Content uploaded by James D Sargent
Author content
All content in this area was uploaded by James D Sargent on May 25, 2015
Content may be subject to copyright.
http://pss.sagepub.com/
Psychological Science
http://pss.sagepub.com/content/early/2012/07/18/0956797611435529
The online version of this article can be found at:
DOI: 10.1177/0956797611435529
published online 18 July 2012Psychological Science
Ross E. O'Hara, Frederick X. Gibbons, Meg Gerrard, Zhigang Li and James D. Sargent
Sexual Risk Taking
Greater Exposure to Sexual Content in Popular Movies Predicts Earlier Sexual Debut and Increased
Published by:
http://www.sagepublications.com
On behalf of:
Association for Psychological Science
can be found at:Psychological ScienceAdditional services and information for
http://pss.sagepub.com/cgi/alertsEmail Alerts:
http://pss.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
What is This?
- Jul 18, 2012OnlineFirst Version of Record >>
at DARTMOUTH COLLEGE on July 24, 2012pss.sagepub.comDownloaded from
Psychological Science
XX(X) 1 –10
© The Author(s) 2012
Reprints and permission:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0956797611435529
http://pss.sagepub.com
The effects of media on adolescents’ risk behaviors, including
tobacco use (National Cancer Institute, 2008), alcohol use
(P. Anderson, de Bruijn, Angus, Gordon, & Hastings, 2009),
and aggression (C. A. Anderson et al., 2003), have been widely
documented. Relatively less is known, however, about how
media influence adolescents’ sexual behavior, including their
age of sexual debut and subsequent sexual risk taking. Early
sexual debut is associated with an increased number of sexual
partners and inconsistent condom use, as well as an increased
risk of sexually transmitted infections (STIs; Kaestle, Halpern,
Miller, & Ford, 2005). Delaying adolescents’ sexual debut,
therefore, could curb U.S. rates of STIs (more than 9 million
new cases occur annually among adolescents; Weinstock,
Berman, & Cates, 2000), and could potentially reduce instances
of unplanned pregnancy (roughly 64 unplanned pregnancies
occur for every 1,000 female adolescents age 19 or younger;
Guttmacher Institute, 2010). Identifying risk factors for early
sexual debut and sexual risk taking, therefore, is an important
public-health concern. One significant influence on engage-
ment in risky sexual behavior may be media (Wright, 2011)—
specifically, movie sexual exposure (MSE). In the study
reported here, we examined the association of MSE with sex-
ual debut and engagement in risky sexual behaviors, both
directly and indirectly through changes in sensation seeking.
Sex in the Movies
Popular movies provide adolescents with a wealth of sexual
exposure, much of which may promote risk behaviors. A survey
of movies released from 1950 to 2006 revealed that more than
84% contained sexual content (68% of G-rated movies, 82% of
PG-rated movies, 85% of PG-13-rated movies, and 88% of
R-rated movies; Nalkur, Jamieson, & Romer, 2010). Also, the
sexual explicitness of PG-13-rated and R-rated movies has
increased over the past decade (Nalkur et al., 2010). Potentially
even more important for adolescents’ sexual health, however, is
that most of these movies do not portray safe sex. A content
analysis revealed that 70% of the sexual acts depicted in movies
released from 1983 to 2003 occurred between newly acquainted
partners, 98% included no reference to contraception, and
89% resulted in no consequences (Gunasekera, Chapman, &
Campbell, 2005). Additionally, Pardun, L’Engle, and Brown
(2005) found that only 9% of sexual content in movies con-
tained messages promoting sexual health. Adolescents who
Corresponding Author:
Ross E. O’Hara, University of Missouri, 147 Psychology Building, 200 South
7th St., Columbia, MO 65211
E-mail: oharar@missouri.edu
Greater Exposure to Sexual Content in
Popular Movies Predicts Earlier Sexual
Debut and Increased Sexual Risk Taking
Ross E. O’Hara1, Frederick X. Gibbons1, Meg Gerrard2,
Zhigang Li3, and James D. Sargent4
1Department of Psychological and Brain Sciences, Dartmouth College ; 2Department of Psychiatry, Geisel School
of Medicine, Dar tmouth College; 3Department of Community and Family Medicine, Geisel School of Medicine,
Dartmouth College ; and 4Department of Pediatrics, Geisel School of Medicine, Dartmouth College
Abstract
Early sexual debut is associated with risky sexual behavior and an increased risk of unplanned pregnancy and sexually
transmitted infections later in life. The relations among early movie sexual exposure (MSE), sexual debut, and risky sexual
behavior in adulthood (i.e., multiple sexual partners and inconsistent condom use) were examined in a longitudinal study
of U.S. adolescents. MSE was measured using the Beach method, a comprehensive procedure for media content coding.
Controlling for characteristics of adolescents and their families, analyses showed that MSE predicted age of sexual debut,
both directly and indirectly through changes in sensation seeking. MSE also predicted engagement in risky sexual behaviors
both directly and indirectly via early sexual debut. These results suggest that MSE may promote sexual risk taking both by
modifying sexual behavior and by accelerating the normal rise in sensation seeking during adolescence.
Keywords
mass media, sex
Received 8/24/11; Revision accepted 12/16/11
Research Article
Psychological Science OnlineFirst, published on July 18, 2012 as doi:10.1177/0956797611435529
at DARTMOUTH COLLEGE on July 24, 2012pss.sagepub.comDownloaded from
2 O’Hara et al.
watch popular movies, therefore, are exposed to a great deal of
sex, most of which is portrayed in an unrealistic and/or risk-
promoting manner.
How Media Influence Sexual Behavior
Wright (2011) posited that the effect of media on sexual
behavior is driven by the acquisition and activation of sexual
scripts. Scripts provide behavioral options in social situations,
including those that may lead to sexual behavior, and the con-
tent of scripts is often influenced by media. As mentioned in
the previous section, movies generally offer permissive and
risky sexual messages to viewers (Gunasekera et al., 2005;
Nalkur et al., 2010), and a higher degree of sexual media expo-
sure has been found to predict more permissive sexual atti-
tudes (Bleakley, Hennessy, Fishbein, Coles, & Jordan, 2009;
Brown, Halpern, & L’Engle, 2005). Furthermore, adolescents
sometimes seek out sexual media, possibly to learn these
scripts (Brown et al., 2005). In fact, 57% of U.S. adolescents
(ages 14–16) reported using media as a primary source of sex-
ual information (Bleakley et al., 2009).
Activated scripts must be applied to guide behavior, and
media can influence which sexual scripts are used (Wright,
2011). Among adolescents with high MSE, the sexual scripts
portrayed in movies may be readily accessible because of the
frequency of prior activations. The more easily a script is acti-
vated, the more likely it is to be applied in a given situation. In
fact, longitudinal studies have shown that a greater degree of
television sexual exposure predicts greater engagement in
noncoital sexual activities among adolescents (Collins et al.,
2004) and earlier sexual debut (Ashby, Arcari, & Edmonson,
2006; Collins et al., 2004; Martino, Collins, Kanouse, Elliott,
& Berry, 2005), controlling for demographic factors, religios-
ity, and parenting. Furthermore, greater exposure to sexual
content in media, including movies, television, music, and
magazines, has been associated with both a higher likelihood
of engaging in noncoital sexual activities among adolescents
(Brown et al., 2006; L’Engle, Brown, & Kenneavy, 2006;
Pardun et al., 2005) and earlier sexual debut (Brown et al.,
2006), controlling for adolescents’ and their parents’ charac-
teristics. Finally, greater exposure to movies and men’s maga-
zines has been associated with earlier sexual debut and a
greater number of casual partners among male college stu-
dents, effects mediated by sexual norms and beliefs (Ward,
Epstein, Caruthers, & Merriwether, 2011). These results lend
support to Wright’s model by showing that sexual media influ-
ence both sexual attitudes and behavior.
Effects of Movies on Adolescents’
Risk Behavior
Despite the plethora of sex depicted in movies and the popu-
larity of movies among youth, much more research has been
devoted to the influence of television on adolescents’ sexual
behavior. We were interested in the effect of MSE on sexual
debut and engagement in risky sexual behaviors because evi-
dence suggests that adolescents’ sexual attitudes and behavior
are influenced more by movies than by other forms of media
(Bleakley et al., 2009; Pardun et al., 2005). For example, in a
study of male college students (Ward et al., 2011), only expo-
sure to movies (as compared with exposure to television,
music videos, and men’s magazines) directly predicted age of
sexual debut, and exposure to movies (along with exposure to
men’s magazines) indirectly predicted number of casual sex-
ual partners. Moreover, longitudinal studies of movies’ influ-
ence on adolescents’ substance use have shown strong and
consistent effects: Exposure to tobacco use in movies predicts
initiation and escalation of smoking (Dalton et al., 2009;
Sargent et al., 2007; Wills et al., 2010), and exposure to drink-
ing in movies predicts initiation and escalation of alcohol use
(Dal Cin et al., 2009; Gibbons et al., 2010; Sargent, Wills,
Stoolmiller, Gibson, & Gibbons, 2006; Wills et al., 2010).
Our goal in the study reported here was to examine the
effects of early MSE (i.e., before age 16) on age of sexual
debut and risky sexual behaviors (i.e., multiple sexual partners
and inconsistent condom use) in adulthood. These relations
were assessed using data from a longitudinal study of
U.S. adolescents (Sargent et al., 2007). We employed the
Beach method to estimate participants’ exposure to risky sex-
ual behaviors portrayed in movies (Sargent, Worth, Beach,
Gerrard, & Heatherton, 2008). This method entails second-by-
second coding of risk behaviors in movies to maximize valid-
ity and reliability, and it allowed for a more comprehensive
sampling of popular movies than has been used in previous
studies. It has been validated in studies among youth of mov-
ies’ effects on smoking (e.g., Sargent et al., 2007) and on alco-
hol use and alcohol-related problems (e.g., Dal Cin et al.,
2009). The current study was the first to use the Beach method
to estimate MSE and examine its relations with sexual debut
and risky sexual behaviors.
Effects of Movies on Sensation Seeking
There is reason to believe that MSE influences sexual behav-
ior indirectly by increasing sensation seeking—the tendency
to seek novel and intense stimulation (Arnett, 1994). Sensa-
tion seeking rises during adolescence, peaking between the
ages of 10 and 15, and then declines through late adolescence
(Steinberg et al., 2008). Greater sensation seeking is associ-
ated with both earlier sexual debut (Donohew et al., 2000) and
more frequent engagement in casual sex in adulthood (Arnett,
1994). It is important to note that sensation seeking arises from
both biological and socialization factors (Arnett, 1994), which
suggests that environmental influences, such as MSE, could
affect the development of this trait. In fact, research with the
sample used in the current study showed that watching R-rated
movies was associated with later increases in sensation seek-
ing during adolescence (but not vice versa), which in turn
increased the risk of tobacco and alcohol use during adoles-
cence (de Leeuw et al., 2011; Stoolmiller, Gerrard, Sargent,
at DARTMOUTH COLLEGE on July 24, 2012pss.sagepub.comDownloaded from
Movie Sexual Exposure 3
Worth, & Gibbons, 2010). However, the mediating effect of
sensation seeking has not been tested, to our knowledge, with
respect to media’s influence on sexual behavior. In the current
study, therefore, we examined whether changes in sensation
seeking mediated the anticipated relations of MSE with sexual
debut and with risky sexual behaviors.
The Current Study
This study expanded upon past research in several ways. First,
previous studies combined MSE with exposure to sexual con-
tent in other media (e.g., Brown et al., 2006), thereby obscur-
ing the effect of MSE. In our analysis, we focused exclusively
on MSE. Second, the Beach method allowed for the content
coding of more than 600 popular movies released over a 6-year
period, a much larger sampling than has been used in previous
research. Third, few studies of movies’ effects on sexual
behavior have been longitudinal; the duration over which the
data analyzed in our study were collected allowed us to exam-
ine both sexual debut and postdebut sexual outcomes that
could result in STIs or unplanned pregnancies. Finally, this
study was the first to examine whether media effects on
engagement in risky sexual behaviors are mediated by changes
in sensation seeking. Specifically, our hypotheses were as
follows:
•Hypothesis 1: Early MSE predicts age of sexual
debut, an effect mediated by increases in sensation
seeking.
•Hypothesis 2: Early MSE predicts engagement in
risky sexual behaviors (i.e., increased number of
sexual partners and frequency of casual sex without a
condom) approximately 6 years later, an effect medi-
ated by age of sexual debut.
Method
Participants and procedure
These data were collected in a six-wave longitudinal study
spanning from June 2003 to October 2009. At Time 1, data
were collected in a random-digit-dial telephone survey of
6,522 adolescents, from 10 to 14 years of age, living in
the United States. The subsequent three follow-up surveys
were conducted approximately every 8 months; the final two
follow-ups occurred approximately 5 years and 7 years after
Time 1. At Time 6, 2,718 participants responded (38.2% reten-
tion), but only participants who were 18 years of age or older
(n = 1,300) were asked to report their sexual behavior. To
ensure that MSE had occurred before sexual debut, we omitted
from the analysis participants whose sexual debut occurred
before Time 2 (n = 72), which left a final sample of 1,228 par-
ticipants. Participants in the final sample were between 12 and
14 years old at Time 1 (M = 12.89 years, SD = 0.79) and
between 18 and 21 years old at Time 6 (M = 18.90 years, SD =
0.81). The sample comprised 611 males (49.8%) and 617
females (50.2%); 891 were European American (72.6%), 159
were Hispanic (12.9%), 71 were African American (5.8%),
and 107 were of other racial or ethnic backgrounds (8.7%).
Participants lost at follow-up were at greater risk for early
sexual debut and engagement in risky sexual behavior at Time
1 than were those who were retained in the sample. The par-
ticipants who were not retained reported higher MSE and sen-
sation seeking and lower maternal responsiveness, and were
more likely to have a television in their bedroom (ps < .001).
Also, significantly more minorities than European Americans
were lost at follow-up (p < .02).
Measures
MSE was measured using the Beach method. At Time 1, the
523 top-grossing movies released between 1998 and 2003
were coded for the number of seconds of sexual content, which
was defined as instances of sexual behavior, such as heavy
kissing or intercourse. Each movie was rated by one of two
trained coders, and a random subsample of 10% of the movies
was double-coded (interrater agreement: r = .92). Each partici-
pant received a unique list of 50 movies randomly selected
from the larger pool and reported which of those movies he or
she had seen. These data were used to extrapolate participants’
total exposure to sexual content from all 523 movies. The
same procedure was used at Time 2 with a smaller pool of
movies (161), comprising the top-grossing movies released
since the previous content coding. (The number of seconds of
sexual content in selected representative movies is presented
in Table S1 in the Supplemental Material available online.) We
calculated MSE by converting seconds of sexual content into
hours, summing the hours of sexual content viewed at Time 1
and Time 2, and performing a square-root transformation to
correct for positive skew.
Sensation seeking was measured with a four-item scale
designed for children (Time 1: α = .60; Time 2: α = .58;
Stephenson, Hoyle, Palmgreen, & Slater, 2003). This measure
tapped two of four constructs identified by Zuckerman (1994)
as important components of sensation seeking, thrill/adventure
seeking and boredom susceptibility; in addition, it tapped inten-
sity seeking, a component of the Arnett Inventory of Sensation
Seeking (Arnett, 1994). Participants responded to each item
using a scale from 1 to 4, with higher scores indicating higher
sensation seeking. Each participant’s scores were summed. This
measure has been validated for predicting adolescents’ tobacco
and alcohol use (de Leeuw et al., 2011; Stoolmiller et al., 2010).
Age of sexual debut was reported by participants at Time 6.
Risky sexual behavior was measured at Time 6 and comprised
two components: lifetime number of vaginal- or oral-sex part-
ners (open response) and number of instances of casual sex
(defined as vaginal sex not with a “serious or steady dating
partner”) without a condom (reported using a scale from 0,
never, to 5, five or more times). Scores for these two items
were recoded into ordinal variables and combined, α = .62.1
at DARTMOUTH COLLEGE on July 24, 2012pss.sagepub.comDownloaded from
4 O’Hara et al.
Covariates related to both MSE and sexual behavior
(including sensation seeking) were measured at Time 1. Gen-
der, race, and age were reported by participants’ parents. Par-
ticipants reported how often they attended church or engaged
in religious activities, how many hours of television they
viewed each day, whether they had a television in their bed-
room, and with whom they lived (a measure used to code fam-
ily structure as either intact or divided). Participants also
completed a nine-item maternal-responsiveness measure (α =
.71) and a seven-item maternal-demandingness measure (α =
.59; Jackson, Henriksen, & Foshee, 1998). Finally, we con-
trolled for MSE that occurred between Time 2 and partici-
pants’ sexual debut.2 Including this covariate allowed us to
focus specifically on early MSE (i.e., before age 16) as a pre-
dictor of sexual debut and engagement in risky sexual behav-
iors, controlling for subsequent MSE.
Results
Descriptive statistics
The median MSE was 0.93 hr (interquartile range: 0.43 hr–
1.32 hr). Sensation seeking was generally low, M = 7.90 (SD =
2.39) at Time 1 and M = 8.07 (SD = 2.32) at Time 2. By Time
6, 774 participants (63.0%) had sexually debuted: 40 (5.2%)
before age 15, 79 (10.2%) at age 15, 190 (24.5%) at age 16,
223 (28.8%) at age 17, and 242 (31.2%) at age 18 or older.
Among sexually active participants, the median number of
lifetime sexual partners was 2 (interquartile range: 1–4 part-
ners), and 195 of these participants (25.2%) reported that they
had had casual sex without a condom.
Gender differences
Male and female participants were equally likely to have sexu-
ally debuted by Time 6; moreover, males and females sexually
debuted at approximately the same age and reported similar
MSE. Males, however, reported having more sexual partners
(M = 3.43, SD = 5.94) than did females (M = 2.48, SD = 3.91),
t(1221) = 3.48, p = .001, and engaging in casual sex without a
condom more frequently (M = 0.43, SD = 1.14) than did
females (M = 0.29, SD = 0.87), t(1223) = 2.37, p < .02. Males
also reported higher sensation seeking than females at both
Time 1 and Time 2, ts(≥ 1195) ≥ 3.70, ps < .001.
Zero-order correlations
Table 1 displays the full correlation matrix, by gender. Higher
MSE was associated with earlier sexual debut, more sexual
partners, more frequent casual sex without a condom, and
higher sensation seeking for both genders, ps < .001. The rela-
tionship between MSE and sexual debut, however, was sig-
nificantly stronger for males, r(595) = −.33, than for females,
r(585) = −.21; z = 2.19, p < .03. Higher sensation seeking
was also associated with earlier sexual debut, more sexual
partners, and more frequent casual sex without a condom
among both genders, ps < .01. Finally, earlier sexual debut was
associated with more sexual partners and more frequent casual
sex without a condom for both genders, ps < .001.
Survival analysis
To examine predictors of sexual debut, we performed a Cox
proportional-hazards regression with MSE at Times 1 and 2,
sensation seeking at Time 2, and the covariates entered into
the model (Table 2). Omnibus tests of the model’s coefficients
showed that the model was significant, χ2(13, N = 1,133) =
805.01, p < .001. The hazard ratio for MSE was 5.38, p < .001,
indicating that for each 1-hr increase in MSE on a square-root
scale, the hazard of debut at each age increased more than 5
times. Other significant predictors of sexual debut included
change in sensation seeking (hazard ratio = 1.11, p < .001),
gender (females debuted later than males; hazard ratio =
0.81, p = .006), family structure (participants from divided
homes debuted earlier than did those from intact homes; haz-
ard ratio = 1.22, p = .030), and having a television in the bed-
room (hazard ratio = 1.20, p = .024). The survival analysis was
also performed separately for each gender. The model for each
gender was significant, but the influence of MSE on sexual
debut was stronger for males (hazard ratio = 6.71, p < .001)
than for females (hazard ratio = 4.24, p < .001). Including an
MSE × Gender interaction in the model showed this difference
to be significant, p = .01 (see Fig. 1).
Structural equation model
A structural equation model predicting risky sexual behavior
at Time 6 was evaluated using the robust weighted least
squares approach in Mplus 6.12 (Muthén & Muthén, 1998-
2007). Sexual debut was recoded as an ordinal variable (1 = 14
years of age or younger, 2 = 15 years of age, 3 = 16 years of
age, 4 = 17 years of age, 5 = ≥ 18 years of age or older; partici-
pants who were virgins at Time 6 were coded as 5). MSE
summed from Times 1 and 2 was exogenous in the model;
sensation seeking at Time 2, age of sexual debut, and risky
sexual behavior at Time 6 were endogenous. Sensation seek-
ing was specified as a multi-indicator manifest variable, MSE
and age of sexual debut were specified as single-indicator
manifest variables, and risky sexual behavior at Time 6 was
specified as a latent variable with two indicators: number of
lifetime partners (factor loading = .90) and instances of casual
sex without a condom (factor loading = .81).
The structural model (Fig. 2) provided an excellent fit to
the data, χ2(12, N = 1,133) = 11.11, p > .51; root-mean-square
error of approximation (RMSEA) < .001; confirmatory fit
index = 1.00; Tucker-Lewis index > 1.00. This model explained
72% of the variance in age of sexual debut and 58% of
the variance in risky sexual behavior at Time 6. Results sup-
ported Hypothesis 1: The indirect effect of MSE on age of
sexual debut via changes in sensation seeking was significant,
at DARTMOUTH COLLEGE on July 24, 2012pss.sagepub.comDownloaded from
5
Table 1. Correlations Between Study Variables
Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14
1. Age of sexual debut — −.48*** −.37*** −.21*** −.09* −.13*** −.24*** −.30*** .12** .17*** .08* −.14*** .07 .14***
2. Total number of life-
time sexual partners
−.43*** — .56*** .22*** .06 .07 .18*** .22*** −.10* −.14*** −.09* .11** .00 .05
3. Instances of casual sex
without a condom
−.29*** .49*** — .15*** .03 .10* .12** .16*** −.15*** −.08* −.13*** .07 −.01 −.07
4. Time 1 + Time 2 MSE −.33*** .28*** .22*** — .12** .24*** .31*** .29*** −.14*** −.13** −.12** .11** −.08 .19***
5. Daily television view-
ing at Time 1
−.02 .05 .03 .05 — .22*** .15*** .13*** −.02 .01 −.01 .15*** −.13*** .06
6. Television in bedroom
at Time 1
−.09* .08* .06 .21*** .16*** — .16*** .21*** .01 −.05 −.11** .11** −.12** .01
7. Sensation seeking at
Time 1
−.28*** .26*** .24*** .37*** .07 .09* — .67*** −.16*** −.42*** −.25*** .07 −.08* .11**
8. Sensation seeking at
Time 2
−.33*** .25*** .22*** .32*** .06 .07* .65*** — −.14*** −.34*** −.23*** .11** −.10* .04
9. Engagement in
religious activities at
Time 1
.13*** −.17*** −.15*** −.14*** −.13*** −.03 −.10* −.11** — .12** .20*** −.04 −.03 −.07*
10. Maternal responsive-
ness at Time 1
.08* −.13*** −.12** −.12** −.06 −.05 −.32*** −.24*** .06 — .46*** −.06 .07 −.02
11. Maternal demanding-
ness at Time 1
.02 −.08* −.11** −.14*** −.01 .02 −.15*** −.12** .12** .33*** — .03 −.02 −.09*
12. Family structure at
Time 1
−.13** .04 .07* .17*** .13** .14*** .14*** .14*** −.07 −.12** −.02 — −.22*** −.09*
13. Race −.01 .06 .01 −.13*** −.05 −.14*** .06 .06 −.06 .08* −.05 −.09* — .03
14. Age .07 .14*** .07* .22*** .02 −.02 .13*** .12** −.05 −.08* −.16*** −.05 −.01 —
Note: Correlations below the diagonal are for males (pairwise n ≥ 591); correlations above the diagonal are for females (pairwise n ≥ 582). Movie sexual exposure (MSE) was square-root trans-
formed. Family structure at Time 1 was coded 0 for intact and 1 for divided. Race was coded 0 for minority and 1 for European American.
*p ≤ .05. **p ≤ .01. ***p ≤ .001.
at DARTMOUTH COLLEGE on July 24, 2012pss.sagepub.comDownloaded from
6 O’Hara et al.
Table 2. Results of the Cox Regression for Age of Sexual Debut
Overall Males (n = 572) Females (n = 561)
Predictor βHazard ratio βHazard ratio βHazard ratio
Media variables
Time 1 + Time 2 MSE 1.68*** 5.38*** 1.90*** 6.71*** 1.45*** 4.24***
Later MSE –2.70*** 0.07*** –2.70*** 0.07*** –2.75*** 0.06***
Daily television viewing at Time 1 0.01 1.01 0.01 1.01 0.02 1.02
Television in bedroom at Time 1 0.18* 1.20* 0.18 1.20 0.22 1.24
Family variables
Maternal responsiveness at Time 1 –0.09 0.99 –0.10 0.99 –0.01 0.99
Maternal demandingness at Time 1 0.01 1.01 0.02 1.02 0.00 1.00
Engagement in religious activities at Time 1 –0.08 0.93 –0.07 0.93 –0.07 0.93
Family structure at Time 1 0.20* 1.22* 0.21 1.23 0.17 1.18
Individual variables
Sensation seeking at Time 1 0.02 1.02 0.03 1.03 0.00 1.00
Sensation seeking at Time 2 0.10*** 1.11*** 0.12*** 1.13*** 0.09** 1.09**
Age –0.51*** 0.60*** –0.46*** 0.63*** –0.56*** 0.57***
Race 0.08 1.08 0.11 1.12 0.04 1.04
Gender –0.21** 0.81** — — — —
Note: Movie sexual exposure (MSE) was square-root transformed. Family structure was coded 0 for intact and 1 for divided. Race was coded 0
for minority and 1 for European American. Gender was coded 0 for female and 1 for male.
*p ≤ .05. **p ≤ .01. ***p ≤ .001.
10 12 14 16 18 20 22
Age (Years)
Female, High MSE
Male, High MSE
Male, Low MSE
Female, Low MSE
.0
.2
.4
.6
.8
1.0
Proportion Yet to Sexually Debut
Fig. 1. Adjusted survival curves for age of sexual debut for males and
females with high versus low movie sexual exposure (MSE), as determined on
the basis of median splits.
β = −0.01, p < .002 (MSE → changes in sensation seeking:
β = 0.09, p < .001; changes in sensation seeking → age of
sexual debut: β = −0.14, p < .001). Also, MSE directly pre-
dicted age of sexual debut, β = −0.33, p < .001. Results also
supported Hypothesis 2: MSE indirectly predicted risky
sexual behavior at Time 6. The indirect effect of MSE on risky
sexual behavior at Time 6 via age of sexual debut was signifi-
cant, β = 0.21, p < .001 (age of sexual debut → risky sexual
behaviors at Time 6: β = −0.64, p < .001), as was the indirect
effect via changes in sensation seeking and age of sexual
debut, β = 0.01, p < .005. Finally, MSE directly predicted risky
sexual behavior at Time 6, β = 0.10, p < .05.
This model was modified to allow paths to vary by gender.
The multigroup model also provided an excellent fit to the
data, χ2(43, N = 1,133) = 30.38, p > .92; RMSEA < .001; con-
firmatory fit index = 1.00; Tucker-Lewis index > 1.00. Releasing
the equality constraint on the path from MSE to age of sexual
debut significantly improved model fit, χ2(1, N = 1,133) = 8.28,
p < .005. The direct effect of MSE on age of sexual debut was
stronger for males, b = −2.41, p < .001, than for females, b =
−1.38, p < .001; however, the total indirect effects of MSE on
risky sexual behavior at Time 6 were similar for males, β =
0.24, p < .001, and females, β = 0.17, p < .001.
Discussion
Higher early MSE (before age 16) predicted more risky sexual
behaviors (i.e., a higher number of lifetime sexual partners and
more frequent casual sex without a condom) in adulthood, and
it did so both directly and indirectly, via earlier sexual debut.
This result supports previous findings that sexual media diet
predicts age of sexual debut (e.g., Brown et al., 2006), and it
at DARTMOUTH COLLEGE on July 24, 2012pss.sagepub.comDownloaded from
Movie Sexual Exposure 7
extends those findings to suggest that MSE has a lasting influ-
ence on risky sexual behaviors in adulthood (Ward et al.,
2011). MSE also predicted sexual debut indirectly through an
increase in sensation seeking. This finding provides further
evidence that exposure to movies with sexual content may
accelerate the normal rise in sensation seeking during adoles-
cence (Steinberg et al., 2008), thereby promoting risky behav-
ior generally (de Leeuw et al., 2011; Stoolmiller et al., 2010).
Finally, the influence of MSE on sexual debut and risky sexual
behavior at Time 6 was stronger among males than females,
although its influence on sensation seeking was similar
between the genders. It is worth noting that the sizes of
the effects of MSE on sexual behavior ranged from medium
(|.33|) to small (|.01|). However, the largest direct effect was
found for the influence of MSE on sexual debut. These results
suggest that MSE may have a greater impact on other potential
mediating mechanisms, such as changes in attitudes (Brown
et al., 2005) or sexual scripts (Wright, 2011). Given the preva-
lence of MSE among adolescents, we believe that even small
effects of MSE have important implications for adolescents’
sexual health.
Reducing risky sexual behaviors
Our results suggest that restricting adolescents’ MSE would
delay their sexual debut and also reduce their engagement in
risky sexual behaviors later in life. This strategy could attenu-
ate the direct influence of media on adolescents’ sexual behav-
ior by limiting the acquisition of risky sexual scripts and/or
reducing their likelihood of activation (Wright, 2011). In addi-
tion, restricting MSE may retard the increase in sensation
seeking normally experienced during adolescence (Steinberg
et al., 2008), which, in turn, could delay sexual debut and sub-
sequent engagement in risky sexual behaviors (Arnett, 1994;
Donohew et al., 2000). Limiting youths’ MSE may be a diffi-
cult task, however, given the copious amounts of sex portrayed
in movies (Gunasekera et al., 2005; Nalkur et al., 2010).
One promising approach would involve incorporating media-
literacy training into sexual education. A recent intervention
showed that a peer-led sexual-media-literacy curriculum
increased ninth-grade students’ self-efficacy in resisting peer
pressure with regard to sexual behavior, reduced their percep-
tion of the normative prevalence of sexual activity during ado-
lescence, and improved their attitudes toward abstinence
(Pinkleton, Austin, Cohen, Chen, & Fitzgerald, 2008).
Limitations and future directions
Some limitations of our study should be acknowledged. First,
participants lost at follow-up were at greater risk for early
sexual debut and risky sexual behavior than were those
retained in the study, a pattern typical in longitudinal research
Time 6
Instances of
Casual
Sex Without
Condom
R 2 = .72
Time 6
Risky Sex
R 2 = .58
Sexual
Debut
0.10*
Time 1 +
Time 2
MSE
∆ Sensation
Seeking
Time 6
Lifetime
Partners
(–0.33***)
0.09***
–0.14***
–0.64***
0.90***
0.81***
–2.41***
–1.38*** Covariates
Sensation Seeking at Time 1
Maternal Responsiveness at Time 1
Maternal Demandingness at Time 1
Engagement in Religious Activities
at Time 1
Daily TV Viewing at Time 1
TV in Bedroom at Time 1
Gender
Race
Age
Later MSE
Fig. 2. Ef fects of movie sexual exposure (M SE) on risky sexual behavior as mediated by changes in sensation
seeking and age of sexual debut. MSE was measured at Times 1 and 2; the two components of risky sexual
behavior (i.e., number of lifetime sexual partners and instances of sex without a condom) were me asured
at Time 6. Coefficients for the moderated path from MSE to sexual debut are shown separately for males
(above the line) and females (below the line), with the coefficient for the whole sample noted in parentheses.
Asterisks indicate significant paths (*p ≤ .05, ***p ≤ .0 01).
at DARTMOUTH COLLEGE on July 24, 2012pss.sagepub.comDownloaded from
8 O’Hara et al.
(Boys et al., 2005). This biased attrition may have resulted in
an underestimation of the true effect of MSE on sexual out-
comes. Second, our results may not generalize to nations with
sexual-education curricula and sexual norms that differ from
those of the United States, although effects of media exposure
on alcohol and tobacco use have been found to be similar
among U.S. adolescents and samples from other countries
(e.g., Morgenstern et al., 2011). Third, participants did not
report their sexual behavior until they were at least 18 years
old, and their retrospective memory for age of sexual debut,
number of sexual partners, and instances of casual sex without
a condom may therefore have been biased. This would be more
problematic if these biases were associated with MSE (e.g., if
adolescents who watched more movies with sexual content
were more likely to exaggerate their sexual experience).
Our data also did not include measures of other factors that
may confound relations between MSE and sexual behavior,
such as the sexual behavior of siblings and peers, parental atti-
tudes toward sex, and pubertal status (although we controlled
for age). Likewise, we were unable to examine cognitive or
psychosocial mediators of the effect of MSE on age of sexual
debut and engagement in risky sexual behaviors. Previous
studies using the same data for these mediators have shown
that movies’ effects on substance use are mediated by changes
in perceived favorability of typical substance users (i.e., sub-
stance-user prototypes), behavioral willingness to use sub-
stances, expectancies regarding substance use, and substance
use among peers (Dal Cin et al., 2009; Gibbons et al., 2010;
Wills et al., 2010). Future research should examine potential
mediators to explore why seeing sex on the big screen trans-
lates into having sex in the real world.
Future studies should also attempt to differentiate the
effects of MSE from the effects of exposure to portrayals of
other risk behaviors in popular movies, especially with regard
to changes in sensation seeking. It is unclear whether changes
in sensation seeking were related specifically to MSE or to
other co-occurring elements of adult-oriented movies (e.g.,
alcohol use; Stoolmiller et al., 2010). An important avenue for
future work will be to examine whether the effects of MSE on
sexual behavior are partly attributable to exposure to portray-
als of drinking in movies and subsequent alcohol use (e.g., Dal
Cin et al., 2009), given that adolescents’ alcohol use and risky
sexual behaviors are inherently intertwined (Cooper, 2002).
Finally, our results may have been moderated by race. Afri-
can Americans tend to sexually debut at a younger age, engage
in more risky sexual behaviors, and contract more STIs than
do European Americans (Cavazos-Rehg et al., 2009; Halpern
et al., 2004; Kaestle et al., 2005). However, African Americans
also tend to be less responsive than European Americans to
media depictions of sex (Brown et al., 2006) and alcohol use
(Gibbons et al., 2010). Unfortunately, our study’s sample
included too few African Americans for us to test moderation
by race. Future research may allow for a better understanding
of when and how movies influence youth, and how to impede
this influence in order to promote healthier sexual behavior.
Declaration of Conflicting Interests
The authors declared that they had no conflicts of interest with
respect to their authorship or the publication of this article.
Funding
This research was funded by National Institutes of Health Grants
CA077026 and AA015591 to James D. Sargent.
Supplemental Material
Additional supporting information may be found at http://pss.sagepub
.com/content/by/supplemental-data
Notes
1. Participants who were virgins at Time 6 were coded as never hav-
ing had casual sex without a condom. However, because number of
lifetime partners included oral-sex partners, 105 participants who
were virgins (23.1% of virgins) had a risky-sexual-behavior score
greater than zero.
2. For example, this measure comprised MSE at Time 3 for partici-
pants whose sexual debut was prior to Time 4, but comprised MSE
at Times 3, 4, and 5 for participants whose sexual debut was prior to
Time 6.
References
Anderson, C. A., Berkowitz, L., Donnerstein, E., Huesmann, L. R.,
Johnson, J. D., Linz, D., . . . Wartella, E. N. M. (2003). The influ-
ence of media violence on youth. Psychological Science in the
Public Interest, 4, 81–110.
Anderson, P., de Bruijn, A., Angus, K., Gordon, R., & Hastings, G.
(2009). Impact of alcohol advertising and media exposure on
adolescent alcohol use: A systematic review of longitudinal stud-
ies. Alcohol and Alcoholism, 44, 229–243.
Arnett, J. (1994). Sensation seeking: A new conceptualization and a
new scale. Personality and Individual Differences, 16, 289–296.
Ashby, S. L., Arcari, C. M., & Edmonson, M. B. (2006). Televi-
sion viewing and risk of sexual initiation by young adolescents.
Archives of Pediatric and Adolescent Medicine, 160, 375–380.
Bleakley, A., Hennessy, M., Fishbein, M., Coles, H. C., & Jordan,
A. (2009). How sources of sexual information relate to adoles-
cents’ beliefs about sex. American Journal of Health Behavior,
33, 37–48.
Boys, A., Marsden, J., Stillwell, G., Hatchings, K., Griffiths, P., &
Farrell, M. (2005). Minimizing respondent attrition in longitudi-
nal research: Practical implications from a cohort study of adoles-
cent drinking. Journal of Adolescence, 26, 363–373.
Brown, J. D., Halpern, C. T., & L’Engle, K. L. (2005). Mass media
as a sexual super peer for early maturing girls. Journal of Adoles-
cent Health, 36, 420–427.
Brown, J. D., L’Engle, K. L., Pardun, C. J., Guo, G., Kenneavy, K.,
& Jackson, C. (2006). Sexy media matter: Exposure to sexual
content in music, movies, television, and magazines predicts
Black and White adolescents’ sexual behavior. Pediatrics, 11 7 ,
1018–1027.
Cavazos-Rehg, P. A., Krauss, M. J., Spitznagel, E. L., Schootman,
M., Bucholz, K. K., Peipert, J. F., . . . Bierut, L. J. (2009). Age of
at DARTMOUTH COLLEGE on July 24, 2012pss.sagepub.comDownloaded from
Movie Sexual Exposure 9
sexual debut among U.S. adolescents. Contraception, 80, 158–
162.
Collins, R. L., Elliott, M. N., Berry, S. H., Kanouse, D. E., Kunkel,
D., Hunter, S. B., & Miu, A. (2004). Watching sex on television
predicts adolescent initiation of sexual behavior. Pediatrics, 114,
e280–e289. Retrieved from http://www.pediatricsdigest.mobi/
content/114/3/e280.full
Cooper, M. L. (2002). Alcohol use and risky sexual behavior among
college students and youth: Evaluating the evidence. Journal of
Studies on Alcohol, 14, 101–117.
Dal Cin, S., Worth, K. A., Gerrard, M., Gibbons, F. X., Stoolmiller,
M., Wills, T. A., & Sargent, J. D. (2009). Watching and drinking:
Expectancies, prototypes, and friends’ alcohol use mediate the
effect of exposure to alcohol use in movies on adolescent drink-
ing. Health Psychology, 28, 473–483.
Dalton, M. A., Beach, M. L., Adachi-Mejia, A. M., Longacre,
M. R., Matzkin, A. L., Sargent, J. D., . . . Titus-Ernstoff, L. (2009).
Early exposure to movie smoking predicts established smoking
by older teens and young adults. Pediatrics, 123, 551–558.
de Leeuw, R. N. H., Sargent, J. D., Stoolmiller, M., Scholte, R. H. J.,
Engels, R. C. M. E., & Tanski, S. E. (2011). Association of smok-
ing onset with R-rated movie restrictions and adolescent sensa-
tion seeking. Pediatrics, 127, 96–105.
Donohew, L., Zimmerman, R., Cupp, P. S., Novak, S., Colon, S., &
Abell, R. (2000). Sensation seeking, impulsive decision-making,
and risky sex: Implications for risk-taking and design of inter-
ventions. Personality and Individual Differences, 28, 1079–1091.
Gibbons, F. X., Pomery, E. A., Gerrard, M., Sargent, J. D., Weng,
C., Wills, T. A., . . . Yeh, H. (2010). Media as social influence:
Racial differences in the effects of peers and media on adolescent
alcohol cognitions and consumption. Psychology of Addictive
Behaviors, 24, 649–659.
Gunasekera, H., Chapman, S., & Campbell, S. (2005). Sex and drugs
in popular movies: An analysis of the top 200 films. Journal of
the Royal Society of Medicine, 98, 464–470.
Guttmacher Institute. (2010). U.S. teenage pregnancies, births, and
abortions: National and state trends and trends by race and eth-
nicity. Retrieved from http://www.guttmacher.org/pubs/USTP-
trends.pdf
Halpern, C. T., Hallfors, D., Bauer, D. J., Iritani, B., Waller, M. W.,
& Cho, H. (2004). Implications of racial and gender differences
in patterns of adolescent risk behavior for HIV and other sexually
transmitted diseases. Perspectives on Sexual and Reproductive
Health, 36, 239–247.
Jackson, C., Henriksen, L., & Foshee, V. A. (1998). The Authoritative
Parenting Index: Predicting health risk behaviors among children
and adolescents. Health Education & Behavior, 25, 319–337.
Kaestle, C. E., Halpern, C. T., Miller, W. C., & Ford, C. A. (2005).
Young age at first sexual intercourse and sexually transmitted
infections in adolescents and young adults. American Journal of
Epidemiology, 161, 774–780.
L’Engle, K. L., Brown, J. D., & Kenneavy, K. (2006). The mass
media are an important context for adolescents’ sexual behavior.
Journal of Adolescent Health, 38, 186–192.
Martino, S. C., Collins, R. L., Kanouse, D. E., Elliott, M., & Berry,
S. H. (2005). Social cognitive processes mediating the relation-
ship between exposure to television’s sexual content and ado-
lescents’ sexual behavior. Journal of Personality and Social
Psychology, 89, 914–924.
Morgenstern, M., Poelen, E. A. P., Scholte, R., Karlsdottir, S.,
Jonsson, S. H., Mathis, F., . . . Hanewinkel, R. (2011). Smoking
in movies and adolescent smoking: Cross-cultural study in six
European countries. Thorax, 66, 875–883.
Muthén, L. K., & Muthén, B. O. (1998–2007). Mplus user’s guide
(5th ed.). Los Angeles, CA: Author.
Nalkur, P. G., Jamieson, P. E., & Romer, D. (2010). The effectiveness
of the Motion Picture Association of America’s rating system in
screening explicit violence and sex in top-ranked movies from
1950 to 2006. Journal of Adolescent Health, 47, 440–447.
National Cancer Institute. (2008, June). The role of the media in pro-
moting and reducing tobacco use (Tobacco Control Monograph
No. 19; NIH Pub. No. 07–6242). Bethesda, MD: U.S. Depart-
ment of Health and Human Services, National Institutes of
Health, National Cancer Institute.
Pardun, C. J., L’Engle, K. L., & Brown, J. D. (2005). Linking expo-
sure to outcomes: Early adolescents’ consumption of sexual con-
tent in six media. Mass Communication and Society, 8, 75–91.
Pinkleton, B. E., Austin, E. W., Cohen, M., Chen, Y., & Fitzgerald, E.
(2008). Effects of a peer-led media literacy curriculum on adoles-
cents’ knowledge and attitudes toward sexual behavior and media
portrayals of sex. Health Communication, 23, 462–472.
Sargent, J. D., Stoolmiller, M., Worth, K. A., Dal Cin, S., Wills,
T. A., Gibbons, F. X., . . . Tanski, S. (2007). Exposure to smok-
ing depictions in movies: Its association with established ado-
lescent smoking. Archives of Pediatric & Adolescent Medicine,
161, 849–856.
Sargent, J. D., Wills, T. A., Stoolmiller, M., Gibson, J., & Gibbons,
F. X. (2006). Alcohol use in motion pictures and its relation with
early-onset teen drinking. Journal of Studies on Alcohol, 67,
54–65.
Sargent, J. D., Worth, K. A., Beach, M., Gerrard, M., & Heatherton,
T. F. (2008). Population-based assessment of exposure to risk
behavior in motion pictures. Communication Methods and Mea-
sures, 2, 134–151.
Steinberg, L., Albert, D., Cauffman, E., Banich, M., Graham, S., &
Woolard, J. (2008). Age differences in sensation seeking and
impulsivity as indexed by behavior and self-report: Evidence for a
dual systems model. Developmental Psychology, 44, 1764–1778.
Stephenson, M. T., Hoyle, R. H., Palmgreen, P., & Slater, M. D.
(2003). Brief measures of sensation seeking for screening and
large-scale surveys. Drug and Alcohol Dependence, 72, 279–
286.
Stoolmiller, M., Gerrard, M., Sargent, J. D., Worth, K. A., & Gibbons,
F. X. (2010). R-rated movie viewing, growth in sensation seek-
ing, and alcohol initiation: Reciprocal and moderation effects.
Prevention Science, 11, 1–13.
Ward, L. M., Epstein, M., Caruthers, A., & Merriwether, A. (2011).
Men’s media use, sexual cognitions, and sexual risk behavior:
at DARTMOUTH COLLEGE on July 24, 2012pss.sagepub.comDownloaded from
10 O’Hara et al.
Testing a mediational model. Developmental Psychology, 47,
592–602.
Weinstock, H., Berman, S., & Cates, W. (2000). Sexually transmitted
diseases among American youth: Incidence and prevalence esti-
mates, 2000. Perspectives on Sexual and Reproductive Health,
36, 6–10.
Wills, T. A., Gibbons, F. X., Sargent, J. D., Gerrard, M., Lee, H.-R.,
& Dal Cin, S. (2010). Good self-control moderates the effect of
mass media on adolescent tobacco and alcohol use: Tests with
studies of children and adolescents. Health Psychology, 29, 539–
549.
Wright, P. J. (2011). Mass media effects on youth sexual behavior:
Assessing the claim for causality. Communication Yearbook, 35,
343–386.
Zuckerman, M. (1994). Behavior expressions and biosocial bases of
sensation seeking. New York, NY: Cambridge University Press.
at DARTMOUTH COLLEGE on July 24, 2012pss.sagepub.comDownloaded from
