The Impact of Background Television on Parent-Child Interaction

Abstract

This study investigated the hypothesis that background television affects interactions between parents and very young children. Fifty-one 12-, 24-, and 36-month-old children, each accompanied by 1 parent, were observed for 1 hr of free play in a laboratory space resembling a family room. For half of the hour, an adult-directed television program played in the background on a monaural television set. During the other half hour, the television was not on. Both the quantity and quality of parent-child interaction decreased in the presence of background television. These findings suggest one way in which early, chronic exposure to television may have a negative impact on development.

Full text

The Impact of Background Television on Parent–Child Interaction
Heather L. Kirkorian, Tiffany A. Pempek, Lauren A. Murphy, Marie E. Schmidt,
and Daniel R. Anderson
University of Massachusetts, Amherst
This study investigated the hypothesis that background television affects interactions between parents and
very young children. Fifty-one 12-, 24-, and 36-month-old children, each accompanied by 1 parent, were
observed for 1 hr of free play in a laboratory space resembling a family room. For half of the hour, an adult-
directed television program played in the background on a monaural television set. During the other half
hour, the television was not on. Both the quantity and quality of parent–child interaction decreased in the
presence of background television. These findings suggest one way in which early, chronic exposure to tele-
vision may have a negative impact on development.
Very young children are frequently present when
parents or other family members watch programs
intended for adults. Although infants and toddlers
ordinarily pay low levels of overt attention to adult
programs (Schmidt, Pempek, Kirkorian, Lund, &
Anderson, 2008; Schmitt, Anderson, & Collins,
1999), it is possible that these programs can exert a
developmentally important influence. The amount
of exposure is likely substantial: Over one third of
children under age 3 years live in homes that care-
takers characterize as having a television on ‘‘most
of the time’’ or ‘‘always,’’ even if no one is watching
(Rideout & Hamel, 2006; Rideout, Vandewater, &
Wartella, 2003). Moreover, a substantial proportion
of very young children’s solitary toy play and play
with parents occurs in the presence of television
(Masur & Flynn, 2008; Schmitt, Woolf, & Anderson,
2003). This type of exposure to television has been
referred to as ‘‘background television.’’ Anderson
and Evans (2001) hypothesized that background
television serves as an audiovisual distractor that
could in principle disrupt young children’s ongoing
behavior such as toy play. They also hypothesized
that it could be a disruptive influence on parent–
child interactions.
A recent experiment supports the first hypothe-
sis. Schmidt et al. (2008) observed 12-, 24-, and
36-month-olds’ solitary toy play both with and
without background television. When the television
was on, the children exhibited shorter play epi-
sodes and less focused attention than they did
when the television was off. These findings indicate
that adult-directed television programs may func-
tion as dynamic audiovisual distractors for very
young wchildren. Using a similar experimental
design with a different sample, the present experi-
ment tests Anderson and Evans’s (2001) second
hypothesis that background television influences
parent–child interactions. Television that is in the
background for the child may effectively be in the
foreground for the parent, insofar as the parent
chose the program to watch, and thus may be diffi-
cult for the parent to ignore.
There has been little prior research on the effects
of background television, primarily because
researchers have rarely distinguished between
exposure to children’s programs and exposure to
adult-directed programs. Rather, the typical investi-
Tiffany A. Pempek is now at Otterbein College in Westerville,
OH. Lauren A. Murphy is now at Portland State University,
Portland, OR. Marie Evans Schmidt is now at the Center on
Media and Child Health, Children’s Hospital, Boston, MA.
Aspects of this research were presented at the meetings of the
Society for Research in Child Development and the International
Communication Association. This research was supported by
grants from the National Science Foundation (BCS-0111811, BCS-
0519197). Findings and opinions expressed in this manuscript do
not reflect endorsement by the National Science Foundation. We
wish to acknowledge the helpful comments of Elizabeth Harvey
and Erica Scharrer. We also thank the videotape coders for this
project: Kesina Gray, Brittany Hutton, Alexis Lauricella, Angie
Naniot, and Sarah Rudolv.
Correspondence concerning this article should be addressed to
Daniel R. Anderson, Department of Psychology, University of
Massachusetts at Amherst, Amherst, MA 01003. Electronic mail
may be sent to anderson@psych.umass.edu.
Child Development, September/October 2009, Volume 80, Number 5, Pages 1350–1359
2009, Copyright the Author(s)
Journal Compilation 2009, Society for Research in Child Development, Inc.
All rights reserved. 0009-3920/2009/8005-0004

gation simply asks parents how much their
children ‘‘watch’’ television without defining what
is meant by this term. Given that caveat, correla-
tional studies suggest that early exposure to televi-
sion is associated with negative outcomes (e.g.,
Carew, 1980; Christakis, Zimmerman, DiGiuseppe,
& McCarty, 2004; Nelson, 1973) and some research
suggests that this relation may be greatest for, or
perhaps limited to, television exposure during the
first few years of life (Zimmerman & Christakis,
2005).
Importance of Parent–Child Interaction During Early
Childhood
There is a vast literature examining environmen-
tal factors that affect cognitive and social develop-
ment. In broad terms, a sensitive, stimulating, and
otherwise high-quality home is positively associ-
ated with cognition, achievement, language, and
social competence in children (National Institute of
Child Health and Human Development Early Child
Care Research Network, 2003). Among the most
important influences in development are social
interactions with family, peers, teachers, and others.
Vygotsky (1978) emphasized the role of social inter-
action in cognitive development. He observed that
children can master more sophisticated tasks when
working jointly with adults or more advanced
peers than when playing independently.
Perhaps not surprisingly, much of the research
on early environmental influences on children
focuses on parent–child interaction. Compared to
play without parents, experimental studies find
that children engage in more independent goal-
directed play (Landry, Miller-Loncar, & Swank,
1998), higher qualitative levels of play (Aless-
andri, 1992), and increased focused attention
(Lawson, Parrinello, & Ruff, 1992) during or
immediately following interactive play with par-
ents. Parents’ verbal stimulation may be particu-
larly important for cognitive competence and
language ability (Clarke-Stewart, 1973; Olson,
Bates, & Bayles, 1984). In fact, Olson et al. (1984)
suggest that verbal stimulation alone may
account for socioeconomic status as a predictor
of child competence.
The mere presence of parents, although certainly
important, does not produce optimum conditions
for development. Rather, children flourish when
parents are actively involved in their children’s
play. In an experiment by Slade (1987), toddlers
exhibited substantially longer play episodes and
more sophisticated play when mothers made sug-
gestions and were physically involved compared to
when mothers only passively commented on child-
initiated verbal interactions. Notably, both types of
interactions elicited longer and more mature play
episodes than the absence of interaction. Others
have found an association between the quality of
maternal involvement and children’s concurrent
play maturity (Alessandri, 1992) and subsequent
social and language development (Parks & Bradley,
1991). Equally important is parents’ sensitivity to
their children’s needs and activities. For instance,
parents’ contingent responsiveness to children’s
bids for attention may be particularly important
(e.g., Beckwith & Rodning, 1996; Bornstein &
Tamis-LeMonda, 1989; Lawson et al., 1992; Lewis,
1993; Tamis-Lemonda, Bornstein, Baumwell, &
Damast, 1996).
Background Television and Parent–Child Interaction
One way that background television may affect
parent–child interaction is through interference
via background noise. Although there is some
evidence suggesting that performance on certain
tasks is actually enhanced by ambient noise,
background noise containing speech (compared
to white noise or noise produced by machines)
can be detrimental to performance, particularly
on complex cognitive tasks (Baker & Holding,
1993). Similarly, Armstrong and Greenberg (1990)
found that background television interferes with
complex problem-solving tasks in adults, most
likely by diverting cognitive resources. To the
degree to which interacting with a child is a
complex activity requiring one’s attention, it can
be supposed that television may reduce the qual-
ity of that interaction. There is some evidence
that background noise in the home (of which
television is a major contributor) is negatively
related to maternal responsiveness (Corapci &
Wachs, 2000), possibly because background noise
interferes with parent verbal stimulation, thereby
reducing contingent responsiveness to the child
(Wachs, 1986).
Background television may also reduce parent
responsiveness by capturing parents’ attention. In
this case, what is considered background television
from the perspective of the child becomes fore-
ground television for the parent. Burns and Ander-
son (1993) reported that adults’ sustained attention
to television is associated with deeper engagement
and decreased distractibility away from the tele-
vision. To the extent to which parents watch a
program, they may become less able to shift
Impact of Background TV 1351

attention away from the television in response to
their children.
Overview of the Current Study
The purpose of this study was to investigate the
impact of background television on the quantity
and quality of parent–child interaction. Very young
children (12, 24, and 36 months of age) and their
parents were videotaped during a 1-hr free-play
session. For one half of the hour, a television pro-
gram selected by the parent played in the back-
ground. For the other half hour, the television
remained off. The sessions were videotaped and
subsequently coded for several interactive behav-
iors, including parent responsiveness and active
involvement in object play.
Method
Participants and Design
Participants were children within 1 month of
their first (n= 16, 8 females), second (n= 17, 8
females), or third (n= 18, 9 females) birthdays,
each accompanied by one parent. Two of the 51
parents in the study were fathers. Although there
were not enough fathers to conduct statistical com-
parisons, data from fathers were well within the
distributions of data from mothers on all of the
dependent measures. Parents identified 4 children
as Hispanic; the other 47 were Caucasian. Informa-
tion on socioeconomic status was collected in the
form of the highest level of education completed
by each parent. The highest degree obtained by
either parent was a high school diploma for 25%
of the sample (two thirds of these reported some
college), an undergraduate degree for 45%, and a
graduate degree for 30%. Parents reported that
children were exposed to an average of 4.24 hr
(SD = 2.28) of television per day, with an average
1.57 (SD = 1.52) of those hours being foreground
television (i.e., programs for children to which the
target child paid substantial attention). For 90% of
this sample, foreground television occupied less
than half of their total television exposure. Four
additional parent–child dyads participated but
were not included in the analyses due to equip-
ment failure.
The study was a 3 (age of child: 12, 24,
36 months) ·2 (sex of child: male, female) ·2 (con-
dition: TV, NoTV) mixed design with condition as
a repeated measure. Order of conditions was coun-
terbalanced such that half of the children at each
age were randomly assigned to play in the presence
of television first whereas the other half played in
the presence of television following 30 min of play
without television.
Setting and Apparatus
Two laboratory rooms were used for this study:
a playroom (4.27 m ·3.81 m) and an observation
room. The playroom was furnished with an arm-
chair, a table with current magazines and newspa-
pers, an open-shelved toy chest with a standard set
of age-appropriate toys (McCune-Nicholich &
Fenson, 1984), and a 48.26-cm television set and
video cassette player on a stand. A microphone
was hung from the ceiling in the playroom to
record utterances by the parent and child and a sta-
tionary digital video (DV) camera was hidden in a
box beneath the television stand. From the center of
the playroom floor, the television screen subtended
a visual angle of 17horizontally. At that distance,
the audio volume (preset by the experimenter) at
head level of a typical 24-month-old child averaged
57 db (C weighting over ten 200 ms samples),
about the level of typical human speech.
The observation room was an adjacent space sep-
arated by a one-way mirror. Another DV camera
mounted on a tripod in the observation room was
used as the primary camera whereas the stationary
camera in the playroom was used only when it pro-
vided a better view of the child’s face and hands.
Both cameras and the microphone were connected
to a DV recording deck. A video monitor in the
observation room allowed the observer to see both
camera views simultaneously and alternate
between them using a video switcher. The experi-
menter also controlled the playroom television and
video cassette player from the observation room
using a remote control.
Stimuli
Parents chose from a library of adult-directed
programs approximately 30 min in length. Pro-
grams were recorded from a variety of television
networks and included commercials presented at
the time of recording. The programs were from the
series Friends, Mad About You, Spin City, Frasier, The
Cosby Show, Home Improvement, Everybody Loves
Raymond (all situation comedies), Essence of Emeril
(a cooking show), This Old House Classics (a home
repair show), Jeopardy! (a game show), and A Make-
over Story (a daytime reality show). None of the
1352 Kirkorian, Pempek, Murphy, Schmidt, and Anderson

programs contained violent or overtly sexual
content.
Procedure
Upon arrival, each parent and child was offered
a snack and drink and then ushered into the play-
room where the child was free to play with the
toys. The experimenter started recording as soon as
the parent and child entered the playroom. The
parent was asked to provide written informed
consent and select a television program from a list
of the shows available. The parent was encouraged
to act as he or she normally would at home during
a free time period: free to play with the child, read
the magazines and newspapers provided, or watch
television. After any remaining questions were
answered, the experimenter left the room.
The experimental session began when the experi-
menter left the playroom. For the TV-first order, the
program chosen by the parent started immediately
after the session began. When the program ended,
the experimenter remotely turned off the television
and continued to videotape the parent and child
for an additional 30 min without television. For the
TV-second order, the experimenter set a stopwatch
to time the 30 min without television. After this
time, the experimenter remotely turned on the tele-
vision and started the program. At the end of 1 hr,
the experimenter entered the room for debriefing.
The parent was asked to complete a short question-
naire about demographics and home television-
viewing practices.
Videotape Coding
Verbal behaviors for both the parent and child
were coded in one pass through each tape and the
level of parent involvement was coded in a second
pass. All behaviors were coded in 10-s intervals.
Verbal behaviors were coded for each interval if
they occurred at least once during those 10 s.
Categories of parents’ verbal behavior included
who initiated the interaction, the content of the
interaction (providing information, asking ques-
tions, restricting behavior), and attention-directing
strategy (following, interrupting). Because prelimin-
ary analyses indicated that the effect of background
television was equivalent for every subcategory of
parents’ utterances, analyses presented here are on
an overall measure of parents’ verbal stimulation
collapsing across all subcategories of behavior. That
is, the dependent measure for parents’ utterances is
the proportion of 10-s intervals during which par-
ents engaged in any social verbalization, regardless
of type. The complete coding scheme can be found
in the Appendix. Children’s utterances were also
coded within 10-s intervals and identified as either
social or self-directed. Finally, parents and children
were coded for responsiveness. That is, each 10-s
interval was coded for whether it contained an
explicit request from either social partner (e.g.,
question) and, if so, whether it received a response
within 3 s.
In the second coding pass, each 10-s interval
received one of five codes for the level of parent
involvement characteristic of that interval. The lev-
els of involvement were as follows: not interacting
in any way, monitoring without interacting, inter-
acting without object play (e.g., grooming), and
engaging in object play. The object play category
was further qualified by parents’ passive or active
involvement. Active object play occurred when par-
ents were physically or verbally involved in reci-
procal interaction for the majority of the 10-s
interval. Passive object play occurred whenever
parents’ involvement in their children’s toy play
was not characterized as ‘‘active.’’ This included
intervals during which parents only passively com-
mented on their children’s activity or when par-
ents’ physical involvement was passive (e.g.,
holding out a hand to take a toy offered by the
child without speaking to or looking at the child).
Although it was not possible to conceal the
experimental manipulation, coders were blind to
hypotheses about the study. Each coder worked
with two training tapes until he or she achieved
acceptable reliability with an experienced coder
(phi correlations above .80). The coder was then
asked to rate a test tape. The coder was allowed to
rate tapes from the study if the phi correlation
between his or her coding and that of a supervisor
was .80 or above.
Twelve of the videotapes (approximately 25%)
were coded by two independent observers to assess
interrater reliability. These tapes were distributed
over the course of the study to ensure reliability
over time. All dependent measures achieved
acceptable reliability with intraclass correlations
above .70.
Results
Most of the analyses were mixed analyses of var-
iance (ANOVAs) with age of the child, sex of
the child, and order of condition as between-sub-
ject factors and television condition as a repeated
Impact of Background TV 1353

measure. Post hoc ttests were conducted with
Bonferroni correction. In cases where there were
significant interactions with order of condition
suggesting change in the dependent variable over
time, hierarchical linear modeling (HLM) was
employed to control for change over time and
isolate the effects of the other independent vari-
ables. Level 1 (within-subject) predictors included
linear and quadratic change over time (measured
in twelve 5-min intervals centered at the first
interval) and television condition (dichotomous
variable with 0 indicating NoTV and 1 indicating
TV). The Level 2 (between-subject) predictor was
age (measured in years centered on the youngest
group). Schmidt et al. (2008) employed a similar
procedure and provide a detailed example for
the design used here. See Raudenbush and Bryk
(2002) for a detailed description of the HLM
technique.
Dependent variables included parents’ overall
verbal stimulation (i.e., the proportion of 10-s inter-
vals during which parents engaged in any verbal
interaction), the proportion of intervals parents
spent at each of the five levels of involvement, the
proportion of intervals during which children
engaged in social and self-directed speech, and the
proportion of parent and child requests that
received a response. Descriptive statistics for these
dependent variables are presented in Table 1. Sub-
sequent mixed ANOVAs on the subcategories of
parent verbal behavior (initiator, content, and atten-
tion-directing strategy) produced results that paral-
lel those of overall verbal interaction and are not
presented here.
Verbal Stimulation
The first analysis was a 3 (age: 12, 24,
36 months) ·2 (sex) ·2 (test order: TV first, TV
second) ·2 (condition: NoTV, TV) mixed ANOVA
with condition as a repeated measure on the overall
amount of parent verbal interaction (collapsing
across all subcategories). Figure 1 is a scatter plot of
the proportion of intervals containing at least one
parent utterance in the presence of television in
relation to interactions without television. Points
that fall below the diagonal in Figure 1 represent
parents who exhibited less verbal behavior in
the presence of television. A chi-square test dem-
onstrated that most parents interacted less in the
presence of background television, c
2
(1) = 18.84,
p< .001. This plot demonstrates that the television
effect on overall parent verbal stimulation held for
a vast majority of individuals regardless of baseline
level of interaction without television.
The ANOVA indicated main effects of age and
condition, F(2, 39) = 5.15, p= .010, Cohen’s f= 0.40,
and F(1, 39) = 28.82, p< .001, Cohen’s f= 0.74,
respectively. Post hoc ttests indicated that verbal
interaction was greater for parents of 24- and 36-
month-olds than for parents of 12-month-olds.
Moreover, parents interacted with children 68% of
Table 1
Mean Values (Standard Errors) for Dependent Variables as a Function of Age and TV Condition
12 months 24 months 36 months
NoTV TV NoTV TV NoTV TV
Parent verbal 0.58 (.05) 0.42 (.05) 0.75 (.05) 0.59 (.05) 0.70 (.05) 0.61 (.04)
Level of involvement
Not interacting 0.15 (.03) 0.25 (.03) 0.14 (.03) 0.28 (.03) 0.19 (.03) 0.13 (.03)
Monitoring 0.13 (.03) 0.15 (.02) 0.13 (.03) 0.11 (.03) 0.13 (.03) 0.12 (.02)
Nonobject 0.21 (.03) 0.19 (.02) 0.15 (.03) 0.12 (.02) 0.13 (.03) 0.08 (.02)
Passive object 0.10 (.02) 0.11 (.02) 0.12 (.02) 0.12 (.02) 0.09 (.02) 0.12 (.02)
Active object 0.29 (.05) 0.16 (.04) 0.33 (.05) 0.17 (.04) 0.40 (.05) 0.36 (.04)
Child behavior type
Child social 0.18 (.03) 0.14 (.04) 0.60 (.03) 0.46 (.03) 0.66 (.03) 0.56 (.03)
Self-directed 0.16 (.03) 0.14 (.03) 0.15 (.03) 0.17 (.03) 0.16 (.03) 0.22 (.03)
Responsiveness
Child requests 0.87 (.04) 0.79 (.05) 0.84 (.04) 0.72 (.05) 0.91 (.03) 0.87 (.04)
Parent requests 0.16 (.03) 0.13 (.03) 0.56 (.03) 0.56 (.03) 0.72 (.03) 0.67 (.03)
Note. Dependent variables are the proportion of 10-s intervals during which parents engaged in any verbal interaction, the proportion
of intervals characterized as each of the five levels of parent involvement, the proportion of intervals during which children engaged in
social and self-directed speech, and the proportion of parent and child requests responded to by the social partner within 3 s.
1354 Kirkorian, Pempek, Murphy, Schmidt, and Anderson

the time without television but only 54% of the time
with television. There were no other main effects or
interactions. Thus, parents generally talked less to
12-month-olds than to older children, but parents
of children at all ages interacted less with television
than without.
Parent Involvement
The second analysis was a 3 (age) ·2 (sex) ·2
(order) ·2 (condition) ·5 (level of involvement)
mixed ANOVA with condition and level of
involvement (none, monitoring, nonobject, passive
and active) as repeated measures. There were sig-
nificant interactions between level and age, F(4,
156) = 3.95, p< .001; level and condition, F(4,
156) = 14.61, p< .001; and level, condition, and
order of conditions, F(4, 156) = 8.68, p< .001.
Because of these interactions, lower order 3
(age) ·2 (order) ·2 (condition) ANOVAs were
performed separately for each of the five levels of
parent involvement. There were no significant main
effects or interactions for monitoring or passive
object play. For nonobject interaction, there was a
main effect of age such that this behavior was more
frequent among parents of 12-month-olds than the
older age groups, F(2, 45) = 6.45, p= .003, Cohen’s
f= 0.46. There was also more nonobject interaction
in the second half-hour than the first, regardless of
condition, as evidenced by a significant Condi-
tion ·Order interaction, F(1, 45) = 14.05, p= .001,
Cohen’s f= 0.51.
For the time parents spent not interacting in any
way, there were main effects of age and condition,
F(2, 45) = 4.90, p= .012, Cohen’s f= 0.39, and
F(1, 45) = 16.21, p< .001, Cohen’s f= 0.55, respec-
tively. Parents spent less time interacting with
12- and 24-month-olds than with 36-month-olds,
regardless of condition. Moreover, when the televi-
sion was on, parents more than doubled the time
they spent not interacting with their children. This
effect was due to reduced active involvement,
which decreased by one third in the presence of
television, F(1, 45) = 28.36, p< .001, Cohen’s
f= 0.73 (see Figure 2). The main effect of condition
was qualified by an interaction with order, F(1,
45) = 15.85, p< .001, Cohen’s f= 0.54. HLM was
employed to control for change over time and iso-
late the effect of background television on parents’
active involvement. The final model included sig-
nificant linear decline in the proportion of each 5-
min interval during which parents were actively
involved in their children’s object play, B=).01
(SE < .01), t(50) = )3.74, p= .001. Moreover, when
controlling for this change over time, background
television remained a significant negative predictor
of parents’ active involvement, B=).11 (SE = .02),
t(50) = )5.41, p< .001. In addition to the effect of
background television, parents’ active involvement
increased with age of child, although only the
Figure 2. Mean proportion of 10-s intervals during which parents
were rated as not interacting or actively involved with their
children as a function of television condition. Bars represent
±1SE.
Figure 1. Scatterplot of parents’ verbal interaction (proportion of
10-s intervals) in the presence of background television in
relation to parents’ verbal interaction without background
television. Points below the diagonal represent parents who
interacted less in the presence of background television.
Impact of Background TV 1355

difference between the youngest and oldest chil-
dren was significant, F(2, 45) = 4.16, p= .022, Co-
hen’s f= 0.35.
To summarize results for parent involvement,
parents generally spent more time engaged in
object play with older children and more time in
nonobject interaction (e.g., grooming) with younger
children. Moreover, parents’ active involvement
decreased over the hour. With respect to the pres-
ence of background television, parents spent less
time actively involved in their children’s object play
and more time not interacting in any way when the
television was on. This was true for parents of chil-
dren in all three age groups.
Child Behavior
The third analysis was a 3 (age) ·2 (sex) ·2
(order) ·2 (condition) ·2 (type) mixed ANOVA
with condition and type of child behavior (social,
self-directed) as repeated measures. There were
significant main effects of age, condition, and type,
F(2, 39) = 52.21, p< .001, Cohen’s f= 1.42; F(1,
39) = 11.07, p= .002, Cohen’s f= 0.44; and F(1,
39) = 142.72, p< .001, Cohen’s f= 1.61, respectively.
Specifically, 1-year-olds exhibited less social behav-
ior and self-directed speech than did either of the
two older groups, social behavior occurred more
frequently than did self-directed speech, and
children engaged in fewer social behaviors and self-
directed speech in the presence of background
television. In addition to these main effects, type of
behavior significantly interacted with age and con-
dition, F(2, 39) = 32.54, p< .001, Cohen’s f= 1.09,
and F(1, 39) = 14.33, p= .001, Cohen’s f= 0.51,
respectively; thus, lower order 3 (age) ·2 (condi-
tion) ANOVAs were conducted separately for child
social behavior and self-directed speech. There were
no significant main effects or interactions for self-
directed speech. For social interactions, there was a
main effect of condition such that children’s social
interactions decreased by 19% in the presence of
television, F(1, 48) = 23.58, p< .001, Cohen’s
f= 0.67. There was also a main effect of age such
that 12-month-olds engaged in fewer social behav-
iors than did the two older groups, F(2, 48) = 88.13,
p< .001, Cohen’s f= 1.31.
Parent and Child Responsiveness
Responsiveness refers to the frequency of verbal
or nonverbal responses to bids for attention from
the social partner. Separate 3 (age) ·2 (sex) ·2
(order) ·2 (condition) mixed ANOVAs were con-
ducted on responsiveness of parents and children.
For parents, there was a significant main effect of
condition such that they were less responsive in the
presence of television, responding to child bids
about 87% of the time without television but only
79% of the time with television, F(1, 39) = 8.11,
p= .007.
For children, there were main effects of age and
condition, F(2, 39) = 120.65, p< .001, and F(1, 39) =
5.31, p= .027, respectively. Child responsiveness
increased with age and was greater overall without
background television (48% vs. 45%). However,
these main effects were qualified by a three-way
interaction among age, order, and condition; thus,
HLM was employed to control for change over time
in children’s responsiveness and to isolate the
effects of condition and age. Age significantly pre-
dicted the intercept demonstrating that child
responsiveness increased with age, B= .25 (SE =
.02), t(49) = 11.54, p< .001. There was also signifi-
cant linear change over time, B=).01 (SE < .01,
t(49) = )3.07, p= .004], which varied as a function
of age, B= .01 (SE < .01), t(49) = 2.95, p= .005. This
interaction occurred because responsiveness
decreased somewhat over time for the 1-year-olds,
remained stable for the 2-year-olds, and increased
somewhat for the 3-year-olds. After controlling for
change over time, the effect of background televi-
sion was not significant.
To summarize results for parent and child
responsiveness, parents were less responsive to
their children in the presence of television.
Although child responsiveness increased with age,
there was no effect of background television on
child responsiveness after controlling for change
over time. Thus, perhaps not surprisingly, the
impact of background television (i.e., programs tar-
geting an adult audience) on contingent responsive-
ness was limited to parents.
Discussion
Most research on television’s impact on children
focuses on the effects of program content on learn-
ing and behavior. For example, violent content can
influence children to engage in aggressive behavior
and educational content can increase academically
relevant knowledge (e.g., Anderson, Huston,
Schmitt, Linebarger, & Wright, 2001). These effects
depend in part on children understanding what
they are watching. It is unlikely, however, that
infants and toddlers understand much of the con-
tent of the programs watched by their parents. The
1356 Kirkorian, Pempek, Murphy, Schmidt, and Anderson

present research instead hypothesizes that these
programs produce distracting background stimula-
tion for very young children and become engaging
stimuli for the parents that interfere with social
interactions.
The purpose of this study was to determine
experimentally whether background television has
an impact on the quantity and quality of parent–
child interactions. The results were clear. Not only
did background television reduce overall interac-
tions, this effect was due to a reduction in parents’
active engagement.
Background Television and Parent–Child Interaction
Wachs (1986) suggested that background noise
interferes with parent–child interaction by interrupt-
ing verbal processing and therefore parents’ ability
to verbally stimulate their children. This was clearly
supported in this study. Although parents were not
given specific instructions to watch the television
program, they were much less verbally interactive
with their children when the television was on.
Moreover, as would be predicted from studies of
engagement with television (e.g., Burns & Ander-
son, 1993), parents were less responsive to their chil-
dren’s bids for attention when the television was on.
Adult-directed television programs also decreased
the social behavior of children. In contrast, there
was no decrease in self-directed utterances that are
characteristically frequent among the 24- and 36-
months-olds, suggesting that this was not due to a
general decrease in verbal behavior. Although some
of the effect may be due to auditory interference
from the television, it is equally likely that the
reduction in children’s social behavior results from
the reduced responsiveness of the parent. Further
research is needed to determine the specific mecha-
nisms by which background television influences
children’s social behavior.
Another key finding is that background televi-
sion affects not only the quantity of parent–child
interaction but also the quality of those interactions.
Parents were less likely to be attentive and actively
involved during interactions with their children in
the presence of television than during interactions
without television. When interactions occurred,
they were much more likely to be of a passive
nature on the part of the parent (e.g., verbally
acknowledging the child without actually looking
at or otherwise interacting with the child). This
finding lends support to the hypothesis that back-
ground television has a differential effect on behav-
iors requiring relatively more attention on the part
of the parent. Recall that these more attentive
parental behaviors are also the behaviors that posi-
tively influence concurrent toy play and predict
positive developmental outcomes (e.g., Alessandri,
1992; Parks & Bradley, 1991; Slade, 1987).
Future Directions
Although the focus of the current study is the
impact of adult-directed television presented in the
background (from the perspective of the child), it is
important to consider how these results might dif-
fer with foreground television, that is, programs
actually designed for infants and toddlers. In a pro-
ject by our research group that is nearing comple-
tion, preliminary analyses suggest that the overall
quantity of parent–child interaction also decreases
in the presence of a children’s television program.
However, parents seem to fill that time by coview-
ing and using the program as a basis for interacting
with their children (e.g., labeling objects and
actions on the screen, singing, and even dancing
with the child; Pempek, Demers, Anderson, &
Kirkorian, 2007). Moreover, when coviewing infant-
directed videos in the home, parents’ scaffolding
behavior (questions, labels, descriptives) predicts
infants’ engagement with and attention to the vid-
eos (Barr, Zack, Garcia, & Muentener, 2008). It may
be that the impact of foreground television on par-
ent–child interaction is more positive and that the
long-term effects are less detrimental, at least when
parent and child watch together.
It is particularly important to determine whether
observations in homes would yield the same find-
ings as the laboratory studies. This need is exempli-
fied by the present finding that parents’ active
involvement with their children decreased over
time regardless of whether the television was on,
despite a significant effect of television when con-
trolling for this change. It is possible that parents
were initially and extraordinarily attentive to their
children in the novel laboratory setting while being
observed. In their own homes, however, parents
are likely to become involved in other activities and
to act more naturally. Observations in the home
over extended periods of time will provide ecologi-
cal validity to the current findings.
Summary and Conclusions
The evidence is growing that very early
exposure to television is associated with negative
developmental outcomes. Anderson and Evans
(2001) hypothesized that these effects may be due
Impact of Background TV 1357

at least in part to disruptive influences by back-
ground television on toddlers’ solitary play behav-
iors as well as to disruptions of parent–child
interactions. The experiment by Schmidt et al.
(2008) found that background television does in fact
disrupt solitary toy play. This study confirmed the
hypothesis that background television interferes
with parent–child interactions, disrupting social
input and parent engagement as the parent’s atten-
tion is occupied by the television. Both findings
may in part explain the negative associations found
between early exposure to television and subse-
quent development.
References
Alessandri, S. M. (1992). Mother-child interactional corre-
lates of maltreated and nonmaltreated children’s play
behavior. Development and Psychopathology,4, 257–270.
Anderson, D. R., & Evans, M. K. (2001). Peril and potential
of media for infants and toddlers. Zero to three: National
Center for Infants, Toddlers, and Families,22, 10–16.
Anderson, D. R., Huston, A. C., Schmitt, K. L.,
Linebarger, D. L., & Wright, J. C. (2001). Early
childhood television viewing and adolescent behavior:
The recontact study. Monographs of the Society
for Research in Child Development 68(Serial No. 264),
1–143.
Armstrong, G. B., & Greenberg, B. S. (1990). Background
television as an inhibitor of cognitive processing.
Human Communication Research,16, 355–386.
Baker, M. A., & Holding, D. H. (1993). The effects of noise
and speech on cognitive task performance. The Journal
of General Psychology,120, 339–355.
Barr, R., Zack, E., Garcia, A., & Muentener, P. (2008).
Infants’ attention and responsiveness to television
increases with prior exposure and parent interaction.
Infancy,13, 30–56.
Beckwith, L., & Rodning, C. (1996). Dyadic processes
between mothers and preterm infants: Development at
ages 2 to 5 years. Infant Mental Health Journal,17, 322–
333.
Bornstein, M. H., & Tamis-LeMonda, C. S. (1989). Mater-
nal responsiveness and cognitive development in chil-
dren. In M. H. Bornstein (Ed.), Maternal responsiveness:
Characteristics and consequences (pp. 49–61). San Fran-
cisco: Jossey-Bass.
Burns, J. J., & Anderson, D. R. (1993). Attentional inertia
and recognition memory in adults’ television viewing.
Communication Research,20, 777–799.
Carew, J. V. (1980). Experience and the development of
intelligence in young children at home and in day care.
Monographs for the Society for Research in Child Develop-
ment,45 (Serial No. 187).
Christakis, D. A., Zimmerman, F. J., DiGiuseppe, D. L., &
McCarty, C. A. (2004). Early television exposure and
subsequent attentional problems in children. Pediatrics,
113, 708–713.
Clarke-Stewart, K. A. (1973). Interactions between
mothers and their young children: Characteristics and
Consequences. Monographs of the Society for Research in
Child Development,38 (Serial No. 153).
Corapci, F., & Wachs, T. D. (2000). Does parental mood
or efficacy mediate the influence of environmental
chaos upon parenting behavior? Merrill-Palmer
Quarterly,48, 182–201.
Landry, S. H., Miller-Loncar, C. L., & Swank, P. R. (1998).
Goal-directed behavior in children with Down’s syn-
drome: The role of joint play situations. Early Education
& Development,9, 375–392.
Landry, S. H., Smith, K. E., Miller-Loncar, C. L., & Swank,
P. R. (1998). The relation of change in maternal interac-
tive styles to the developing social competence of
full-term and preterm children. Child Development,69,
105–123.
Lawson, K. R., Parrinello, R., & Ruff, H. A. (1992). Mater-
nal behavior and infant attention. Infant Behavior &
Development,15, 209–229.
Lewis, M. D. (1993). Early socioemotional predictors of
cognitive competency at 4 years. Developmental Psychol-
ogy,29, 1036–1045.
Masur, E. F., & Flynn, V. (2008). Infant and mother-infant
play and the presence of the television. Journal of
Applied Developmental Psychology,29, 76–83.
McCune-Nicholich, L., & Fenson, L. (1984). Methodologi-
cal advances in the study of early pretend play. In
T. D. Hawkey & A. D. Pellegrini (Eds.), Child’s play
(pp. 81–104). Hillsdale, NJ: Erlbaum.
National Institute of Child Health and Human Develop-
ment Early Child Care Research Network. (2003). Do
children’s attention processes mediate the link between
family predictors and school readiness? Developmental
Psychology,39, 581–593.
Nelson, K. (1973). Structure and strategy in learning to
talk. Monographs of the Society for Research in Child Devel-
opment,38 (Serial No. 149), 1–137.
Olson, S. L., Bates, J. E., & Bayles, K. (1984). Mother-
infant interaction and the development of individual
differences in children’s cognitive competence. Develop-
mental Psychology,20, 166–179.
Parks, P. L., & Bradley, R. H. (1991). The interaction of
home environment features and their relation to infant
competence. Infant Mental Health Journal,12, 3–14.
Pempek, T. A., Demers, L. B., Anderson, D. R., &
Kirkorian, H. L. (2007, March). The impact of baby videos.
Paper presented at the biannual meeting of the Society
for Research in Child Development, Boston.
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical lin-
ear models: Applications and data analysis methods (2nd
ed.). Thousand Oaks, CA: Sage.
Rideout, V., & Hamel, E. (2006). The media family: Elec-
tronic media in the lives of infants, toddlers, preschoolers
and their parents. Menlo Park, CA: Henry J. Kaiser
Family Foundation.
1358 Kirkorian, Pempek, Murphy, Schmidt, and Anderson

Rideout, V., Vandewater, E., & Wartella, E. (2003). Zero to
six: Electronic media in the lives of infants, toddlers, and
preschoolers. Menlo Park, CA: Henry J. Kaiser Family
Foundation.
Schmidt, M. K., Pempek, T. A., Kirkorian, H. L., Lund, A.
F., & Anderson, D. R. (2008). The impact of background
television on very young children. Child Development,
79, 1137–1151.
Schmitt, K. L., Anderson, D. R., & Collins, P. A. (1999).
Form and content: Looking at visual features of televi-
sion. Developmental Psychology,35, 1156–1167.
Schmitt, K. L., Woolf, K. D., & Anderson, D. R. (2003).
Viewing the viewers: Viewing behaviors by children
and adults during television programs and commer-
cials. Journal of Communication,53, 265–281.
Slade, A. (1987). A longitudinal study of maternal
involvement and symbolic play during the toddler per-
iod. Child Development,58, 367–375.
Tamis-Lemonda, C. S., Bornstein, M. H., Baumwell, L.,
& Damast, A. M. (1996). Responsive parenting in the
second year: Specific influences on children’s language
and play. Early Development & Parenting,5, 173–183.
Vygotsky, L. S. (1978). Mind in society: The development of
higher psychological processes. Cambridge, MA: Harvard
University Press.
Wachs, T. D. (1986). Ambient background noise and early
development. Children’s Environments Quarterly,3, 23–33.
Zimmerman, F. J., & Christakis, D. A. (2005). Children’s
television viewing and cognitive outcomes: A longitu-
dinal analysis of national data. Archives of Pediatrics and
Adolescent Medicine,159, 619–625.
Appendix
Parent verbal behavior
Initiation
Response—response to verbal or nonverbal social behavior from child
Unsolicited—help or instruction to child that is contingent on child’s behavior but not solicited by child
Spontaneous—comments not in response to child or his or her behavior
Content
Informative—provides information (e.g., descriptions, labels)
Restrictive—uses ‘‘do’’ or ‘‘do not’’ commands and other directives or explicit demonstrations
Suggestive—asks questions or shows or offers objects
Praise—praises or positively acknowledges child
Other—uses some other type of utterance
Attention directing
Following—object of parent’s utterance is same as child’s focus of attention
Interrupting—object of parent’s utterance is different from child’s focus of attention
Child behavior
Social—utterance or gesture that is social in nature
Self-directed—utterance that is not social in nature
Parent and child responsiveness
Response—appropriately responds within 3 s to explicit bid (e.g., question, point) from social partner
Nonresponse—fails to respond within 3 s to explicit bid
Parent involvement
Not interacting—does not look at or interact with child
Monitoring—looks at child but does not interact
Nonobject—interacts with child without engaging in object play (e.g., grooms)
Passive—interacts during object play but is not actively involved (e.g., acknowledges without looking)
Active—interacts physically or verbally in reciprocal interaction surrounding object play
The above table lists all categories of parent and child behavior originally coded. Categories of parent verbal behavior, child behavior,
and responsiveness were coded if they occurred at least once in each interval. For parent involvement, each 10-s interval was assigned
one code for the level of involvement characteristic of that interval. Analyses of individual categories of parent verbal behavior are not
reported here because they did not differ across categories and were identical to results for parents, overall verbal stimulation (i.e., col-
lapsing across all categories of verbal behavior).
Impact of Background TV 1359