagen, 1984), the authors
posited that young children are not attending to the overall educa-
tional lesson, but are instead learning and in turn modeling the rela-
tionally aggressive behaviors (Ostrov et al., 2006).
The hypothesis raised by Ostrov et al. (2006) that young children
may not be learning the social and emotional lesson often found in
educational media is consistent with a recent study by Mares and
Acosta (2008) who examined if kindergarten children (89% were
5-year-olds) could identify the correct moral lesson in an episode
of Clifford the Big Red Dog (a PBS Kids show that was frequently
watched in the Ostrov et al., 2006 study). Interestingly, this study
found that only 19% of the children identified the correct moral lesson
and 89% of the children focused on irrelevant information and misun-
derstood or misinterpreted the intended moral lesson (Mares &
Acosta, 2008). Past research has also found that children have difficul-
ty remembering (i.e., recognition and recall) the motives of TV char-
acters or the consequences of actions within the context of an
action–adventure television program (Collins, Wellman, Keniston, &
Westby, 1978). Collectively, this work suggests that young children
who are exposed to cases of relational aggression within educational
media may not attend to moral lessons of reconciliation or learn con-
flict resolution skills, but rather would focus on and learn about the
aggressive behavior.
Importantly,contentanalysesofchildren'sprogramming(e.g.,Disney
films) have shown that indirect or relational aggression was por-
trayed as high as 9.23 times per hour, although it was primarily
depicted by “bad” characters, which might decrease the reinforcing
value of the content in these cases (Coyne & Whitehead, 2008).
Furthermore, the research on the crossover effect suggests that any
aggressive modeling in children's programming could lead to greater
aggression, although how it is expressed is likely to be moderated by
individual and situational factors (e.g., even young children know
that they will get in trouble if they hit, so they may use relational ag-
gression as it has a lower likelihood of punishment from adults;
Werner, Senich, & Przepyszny, 2006). Given how novel the initial
findings were documenting links between educational media expo-
sure and relational aggression, the originalauthors called forthe rep-
lication and extension of the effect (Ostrov et al., 2006), which is the
central goal of the current study.
Study objectives
Taken together, although the association between media effects
and aggressive thoughts, feelings, and behaviors has been extensively
demonstrated in previous research, significant gaps in the literature
exist. Most notably, research investigating alternate subtypes of ag-
gressive behavior (e.g., relational aggression) in early childhood is
clearly lacking and more focus on educational media content is need-
ed. In order to address these gaps in the literature, the primary goal of
the current study is to evaluate the association between EME and pro-
spective displays of aggressive behaviors in preschoolers. Ostrov et al.
(2006) only demonstrated that educational media was associated
with future relational aggression and in the present study we predict
that educational media will also be associated with increases in
39
J.M. Ostrov et al. / Journal of Applied Developmental Psychology 34 (2013) 38–44

Page 4

Author's personal copy
relational aggression across time. We anticipate that these effects will
be unique to relational aggression and not physical aggression. To
support the robustness of these relations, we further aim to replicate
these effects with both observations and teacher reports of aggressive
behavior. Finding similar effects across the independent informants,
which are known to contribute unique information in the prediction
of various outcomes (e.g., Ostrov et al., 2008), would provide strong
support for our key hypothesis.
An additional improvementon theexistingliterature (e.g.,Ostrov et
al., 2006) is controlling for the alternative subtype of aggression in the
respective models (e.g., initial physical aggression will be controlled in
the relational aggression model). Further, given past associations be-
tween gender, age, and SES with aggressive behavior and media effects
(for reviews see Card et al., 2008; Dodge, Coie & Lynam, 2006; Gentile,
2003) we statistically control for these variables.
Method
Participants
Children were recruited from four nationally accredited, university
affiliated childcare centers in a largecity in the northeast US, participat-
ing in a larger project (e.g., Ostrov et al., 2008). Parents of participating
children were invited to complete a parent questionnaire packet dis-
tributed and collected via US mail, and were compensated with a $10
gift certificate for their time. Of the 75 children participating in the pro-
ject, packets for 47 children (63%) were completed by parents; 43 were
mothers (91.5%) and 4 were fathers (8.5%).
At Time 1, the sample comprised 47 children (17 males and 30 fe-
males), between the ages of 30 and 58 months (M = 42.44 months,
SD=8.02).Thesamplewasrelativelydiverse:Asian(10.6%),Caucasian
(66%), Latino (2.1%), multi-racial (10.6%), Native American (4.3%), and
other/unknown (6.4%). The majority of parents were married (87.2%),
with the remainder being single (8.5%), divorced (2.1%), or in other sit-
uations (2.2%). On average, parents had a four-year degree (ranging
from some high school to a graduate or professional degree) and
mean family income between $55,000 and $100,000 (ranging from
$15,000 to over$100,000),suggestingthechildrenwere from primarily
middle class families. Between time points, seven (four girls) partici-
pants switched schools or moved out of the country, resulting in a
final participant sample at Time 2 of 40 children. An additional two
(onegirl)participantsweremissingmediaexposuredataduetoincom-
pleteparentpackets.Thus, the final sample consisted of 38 children (25
girls).ApoweranalysisusingG*Power3.1.3(Faul,Erdfelder,Buchner,&
Lang, 2009) with α = .05, 1-β = .80, and r2= .37 (based on the Ostrov
et al., 2006 prospective finding between educational media exposure
and relational aggression), suggested we would need 18 participants
for correlations. However, without prior regression models to base our
estimate on, the power analysis for regression models was more diffi-
cult to conduct. We estimated that with α = .05, 1-β = .80, a medium
effect (f2= .20; based in part on Ostrov et al., 2006), and regression
models with six total predictors, we would need a sample size of 42
(Faul et al., 2009). Given the unforeseen circumstances (attrition and
the aforementioned errors) we are slightly underpoweredfor detecting
theseanticipatedmediumeffectsandcautionshouldbeexercisedinthe
interpretation of the findings. Importantly, there were no significant
differences between those that stayed in the study and those that did
not on any of the key study variables, ts b 1.40, n.s.
Procedure
Theprojectwasapproved bythelocalsocialand behavioralsciences
IRB. All children were invited to participate and parental consent was
collected. Observations and teacher reports were conducted at two
time points (fall and spring) with approximately four months in be-
tween. Parent packets were collected during the middle of time 1 (fall).
Measures
Observations of aggression
Children's social interactions were observed and recorded using
the Early Childhood Observation System developed by Ostrov and
Keating (2004) and revised by Crick and colleagues (Crick et al.,
2006) that uses a focal child sampling with continuous recording ap-
proach. One child is randomly chosen from the class roster and ob-
served for ten minutes during free play. Each child was observed
eight times, resulting in 80 min of total observation at each of the
two larger time points (or 160 min per child across the study). No
child was observed more than one time per day. Behavior categories
included physical aggression (e.g., hitting, pushing, kicking) and rela-
tional aggression (e.g., friendship withdrawal threats, excluding child
from playgroup). Behavioral categories were summed to create phys-
ical and relational aggression scores. Previous research has shown ac-
ceptable levels of inter-rater reliability with Intra-Class Correlation
Coefficients (ICCs) ranging from .78 to .91 for physical aggression,
and .70 to .85 for relational aggression (e.g., Ostrov et al., 2006; for
a review see Leff & Lakin, 2005). Observations were conducted by
graduate and undergraduate students who were trained via video-
tapes, vignettes and in-vivo practice reliability observational sessions.
Inter-rater reliability was assessed for 10–15% of total observations
and ICCs were calculated. ICCs with absolute agreement have been
demonstrated to be appropriate given the nature of the current data
(McGraw & Wong, 1996), and have been used with the ECOS in the
past (e.g., Crick et al., 2006; Ostrov & Keating, 2004). ICCs assessing
inter-observer agreement in the current study, across both time
points, for physical aggression and relational aggression were all
greater than .75.
Teacher reports of aggression
The Preschool Social Behavior Scale—Teacher Form (PSBS-TF; Crick,
Casas, & Mosher, 1997) was used to assess children's relational and
physical aggression. This measure contains 16 items including 6 rela-
tional aggression items (e.g., “This child tells a peer they won't be invit-
edtotheirbirthdaypartyunlesss/hedoeswhatthechildwants”),and6
physical aggression items (e.g., “This child kicks or hits others”). Four
additional positively toned filler items are included to avoid a negative
response bias. A 5-point rating scale from 1 (never or almost never true)
to 5 (always or almost always true) was used. The psychometric proper-
ties of this measure have been supported (e.g., Crick et al., 1997, 2006).
Four (two girls) children were missing teacher report data due to in-
complete packets. In the current sample, Cronbach's α's > .80 for all
scales at the two time points.
Parent-reported media exposure
Items evaluating media habits were included in the family infor-
mation questionnaire and come from the MediaQuotient survey
(Gentile & Walsh, 2002). For each media product, parents estimated
how many hours per day their child was engaged with such media,
using a 4-point scale (i.e., 0–1, 1–2, 3–4, 5 or more). Parents rated
how educational they considered the “television/movies” and
“video/computer games” their child played during the school year.
More specifically, for each media product (i.e., television/movies,
video/computer games), parents rated how educational they consid-
ered each media product using a 7-point scale from 1 (Almost
Never) to 7 (Almost Always). An educational exposure score was com-
puted for each medium by multiplying the number of hours spent
with each media product by its subjective educational rating (for sim-
ilar procedures, see Gentile, Coyne et al., 2011; Gentile, Mathieson et
al., 2011). An overall index of educational media exposure (EME) was
calculated by summing the educational rating provided for television/
movies and video/computer games, respectively (Ostrov et al., 2006).
Media researchers have commented that it may be unreasonable to
expect specific media scales, such as educational media indices, to have
40
J.M. Ostrov et al. / Journal of Applied Developmental Psychology 34 (2013) 38–44

Page 5

Author's personal copy
high reliabilities. Gentile and colleagues (2004) noted that it is likely
that some children prefer media differently across various mediums,
and as a result, the specific media index may appear unreliable when
in fact the scale is measuring exactly what it is designed to measure
(Gentile, Lynch, Linder, & Walsh, 2004; Gentile & Walsh, 2002). Given
this recommendation, Cronbach's α was not computed. However,
responses were significantly correlated (e.g., r = .42, p = .004, for the
two items assessing educational exposure).
Importantly, the measure of EME (and similar approaches for vio-
lence exposure; e.g., Gentile, Coyne et al., 2011; Gentile, Mathieson et
al., 2011) has demonstrated acceptable validity and reliability (Ostrov
et al., 2006). For example, Ostrov and colleagues asked parents to pro-
vide the frequency of each child's three favorite television programs
and an inspection of the content of these shows suggests that the
most frequent “programming to which the participants were exposed
was found on PBS, Disney Channel, and Nickelodeon and may be
coded as prosocial or educational in nature” (Ostrov et al., 2006, p.
619). Ostrov et al. (2006) further noted that these programs were
rated as “educational” and were not rated as “violent” by the parents.
Given the similarities in the present sample to the Ostrov et al. (2006)
study with respect to age, ethnicity, SES, and marital status (as well as
the quality of child care center from which the participants were
recruited), and due to space restrictions in our parent packet we did
not ask parents to report on the specific programs. However, parents
uniformly reported that children were exposed to “educational” rath-
er than “violent” media. Violent media exposure was assessed using
the same types of items and procedures as EME (i.e., parents indicat-
ed how violent they considered the television/movies and video/
computer games their child played during the school year on a 7
point scale from 1 “not at all violent” to 7 “extremely violent”). Using
the same procedures as described by Ostrov et al. (2006) the current
measure of violent media exposure was not significantly correlated
with EME, r = .28, p = .11, which demonstrates that parents discrimi-
nated between “educational” and “violent” content. In addition, parents
reported that children were exposed to significantly more educational
media (M = 10.62; SD = 3.65) than violent media (M = 8.54; SD =
4.79),t(34) =2.38,p=.023,d=.49,aswouldbeexpectedforchildren
of this age.
Follow-up assessment
To bolster the validity of the EME, a follow-up assessment
(i.e., Time 3) was conducted. Participants and a parent were invited to
visit the laboratory to complete a small parent report packet for which
they were compensated ($20 gift card). The lag between Time 1 and
Time3assessmentswasslightlymorethantwoyearsandthreemonths
(M = 27.92 months; SD = 1.91; Range = 25.97–31.47). Of the original
38 families, 27 (71%) completed the follow-up study. There were 18
girls and 9 boys and theethnicity percentages were similar to theinitial
time point (14.8% Asian, 66.7% Caucasian, 3.7% Latino, 7.4% multi-racial,
7.4% other races or ethnicities). For one child, a different parent com-
pleted the measures at Time 3 compared to Time 1 (i.e., father partici-
pated at Time 1 and mother participated at Time 2).
Follow-up measures included parent reports of child aggressionand
media exposure.TheChildren'sSocialBehavior—ParentReport(CSB-P),
which was revised from the Children's Social Experience measure used
byCasasetal.(2006)anddevelopedbyCrick(2006),wasusedtoassess
relational and physical aggression. The CSB-P has 13 items on a 5-point
scale ranging from 1 (Never true) to 5 (Almost always true). The
relational aggression subscale contains 5 items (e.g., “spreads rumors,
secrets, or gossips about other kids”), and the physical aggression
subscale has 4 items (e.g., “hits or kicks other kids”). Four additional
prosocial items are included to avoid a negative response bias. Casas
etal.(2006)supportedthevalidityoftheoriginalmeasurewithmoder-
ate associations between mother and father reports for both relational
and physical aggression. Ostrov and Bishop (2008) further supported
the validity of the measure with significant associations between
parent and teacher reports for physical (r = .45, p b .01) and rela-
tional (r = .40, p b .01) aggression. In the present study, the CSB-P
was reliable for physical aggression (Cronbach's α = .71), but was
slightly lower than conventional thresholds for relational aggression
(Cronbach's α = .67) at Time 1. At Time 3, the CSB-P was reliable for
both physical (Cronbach's α = .76) and relational (Cronbach's α =
.73) aggression.
Parents also completed media exposure measures at the follow-up
assessment. Parents first completed the aforementioned measure
from Time 1 assessing EME. Parents also completed theParental Survey
ofMediaExposure(PSME,Ostrovetal.,2006),whichasksparentstore-
port on their child's three favorite television shows and their child's
three favorite movies/videos. For each named media product (i.e., tele-
vision shows aswell asmovies/videos),parentswere asked torate how
“educational”they considereachmediaproducttobeona7-pointscale
(1 = Not at all educational, 7 = Extremely educational) and items were
summed to create an overall index. The majority of shows that were
listed included shows that are educational or informational in nature
with a social and emotional emphasis and are found on channels such
as PBS and Nick Jr. including: Arthur, Caillou, Clifford the Big Red Dog,
Curious George, Franklin, and Reading Rainbow. Additional programs
were reported and included Hannah Montana, Full House, and Sponge
Bob Square Pants,althoughtheseshows generally received low“educa-
tional” ratings by the parents.
The partial correlation (controlling for SES) between general EME
at Time with the PSME (i.e., educational ratings of named programs)
at Time 3 was significant and moderate providing important validity
for the EME used in the present study, rp= .52, p = .026. In addition,
the stability of the EME from times 1 to 3 was significant even when
controlling for SES, rp= .53, p = .023. Collectively, these findings sup-
port the validity of the EME.
Results
Preliminary analyses indicated that there were no skew (b1.73) or
kurtosis (b2.53) concerns (Kline, 2005). There was one outlier for
each of the four observed aggression variables and these were respec-
tively reduced to the magnitude of 3 SD above the mean (Kline,
2005). Descriptive statistics and correlations among the key study vari-
ables are presented in Table 1. Correlations revealed that observations
of physical and relational aggression were not significantly associated
at Time 1 or 2, but correlations between physical and relational aggres-
sion were moderately correlated at times 1 and 2 for teacher reports.
Despite the lack of statistically significant associations among the ob-
served constructs we control for the alternative subtype of aggression
in all the respective models given our a priori goal of testing unique ef-
fects and the knowledge that some of these correlations are in the the-
oretically predicted range despite the lack of significance (likely due to
thesamplesize).Allsubsequentmodelswereruncontrollingforviolent
media exposure and there were no differences between those models
with and without the covariate and so it was removed for simplicity.
Association between educational media exposure and aggressive
behaviors at Time 2
To evaluate the stated objectives, four regression models were
conducted: assessment of the relation between EME and prospective
displays of observed and teacher-reported (in separate models)
(a) relationalaggressionand(b)physicalaggression.Inordertopredict
the respective behavior category (e.g., physical aggression at Time 2),
observations (or teacher reports in separate models) of the behavior
at Time 1 (e.g., physical aggression at Time 1) as well as the alternative
aggression subtype at Time 1, entered the regression model at step 1,
and EME was entered at step 2. In addition, age, gender, and SES were
all entered at step 1 as covariates. As seen in Table 2, EME significantly
41
J.M. Ostrov et al. / Journal of Applied Developmental Psychology 34 (2013) 38–44

Page 6

Author's personal copy
predicted increases in observed relational aggression. It is notable that
theseeffectswithrelationalaggressionwerealsofoundwithteacherin-
formants. EME was not associated with changes in either observed or
teacher-reported physical aggression.
Associations between educational media exposure and aggressive
behaviors at Time 3
To test if the effects held beyond the academic year, two regression
models were conducted that examined if EME at Time 1 was associated
with increases in parent reported relational aggression from times 1 to
3 (see Table 3). Given the small sample size in the follow-up portion of
the study as well as shared method variance concerns (i.e., only parent
reportwasavailableatbothtimepoints),cautionshouldbeexercisedin
the interpretation of the findings. Consistent with the aforementioned
findings, EME predicted increases in parent reported relational aggres-
sionacross onaveragetwoyearsandthreemonths,evenwhencontrol-
ling for initial physical aggression, age, gender, and SES. In addition, as
shown in Table 3, EME did not significantly predict changes in parent
reported physical aggression.
Discussion
The present study tested the association between educational
media exposure (EME) and subtypes of aggression in early childhood
using observational methodology. Results of this research provide ev-
idence that EME can predict the exhibition of relational aggression in
preschool children. The current study replicated past research
(Ostrov et al., 2006) in which EME significantly predicted future rela-
tional aggression in preschoolers, but extended the past findings by
revealing that EME was associated with increases in relational aggres-
sion. Moreover, our conservative model also controlled for initial
physical aggression, as well as SES, gender, and age. In addition, we
are also the first to demonstrate these effects using multiple methods
and informants. Although most research on EME has documented a
significant effect on positive academic and social outcomes, these re-
sults suggest that EME may simultaneously have a detrimental effect
on children's social behavior. Ostrov et al. (2006) posited that it is
possible that children may be exposed to relationally aggressive
models within these programs and may not comprehend the conflict
resolution skills that typically are depicted at the end of the program,
(or may not understand how they relate to the earlier conflict) rather
the young children focus on and learn the modeled behaviors. In fact,
younger children tend to pay greater attention to perceptually-salient
information (e.g., action, music; Schmitt, Anderson, & Collins, 1999),
relativetoolderchildren,whoattendmoretocuesthatareplot-relevant
(Calvert,Huston,Watkins,&Wright,1982).Futureexperimentalstudies
are needed to test the hypothesis that children are modeling behaviors
from peer conflict scenarios seen in educational programs.
Despite the contributions of the current study, there are several
limitations. First, the limited sample size reduced the power to find
significant results. Although the sample size was similar to that used
in other observational studies of aggressive behavior in early childhood
(e.g.,McEvoy,Estrem,Rodriguez&Olson,2003;Ostrov&Keating,2004;
Stauffacher & DeHart, 2005) and was generally consistent with power
analysis recommendations, attrition between time points constrained
the study's power. Clearly, replication with larger, ethnically diverse
samples is needed. Second, even though the follow-up analyses
suggested moderate levels of agreement between the general EME
and PSME ratings in which parents named programs, and most of the
programs listed are generally believed and rated by the media industry
tobe“educationalandinformational,”itisunclearhowwellparentscan
evaluate the educational value of children's media. More importantly,
perhaps, future studies should measure both the perceived educational
value and the amount of relational aggression in each media product
(although it is similarly unclear how well parents can judge that). In
Table 1
Descriptive statistics and correlations for key study variables.
123456789
M SD
Range
1. RA-O T1
2. PA-O T1
3. RA-TR T1
4. PA-TR T1
5. RA-O T2
6. PA-O T2
7. RA-TR T2
8. PA-TR T2
9. EME-P T1
–
.03 .30
.04
.13
.15
.55⁎⁎⁎
.52⁎⁎⁎
-.004
−.06
.01
–
.25
.42⁎⁎
−.01
.52⁎⁎⁎
.30
–
.47⁎⁎
−.20
.59⁎⁎⁎
.37⁎
.38⁎
.07
–
.09
−.15
−.03
−.09
.07
.21
.12
.01
.07
–
0.94
2.54
10.74
10.47
2.43
1.66
12.27
9.83
10.62
1.19
2.60
5.42
4.89
3.07
2.16
5.88
4.83
3.65
0.00–5.00
0.00–10.00
6.00–23.00
6.00–22.00
0.00–12.17
0.00–8.57
6.00–25.00
6.00–20.00
2.00–20.00
–
−.01
.29
.70⁎⁎⁎
−.01
.39⁎
.38⁎
–
–
–
Note. RA = Relational Aggression; PA = Physical Aggression; EME = Educational Media Exposure; O = Observation; TR = Teacher Report; P = Parent Report; T1 = Time 1; T2 =
Time 2.
⁎ p b .05.
⁎⁎ p b .01.
⁎⁎⁎ p b .001.
Table 2
Hierarchical multiple regressions: Associations between educational media exposure
and aggressive behavior at Time 2.
Outcome, step, predictors
β
F, ΔFR2, ΔR2
I. Relational aggression T2 (O)
1. Gender
Age
SES
Relational aggression T1 (O)
Physical aggression T1 (O)
2. EME T1 (parent report)
II. Physical aggression T2 (O)
1. Gender
Age
SES
Relational aggression T1 (O)
Physical aggression T1 (O)
2. EME T1 (parent report)
III. Relational aggression T2 (TR)
1. Gender
Age
SES
Relational aggression T1 (TR)
Physical aggression T1 (TR)
2. EME T1 (parent report)
IV. Physical aggression T2 (TR)
1. Gender
Age
SES
Relational aggression T1 (TR)
Physical aggression T1 (TR)
2. EME T1 (parent report)
.06
.20
F(5, 33) = 1.25, n.s.
.159
−.07
.28
.07
.34⁎
ΔF(1, 32) = 4.41, p = .04.102
−.28
−.24
.16
.37⁎
.35⁎
.14
F(5, 33) = 4.51, p = .003.406
ΔF(1, 32) = 0.94, n.s.
.017
.28
F(5, 29) = 7.38, p b .001.560
−.05
−.18
.42⁎
.30
.25⁎
ΔF(1, 28) = 4.11, p = .05 .056
.05
F(5, 29) = 5.48, p = .001.486
−.06
.11
−.03
.69⁎⁎
.07
ΔF(1, 28) = 0.27, n.s.
.005
Note. O = observations; TR = teacher report; T1 = Time 1; T2 = Time 2; EMI =
educational media exposure. ⁎ p b b.05. ⁎⁎ p b b.01.
42
J.M. Ostrov et al. / Journal of Applied Developmental Psychology 34 (2013) 38–44

Page 7

Author's personal copy
addition, thefindings from the follow-upportion of the study should be
interpreted with caution given the small sample size and relatively
small number of boys that participated. Third, although the format
was similar to that used in previous studies (e.g., Anderson & Dill,
2000), social desirability could certainly have influenced the parental
ratings in the present study and thus not only should future studies ex-
amine for and possibly control for these biases, but it may also be best
for future studies to conduct content analyses of the nominated
shows, similar to the approach taken by Linder and Gentile (2009)
with older school-aged children. Future studies are needed that incor-
porate multiple measures of media exposure (e.g., parent report,
media diaries, specific program content analyses) from multiple infor-
mants (e.g., children, siblings, parents, and peers) across mediums
(e.g., television, video games, movies, and music) in order to determine
the “gold standard” for assessing media exposure. Moreover, future re-
search should specifically assess the actual TV programs, movies, and
video games that children are exposed to, the relative amounts of
time that they are exposed to each program, and expert raters should
be used to assess the specific content of each show, movie, and
videogame.
It is certainly possible that children who consume high levels of
media receive less scaffolding from their parents with respect to how
to navigate social relationships. In fact, we echo the calls of others that
suggest that parents not just co-view, but actively mediate the content
of themedia (Warren, 2003). This mediationmay allow young children
the assistance that they need to appropriately connect the aggressive
behaviorsandfriendshipconflictswiththemorallessonoftheprogram.
We believe that this process requires active mediation and not just
co-viewing (Warren, 2003), as co-viewing without the active engage-
ment and scaffolding of the content may imply tacit approval of the ag-
gressive behaviors. Perhaps relationally aggressive behavior would
decline if parents actively mediated the educational programs with
their young children and helped them to make the connections be-
tween the relationally aggressive behaviors and conflict resolution
strategies. The fact that children are not following the plot line and
learning the character development lesson in media programs and
may just be attending to reinforced (i.e., the character gets what they
want)relationallyaggressivebehaviorsthattheyseedisplayedsuggests
the need for simpler and more explicit lessonsfor youngchildren's pro-
gramming (Mares & Acosta, 2008). A second implication, however, is
that even shows that are considered to be child-friendly educational
shows can have negative consequences if they model and reinforce re-
lational or indirect aggression(Linder & Gentile, 2009). The currentrat-
ings (and the associated V-chip) do not discuss this type of information
(Linder & Gentile), which leaves parents with limited information for
making informed decisions for their children.
In conclusion, the present study documented that exposure to
media that parents label as educational is prospectively associated
with increases in relational aggression over time. These effects were
replicated using three independent measures of relational aggression
(i.e., observations, teacher reports, and parent reports) and after con-
trolling for physical aggression, age, gender, and SES. In keeping with
past work and current predictions the links were limited to relational
aggressionandnotphysicalaggression.Thesefindingssuggestthatpar-
ents, educators, media professionals, policymakers, and researchers
should work collaboratively to reduce these potential harmful effects
for young children.
References
Anderson, C. A. (1997). Effects of violent movies and trait hostility on hostile feelings
and aggressive thoughts. Aggressive Behavior, 23, 161–178.
Anderson, C. A., Anderson, K. B., & Deuser, W. E. (1996). Examining an affective aggres-
sion framework: Weapon and temperature effects on aggressive thoughts, affect,
and attitudes. Personality and Social Psychology Bulletin, 22, 366–376.
Anderson, C. A., & Bushman, B. J. (2001). Effects of violent video games on aggressive be-
havior, aggressive cognition, aggressive affect, physiological arousal, and prosocial
behavior: A meta-analytic review of the scientific literature. Psychological Science,
12, 353–359.
Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of Psychology,
53, 27–51.
Anderson, C. A., & Dill, K. E. (2000). Video games and aggressive thoughts, feelings, and
behavior in the laboratory and real life. Journal of Personality and Social Psychology,
78, 772–790.
Anderson, D. R., Huston, A. C., Schmitt, K., Linebarger, D. L., & Wright, J. C. (2001). Early
childhood television viewing and adolescent behavior. Monographs of The Society
for Research on Child Development, 66, (1, Serial No. 264). (pp. 1–147).
Anderson, C. A., Shibuya, A., Ihori, N., Swing, E. L., Bushman, B. J., Sakamoto, A., &
Saleem, M. (2010). Violent video game effects on aggression, empathy, and
prosocial behavior in eastern and western countries: A meta-analytic review.
Psychological Bulletin, 136(2), 151–173, http://dx.doi.org/10.1037/a0018251.
Bryant, J., & Anderson, D. R. (Eds.). (1983). Children's understanding of TV: Research on
attention and comprehension. New York, NY: Academic Press.
Calvert, S. L., Huston, A. C., Watkins, B. A., & Wright, J. C. (1982). The relations between
selective attention to television forms and children's comprehension of content.
Child Development, 53, 601–610.
Card, N. A., Stucky, B. D., Sawalani, G. M., & Little, T. D. (2008). Direct and indirect ag-
gression during childhood and adolescence: A meta-analytic review of gender dif-
ferences, intercorrelations, and relations to maladjustment. Child Development, 79,
1185–1229.
Carnagey, N. L., & Anderson, C. A. (2003). Theory in the study of media violence: The gen-
eralaggressionmodel.InD.A.Gentile(Ed.),Mediaviolenceandchildren.Westport,CT:
Praeger.
Casas, J. F., Weingel, S. M., Crick, N. R., Ostrov, J. M., Woods, K. E., Jansen Yeh, E. A., &
Huddleston-Casas, C. A.(2006). Early parenting and children'srelational and physical
aggression in the preschool and home contexts. Journal of Applied Developmental
Psychology, 27, 209–227, http://dx.doi.org/10.1016/j.appdev.2006.02.003.
Collins, W. A., Wellman, H., Keniston, A. H., & Westby, S. D. (1978). Age-related aspects of
comprehensionand inferencefrom a televiseddramaticnarrative.Child Development,
49, 389–399.
Coyne, S. M., & Archer, J. (2005). The relationship between indirect and physical ag-
gression on television and in real life. Social Development, 2, 324–338.
Coyne, S. M., Archer, J., & Eslea, M. (2004). Cruel intentions on television and in real life:
Can viewing indirect aggression increase viewers' subsequent indirect aggression?
JournalofExperimentalChildPsychology,88,234–253,http://dx.doi.org/10.1016/j.jecp.
2004.03.001.
Coyne, S. M., Nelson, D. A., Lawton, F., Haslam, S., Rooney, L., Titterington, L., & Ogunlaja,
L. (2008). The effects of viewing physical and relational aggression in the media:
Evidence for a cross-over effect. Journal of Experimental Social Psychology, 44,
1551–1554, http://dx.doi.org/10.1016/j.jesp. 2008.06.006.
Coyne, S. M., & Whitehead, E. (2008). Indirect aggression in animated Disney films.
Journal of Communication, 58, 382–395.
Crick, N. R. (2006). Children's Social Behavior—Parent Report. Unpublished manual,
University of Minnesota.
Crick, N. R., Casas, J. F., & Mosher, M. (1997). Relational and physical aggression in pre-
school. Developmental Psychology, 35, 376–385.
Crick, N. R., Ostrov, J. M., Burr, J. E., Cullerton-Sen, C., Jansen-Yeh, E. A., & Ralston, P. (2006). A
longitudinal study of relational and physical aggression in preschool. Journal of Applied
DevelopmentalPsychology,27,254–268,http://dx.doi.org/10.1016/j.appdev.2006.02.006.
Dodge,K.A.,Coie,J.D.,&Lynam,D.(2006).Aggressionandantisocialbehaviorinyouth.In
W. Damon (Series Ed.) & N. Eisenberg (Vol. Ed.), Handbook of Child Psychology: Vol. 3.
Social, emotional, and personality development (6th ed., pp. 719–788). New York, NY:
Wiley.
Faul, F., Erdfelder, E., Buchner, A., & Lang, A. -G. (2009). Statistical power analysis using
G*Power 3.1: Tests for correlation and regression analyses. Behavior Research
Methods, 41, 1149–1160, http://dx.doi.org/10.3758/BRM.41.4.1149.
Fisch, S. M., & Truglio, R. T. (2001). “G” is for growing: Thirty years of research on children
and Sesame Street. Mahwah, NJ: Lawrence Erlbaum Associates.
Table 3
Hierarchical multiple regressions: Associations between educational media exposure
and aggressive behavior at Time 3.
Outcome, step, predictors
β
F, ΔFR2, ΔR2
V. Relational aggression T3 (P)
1. Gender
Age
SES
Relational aggression T1 (P)
Physical aggression T1 (P)
2. EME T1 (parent report)
VI. Physical aggression T3 (P)
1. Gender
Age
SES
Relational aggression T1 (P)
Physical aggression T1 (P)
2. EME T1 (P)
.21
F(5, 17) = 2.37, n.s.
.41
−.29
−.34
.22
.16
.49⁎
ΔF(1, 16) = 7.06, p = .017.18
.08
.02
F(5, 18) = 3.63, p = .019.50
−.15
−.11
.69⁎⁎
.17
ΔF(1, 17) = 0.80, n.s.
.02
Note. P = parent report; T1 = Time 1; T3 = Time 3; EME = educational media expo-
sure. ⁎ p b b.05. ⁎⁎ p b b.01.
43
J.M. Ostrov et al. / Journal of Applied Developmental Psychology 34 (2013) 38–44

Page 8

Author's personal copy
Gentile, D. A. (2003). Media violence and children. Westport, CT: Praeger.
Gentile, D. A., Anderson, C. A., Yukawa, S., Ihori, N., Saleem, M., Ming, L. K.,...Sakamoto,
A. (2009). The effects of prosocial video games on prosocial behaviors: Internation-
al evidence from correlational, longitudinal, and experimental studies. Personality
and Social Psychology Bulletin, 35, 752–763, http://dx.doi.org/10.1177/0146167
209333045.
Gentile, D. A., Coyne, S., & Walsh, D. A. (2011). Media violence, physical aggression, and
relational aggression in school age children: A short-term longitudinal study.
Aggressive Behavior, 37, 193–206, http://dx.doi.org/10.1002/ab.20380.
Gentile, D. A., Mathieson, L. C., & Crick, N. R. (2011). Media violence associations with
the form and function of aggression among elementary school children. Social
Development, 20, 213–232, http://dx.doi.org/10.1111/j.1467-9507.2010.00577.x.
Gentile, D. A., Lynch, P. J., Linder, J. R., & Walsh, D. A. (2004). The effects of violent video
game habits on adolescent hostility, aggressive behaviors, and school performance.
Journal of Adolescence, 27, 5–22.
Gentile, D. A., & Sesma, A. (2003). Developmental approaches to understanding media
effects on individuals. In D. A. Gentile (Ed.), Media violence and children
(pp. 19–37). Westport, CT: Ablex.
Gentile, D.A., & Stone, W. (2005). Violent video game effects on childrenand adolescents:
A review of the literature. Minerva Pediatrica, 57, 337–358.
Gentile, D. A., & Walsh, D. A. (2002). A normative study of family media habits. Journal
of Applied Developmental Psychology, 23, 157–178.
Goldman, S. R., Reyes, M., & Varnhagen, C. K. (1984). Understanding fables in first and
second languages. NABE Journal, 8, 35–66.
Huesmann, L. R. (1998). The role of social information processing and cognitive
schemas in the acquisition and maintenance of habitual aggressive behavior. In
R. G. Geen, & E. Donnerstein (Eds.), Human aggression: Theories, research, and impli-
cations for social policy (pp. 1120–1134). San Diego, CA: Academic Press.
Huesmann, L. R., Moise-Titus, J., Podolski, C. L., & Eron, L. D. (2003). Longitudinal re-
lations between children's exposure to TV violence and their aggressive and vio-
lent behavior in young adulthood: 1977–1992. Developmental Psychology, 39,
201–221.
Huston, A. C., Wright, J. C., Marquis, J., & Green, S. B. (1999). How young children spend
their time: Television and other activities. Developmental Psychology, 35, 912–925.
Kline, R. B. (2005). Principles and Practice of Structural Equation Modeling (2nd Ed).
New York, NY: Guilford Press.
Leff, S. S., & Lakin, R. (2005). Playground-based observational systems: A review and
implications for practitioners and researchers. School Psychology Review, 34,
475–489.
Linder, J. R., & Gentile, D. A. (2009). Is the television rating system valid? Indirect, ver-
bal, and physical aggression in programs viewed by fifth grade girls and associa-
tions with behavior. Journal of Applied Developmental Psychology, 30, 286–297,
http://dx.doi.org/10.1016/j.appdev.2008.12.013.
Lindsay, J. J., & Anderson, C. A. (2000). From antecedent conditions to violent actions:
A general affective aggression model. Personality and Social Psychology Bulletin,
26, 533–547.
Mares, M. -L., & Acosta, E. E. (2008). Be kind to the three-legged dogs: Children's literal
interpretations of TV's moral lessons. Media Psychology, 11, 377–399, http:
//dx.doi.org/10.1080/15213260802204355.
Mares, M. -L., & Woodard, E. (2005). Positive effects on children's social interactions:
A meta-analysis. Media Psychology, 7, 301–322.
McEvoy, M. A., Estrem, T. L., Rodriguez, M. C., & Olson, M. L. (2003). Assessing relational
and physical aggression among preschool children: Intermethod agreement. Topics
in Early Childhood Special Education, 23, 53–63.
McGraw, K. O., & Wong, S. P. (1996). Forming inferences about some intra-class corre-
lation coefficients. Psychological Methods, 1, 30–46.
Möller, I., & Krahé, B. (2009). Exposure to violent video games and aggression in
German adolescents: A longitudinal analysis. Aggressive Behavior, 35, 75–89,
http://dx.doi.org/10.1002/ab.20290.
Ostrov, J. M., & Bishop, C. M. (2008). Preschoolers' aggression and parent–child conflict: A
multi-informantandmulti-methodstudy.JournalofExperimentalChildPsychology,99,
309–322, http://dx.doi.org/10.1016/j.jecp.2008.01.001.
Ostrov, J. M., Gentile, D. A., & Crick, N. R. (2006). Media exposure, aggression, and prosocial
behaviorduringearlychildhood:Alongitudinalstudy.SocialDevelopment,15,612–627.
Ostrov, J. M., & Keating, C. F. (2004). Gender differences in preschool aggression during
free play and structured interactions: An observational study. Social Development,
13, 255–277.
Ostrov, J. M., Ries, E. E., Stauffacher, K., Godleski, S. A., & Mullins, A. D. (2008). Relational
aggression, physical aggression and deception during early childhood: A
multimethod, multi-informant short-term longitudinal study. Journal of Clinical
Child and Adolescent Psychology, 37, 664–675, http://dx.doi.org/10.1080/1537441
0802148137.
Paik, H., & Comstock, G. (1994). The effects of television violence on antisocial behavior:
A meta-analysis. Communication Research, 21, 516–546.
Richert, R. A., & Smith, E. I. (2011). Preschoolers' quarantining of fantasy stories. Child
Development, 82, 1106–1119, http://dx.doi.org/10.1111/j.1467-8624.2011.01603.x.
Roberts, D. F., & Foehr, U. G. (2004). Kids and media in America. New York, NY:
Cambridge University Press.
Schmitt, K. L., Anderson, D. R., & Collins, P. A. (1999). Form and content: Looking at vi-
sual features of television. Developmental Psychology, 35, 1156–1167.
Stauffacher, K., & DeHart, G. (2005). Preschoolers' relational aggression with siblings
and with friends. Early Education and Development, 16, 185–206.
Vandewater, E. A., & Bickham, D. S. (2004). The impact of educational television on young
children's reading in the context of family stress. Journal of Applied Developmental
Psychology, 25, 717–728, http://dx.doi.org/10.1016/j.appdev.2004.09.006.
Warren, R. (2003). Parental mediation of preschool children's television viewing. Journal
of Broadcasting & Electronic Media, 47, 394–417.
Werner, N. E., Senich, S., & Przepyszny, K. A. (2006). Mothers' responses to pre-
schoolers' relational and physical aggression. Journal of Applied Developmental
Psychology, 27, 193–208.
Wilson, B. J. (2008). Media and children's aggression, fear, and altruism. The Future of
Children, 18, 87–118.
44
J.M. Ostrov et al. / Journal of Applied Developmental Psychology 34 (2013) 38–44