Prevention Science

https://doi.org/10.1007/s11121-019-0984-z

Sensation Seeking and Impulsivity Can Increase Exposure to Risky Media

and Moderate Its Effects on Adolescent Risk Behaviors
Atika Khurana 1 & Amy Bleakley 2 & Morgan E. Ellithorpe 3 & Michael Hennessy 2 & Patrick E. Jamieson 4 & Ilana Weitz 4

# Society for Prevention Research 2019

Abstract
Media exposure to risky behaviors (e.g., alcohol use, violence) has been associated with adolescent engagement in risk-taking
behaviors, but not all adolescents are equally at risk. Here we focus on individual differences in impulsivity and sensation seeking
and assess their effects on the relation between media risk exposure and adolescent risk behavior. Survey data from 1990 Black
and White US adolescents (mean age = 15.6 ± 1.10 years; 48% female) and content analysis of top-grossing films and popular
TV shows were analyzed using linear regression models. High levels of impulsivity and sensation seeking were associated with
greater exposure to risky media content, and also operated as moderators, exacerbating the impact of media risk exposure on
adolescent risk behaviors. Prevention efforts targeting negative effects of media on adolescent health should prioritize youth with
high levels of impulsivity and sensation seeking.

Keywords Media effects . Adolescent risk behaviors . Impulsivity . Sensation seeking

Adolescents are spending increasing amounts of time with Peter 2013). By focusing on individual differences in impul-
media (Lenhart 2015; Rideout 2016; Rideout et al. 2010). sivity and sensation seeking (SS), this study examines the role
TV and movie exposure to risk behaviors is prevalent among of these personality dimensions in better understanding the
adolescents and is linked to involvement in risk-taking behav- media exposure–risk behavior relationship in adolescents.
iors (Brown and Bobkowski 2011; Strasburger et al. 2010).
These effects have been reported for alcohol use (Hanewinkel
and Sargent 2009; Wills et al. 2012), sexual risk-taking
(Bleakley et al. 2008b; Brown et al. 2006; Collins et al. Impulsivity, Sensation Seeking,
2004; Hennessy et al. 2009), and aggressive behaviors and Adolescent Risk-Taking
(Bushman and Huesmann 2006). Although past studies have
often assumed that the effect of media exposure on risk be- Impulsivity and SS are robust predictors of adolescent risk be-
haviors operates similarly for all adolescents, some adoles- haviors. Impulsivity is typically defined as a tendency to act on
cents may be more vulnerable than others (Valkenburg and the spur of the moment without adequate consideration of the
consequences and is assessed using self-report measures such as
the Barratt (Patton et al. 1995) or Eysenck (Eysenck et al. 1984)
* Atika Khurana scales. SS on the other hand reflects a propensity toward novel or
atika@uoregon.edu exciting behaviors that is considered to be evolutionarily adaptive
(Spear, 2007) and is commonly assessed using self-report mea-
1
College of Education, University of Oregon, 1655 Alder St,
sures such as Zuckerman’s sensation seeking scale (Hoyle et al.
Eugene, OR 97403, USA 2002; Zuckerman 2006). Developmental studies find that SS
2
Annenberg School for Communication, University of Pennsylvania,
exhibits a universal peak during mid-late adolescence and de-
Philadelphia, PA, USA clines thereafter (Collado et al. 2014; Khurana et al. 2018).
3
Department of Advertising & Public Relations, College of
Impulsivity is a complex multidimensional construct
Communication Arts & Sciences, Michigan State University, East (Whiteside & Lynam 2001). Some dimensions of impulsivity,
Lansing, MI, USA such as the one assessed in this study (i.e., acting without think-
4
Annenberg Public Policy Center, University of Pennsylvania, ing), exhibit a peak during adolescence but only for a sub-group
Philadelphia, PA, USA of youth who have an underlying weakness in cognitive control
 Prev Sci

(Khurana et al. 2018). These difficulties in impulse control can be movie) exposure to alcohol use, sexual content, and violence,
identified at younger ages (Iacono et al. 2008) and tend to get which in turn is linked to elevated risk for involvement in
exacerbated during adolescence when there is a concomitant rise corresponding risk behaviors? (2) Do individual differences
in reward sensitivity (Bjork and Pardini 2015; Moffitt et al. in SS and impulsivity moderate the relationship between me-
2011). Unlike impulsivity, SS does not reflect a lack of top- dia (TV and movie) risk exposure and real-world risk-taking
down control over behavioral urges. Instead, SS tends to be for alcohol use, sexual behaviors and violence? (3) Does ad-
positively related to cognitive control measures like working olescent race moderate the pathways of influence described in
memory (Khurana et al. 2018) and IQ (Raine et al. 2002). It is RQ1 and RQ2, i.e., do the direct, mediated, or moderated
therefore not surprising that while SS is linked to experimenta- effects vary for White and Black adolescents? Alcohol use,
tion with drugs and sexual behaviors, it is not associated with sexual risk, and violence were modeled as separate outcomes.
more adverse forms of risk-taking such as progressive drug use
or dependence that reflect a loss of control (Khurana et al. 2015a;
Khurana et al. 2017). Methods
Guided by the BUses and Gratifications^ theory (Katz et al.
1974; Rubin 2009), several studies have linked SS and impul- This study includes data from a content analysis of popular
sivity to seeking out risky content in the media. Adolescents entertainment media titles and an online survey administered
are not mere passive recipients but instead select media pur- to a sample of 2424 adolescents aged 14–17 years, recruited
posefully to gratify their needs. Such needs could be develop- from online opt-in, volunteer panels through survey company
mental (e.g., identity development) or to validate a particular GfK between November 13 and December 14, 2015. Most
behavior (e.g., validating aggressive tendencies by engaging respondents (76%) were recruited through their parents; re-
with violent media). Studies have found that adolescents high maining were recruited directly. Potential respondents were
in SS are more likely to be exposed to or seek out sexually screened for whether they were a teen aged 14–17, parent of
explicit media content (Bleakley et al. 2011a; Brown and a teen aged 14–17, or not eligible. Parents were asked to
L’Engle 2009; Kim et al. 2006; Slater 2003). Individuals with consent for their child to participate before being asked to
high levels of impulsivity also report greater exposure to vio- bring their teen to complete the rest of the survey. All teens
lent media (Chory and Goodboy 2010; Krcmar and Kean were given assent information before beginning the survey.
2005). Thus, adolescents high in SS and impulsivity may be The survey procedures were approved by the Institutional
at greater risk for experiencing negative effects of media risk Review Board of the sponsoring institution. The survey had
exposure (e.g., involvement in risk-taking behaviors) in part a median length of 26 min. Respondents received an incentive
because of their higher exposure to risky content in the media for participating (e.g., panel points from survey firm). The
(Anderson and Bushman 2001; Strasburger et al. 2010). purpose of the larger study was to examine Black and White
Alternatively, it is also possible that adolescents high in differences in media exposure and its effects on risk behav-
impulsivity and SS are more susceptible to the negative effects iors; therefore, Black adolescents were oversampled for a
of media exposure on risk behaviors. This idea is consistent roughly equal sample of Black and White respondents.
with the Differential Susceptibility Model (Valkenburg and Given the study design, we selected the non-Hispanic Black
Peter 2013) which posits that some youth experience greater (n = 1000) and non-Hispanic White (n = 990) participants to
susceptibility to media effects based on personality character- be included in the current sample (total n = 1990; mean age =
istics, moods, and cognitions. This model argues that person- 15.6 ± 1.1 years; 48.8% female).
ality dimensions such as impulsivity and SS may operate as
moderators (as well as mediators) upon which effects of media Content Analysis
would be conditional. Adolescents high in impulsivity and/or
SS may be more likely to enact media depictions of risk be- Movie Sample The top 30 grossing movies of 2014 according
haviors in the real-world either due to the rewarding aspect or to Variety magazine were selected as representing popular
because of their relative inability to control behavioral urges mainstream movies. Of the top 500 movies for 2013 and
and evaluate long-term consequences. We hypothesized that 2014 (1000 total), according to www.boxofficemojo.com, 33
the association between media risk exposure and risk behavior Black-oriented movies were selected based on following
would be positive and stronger in magnitude among adoles- criteria: (1) Black actors comprised half or more of the main
cents high in impulsivity and SS (Gibbons et al. 2016; Slater characters and/or (2) the movie had a Black-oriented narrative
et al. 2004; Wills et al. 2010). (Allen et al. 1989; Schooler et al. 2004; Sheridan 2006). Top-
By focusing on three common adolescent risk behaviors, grossing movies of both 2013 and 2014 were included for
i.e., alcohol use, sexual risk, and violence, this study ad- Black-oriented movies in order to have a large enough sample
dressed the following research questions: (1) Are high levels for comparison purposes. The final sample included 29 main-
of SS and impulsivity associated with greater media (TV and stream movies and 34 Black-oriented movies (one film
Prev Sci

originally coded with mainstream movies (Ride Along) met Online Survey
the criterion for a predominately Black cast).
Media Exposure to Risk Behaviors Exposure to media content
TV Show Sample Nielsen statistics for adolescents aged 14– was calculated for alcohol, sex, and violence using content
17 were used to determine the narrative television shows analysis of movies and TV shows and participants’ exposure
from the 2014 to 2015 season (from September 22, 2014 to film titles and TV shows. Participants were presented with
and June 28, 2015) that would be coded. The lists were the list of coded films (randomly ordered) in a grid and asked
separated by Black and non-Black adolescents because to indicate whether they had never seen the film, seen the film
Black and non-Black adolescents tend to watch different once, or seen the film more than once. Similarly, they were
kinds of shows (Ellithorpe and Bleakley 2016) in that ad- presented with additional grids for the coded television show
olescents in general tend to gravitate toward content with titles and asked to indicate how often they had watched each
more characters from their own groups. Thus, television show. Content-specific exposure was operationalized by mul-
shows popular with Black adolescents will have more tiplying the proportion of segments for each film containing
Black characters than shows popular with non-Black ado- the risk behavior by each participant’s self-reported exposure
lescents. We created lists of the top 30 shows watched by to the film (indicated on the survey for whether they had seen
Black adolescents and the top 30 shows watched by non- each film: never (0), once (1), or more than once (2)) or tele-
Black adolescents (including primarily White viewers but vision show (how often they watched each coded television
also Hispanic, Asian, and others). Old shows airing repeat show in the past year, on a scale from 0 = never to 3 = often).
episodes and syndicated shows were also included but only Using a previously tested approach (Bleakley et al. 2008a;
when the number of repeats aired were equivalent to at Bleakley et al. 2011b; Hennessy et al. 2009), these scores were
least of one season of that show (Bleakley et al. 2017). summed across all movie and all television titles, separately, to
Four shows appeared on both the top 30 for non-Black and create a measure of exposure to each risk in movies and in
top 30 for list for Blacks; thus, 56 shows were coded. Episodes television. The measures were then standardized.
from the 2014–2015 season were coded, with the exception of
a handful of shows in syndication or reruns (e.g., How I Met Impulsivity We used nine yes/no items adapted from the
Your Mother) in which case the last season produced was Junior Eysenck Impulsivity Scale (Eysenck et al. 1984)
coded. Either three or five episodes were randomly selected (e.g., Do you usually do and say things without stopping to
from each season for coding (Manganello, Franzini, & Jordan, think?). Scores were averaged to create a composite index
2008). If a show had ten or fewer episodes in the 2014–2015 (range = 0–1; M = 0.47, SD = 0.31, α = 0.80).
season, three episodes were randomly selected (n = 5, 8.9% of
shows). If a show had more than ten episodes that season, five Sensation Seeking We used four items (e.g., BI like to do
episodes were randomly selected (n = 47, 83.9% of shows). If frightening things^) representing each of the four dimensions
a show had a 15-min runtime, six episodes were selected to (i.e., experience seeking, boredom susceptibility, thrill/
equate the amount coded to a half-hour show (n = 4, 7.1% of adventure seeking, and disinhibition) of the Brief Sensation
shows). Seeking Scale (Hoyle et al. 2002). Responses were coded on a
5-point scale ranging from strongly disagree to strongly agree
Content Coding Procedures The television and movie con- and averaged to create a composite score (range = 1–5; M =
tents were coded by trained coders in 5-min segments using 3.29, SD = 1.00, α = 0.87).
a directed, quantitative, previously validated coding scheme
(Bleakley et al. 2012; Bleakley et al. 2014; Jamieson and Alcohol Use Drinking frequency was assessed by asking about
Romer 2008). After multiple training sessions, the coders the number of times the respondent had had at least one drink
achieved inter-coder reliability for identifying the presence of alcohol in the past 6 months (Centers for Disease Control
of each behavior as calculated by Krippendorff’s alpha and Prevention 2015) (range: never (0) to about every day (5)
(Hayes and Krippendorff 2007) using a separately validated (M = 0.58, SD = 1.04).
test sample of 59 segments. Each segment was coded for the
portrayal of alcohol, sex, and/or violence. Alcohol portrayal Sexual Activity Progression in sexual activity was measured
was defined as a character being directly involved any activity using an ordered (Guttman) scale of the following dichoto-
related to alcohol, ranging from handling of alcohol bottles to mous items assessing activities engaged in the past 6 months:
observed consumption (α = 0.94). Sexual behavior was de- kissed, touched each other over clothes, touched breasts/had
fined as any type of sexual contact, from kissing on the lips breasts touched, touched a partner’s private parts, saw a part-
to explicit intercourse (α = 0.93). Violence was defined as ner naked, was naked with partner, received oral sex, had
initiated or received intentional acts to inflict injury or harm vaginal sex, and gave oral sex (Hennessy et al. 2008). The
(α = 0.94). items are listed by difficulty: the Loevinger’s H coefficient
 Prev Sci

was 0.89 (van Schuur 2003) and KR-20 for the scale was 0.94 Results
(M = 2.11, SD = 3.01).
Correlations and means (SD) of study variables are presented
Violence This variable was assessed by asking about involve- in Table 1. The media exposure–risk behavior association was
ment in a physical fight in the past 6 months (Centers for significant in case of sexual risk, alcohol use, and violence, for
Disease Control and Prevention 2015). Response options both TV and movie exposure. Impulsivity and SS were posi-
ranged from 0 times (0) to 12 or more times (7). Given the tively associated with TV and movie exposure to sex, alcohol
skewed distribution, the highest three categories with low fre- use, and violence with involvement in all three risk behavior
quencies were collapsed, with the new variable range of 0 outcomes, with impulsivity having stronger associations than
times (0) to 6–7 times or more (4) (M = 0.44, SD = 0.90). SS.
In the model predicting sexual activity, we found direct
Covariates Age, gender (male = 0, female = 1), perceived effects of media risk exposure and of SS and impulsivity.
parental monitoring, maternal education, and daily TV time Some of the effect of impulsivity and SS on sexual activity
were included as covariates because of their relationship to was mediated by media exposure to sexual content. (See
media exposure and/or risk behaviors. In all models, race indirect effect estimates in Table 2). Adolescents with high
(White = 0; Black = 1) was examined as a moderator for levels of impulsivity and SS reported greater exposure to me-
direct, indirect, and moderated effects. If the moderated effect dia sexual content, which in turn was associated with higher
was non-significant, race was included as a control in the levels of sexual involvement. Impulsivity and SS did not mod-
model. Perceived parental monitoring, or the extent to which erate the direct or mediated effects. Further, no racial differ-
an adolescent believes his or her parents knows about their ences were observed in the direct or mediated effects, with the
whereabouts/activities, was measured by eight items ranging exception of SS’s effect on sexual behaviors which tended to
from never (1) to always (5) (α = 0.92, M = 3.80, SD = 0.94) be significant for White youth (B (SE) = 0.26 (0.09), p =
(Kerr et al. 2010). Maternal education was assessed using five 0.003) but not for Black youth (B (SE) = 0.06 (0.09), p =
categories: 1 (some high school) to 5 (graduate degree). We 0.51). However, this moderated effect (i.e., SS × race) (B
calculated daily TV time by asking participants how many (SE) = − 0.37 (0.13), p = 0.004) became non-significant (B
hours they spent watching television in three time periods (SE) = − 0.20 (0.12), p = 0.09), controlling for the effect of
the previous day: before noon, between noon and 6 pm, and alcohol use, and thus was not retained in the final model.
after 6 pm. In order to have a maximum of 24 h, responses The direct effect of TV sex exposure on sexual activity was
greater than 6 h for the time period between noon and 6 pm also not significant at p < 0.05, possibly due to its strong as-
(n = 41, 2.09%) were recoded as 6 h, and responses greater sociation with movie sex exposure (r = 0.69).
than 9 h were recoded as nine for the other two time periods In case of alcohol use, we observed direct positive effects
(before noon n = 28, 1.42%; after 6 pm n = 15, 0.08%). of TV and movie alcohol exposure as well as of impulsivity
Responses were then summed (M = 5.50, SD = 4.61; medi- and SS. Some of the effect of impulsivity and SS was
an = 4.5). channeled through the association with media exposure to
alcohol use. Additionally, SS moderated the effect of TV and
Statistical Analysis movie alcohol exposure on drinking (see Table 2).
Specifically, the association between media alcohol exposure
Descriptive analyses were conducted using t tests or chi- and self-reported drinking frequency was significant only at
square as appropriate, and correlations among the model high levels of SS (i.e., values ≥ 3; variable range = 1–5) (See
variables were assessed. Separate linear regression models Fig. 1a, b). There was also a significant interaction between
were used for the three behavioral outcomes, with robust SS and race. SS was significantly positively associated with
estimation procedures to account for any violations of drinking frequency only in case of White (B (SE) = 0.20
normality. Alcohol use was included as an additional con- (0.03), p < 0.001) but not Black youth (B (SE) = 0.03 (0.03),
trol in models predicting sexual activity and violence. p = 0.36). Race did not moderate any of the other direct or
Descriptive analyses were conducted in STATA 14.0. mediated effects.
Final models were estimated in Mplus v8, with mediated For violence, we observed direct effects of TV and movie
and moderated effects tested in the same model. violence exposure as well as direct effects of impulsivity and
Moderation effects were tested as two-way and three- SS. Some of the effects of impulsivity and SS on violence
way interactions by using product terms with mean cen- were channeled through their association with media exposure
tered variables (Jaccard and Turrisi 2003). The interaction to violence. Both impulsivity and SS also moderated the ef-
effects were tested individually, including main effects of fects of media violence exposure on violence, such that the
predictor variables and covariates in the same model. The association between media violence exposure (TV and
data presented here are unweighted. movies) on violent behavior outcome was stronger for youth
 Prev Sci

Table 1 Correlation matrix of key study variables

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

1 Sexual risk 1.00

2 Alcohol use 0.48 1.00
3 Violence 0.28 0.32 1.00
4 TV sex 0.23 0.19 0.22 1.00
exposure
5 Movie sex 0.25 0.20 0.24 0.69 1.00
exposure
6 TV alcohol 0.25 0.21 0.20 0.96 0.64 1.00
exposure
7 Movie alcohol 0.25 0.19 0.23 0.69 0.99 0.64 1.00
exposure
8 TV violence 0.22 0.17 0.23 0.95 0.66 0.91 0.66 1.00
exposure
9 Movie violence 0.18 0.13 0.18 0.60 0.83 0.58 0.83 0.63 1.00
exposure
10 Impulsivity 0.27 0.22 0.25 0.14 0.14 0.15 0.14 0.14 0.12 1.00
11 Sensation 0.19 0.18 0.07 0.15 0.12 0.15 0.12 0.15 0.14 0.30 1.00
seeking
12 Parental − 0.17 − 0.17 − 0.20 − 0.01 − 0.04 0.004 − 0.03 − 0.02 0.04 − 0.18 − 0.09 1.00
monitoring
13 Gender − 0.10 − 0.05 − 0.14 0.002 0.03 − 0.07 0.02 − 0.06 − 0.07 − 0.04 − 0.01 0.07 1.00
(female = 1)
14 Age 0.20 0.18 0.03 0.02 0.05 0.03 0.05 0.003 − 0.02 − 0.04 − 0.04 − 0.12 0.01 1.00
15 Daily TV time 0.09 0.09 0.05 0.29 0.28 0.27 0.28 0.29 0.27 0.11 0.08 − 0.03 0.03 − 0.02 1.00
16 Race 0.01 − 0.11 0.08 0.24 0.31 0.19 0.31 0.25 0.19 0.003 − 0.03 − 0.22 0.03 − 0.002 0.16 1.00
(Black = 1)
17 Maternal − 0.04 − 0.05 − 0.10 − 0.11 − 0.08 − 0.12 − 0.06 − 0.12 − 0.06 − 0.09 0.01 0.13 0.01 − 0.01 − 0.05 − 0.03 1.00
education
Range; mean 0–9; 0–5; 0–4; 0–19.88; 0–18.44; 0–27.14; 0–19.10; 0–39.07; 0–38.02; 0–1; 1–5; 1–5; 0/1 14–17; 0–24; 0/1 1–5;
(SD) 2.11 0.58 0.44 4.21 3.30 6.56 3.45 8.36 10.20 0.47 3.29 3.80 15.59 5.50 3.12
(3.01) (1.04) (0.90) (3.56) (3.12) (5.24) (3.29) (6.93) (6.89) (0.31) (1.00) (0.94) (1.10) (4.61) (1.19)

Values in bold denote coefficients not significant at p < 0.05. Mean and SD reported for continuous variables only
 Prev Sci

Table 2 Unstandardized regression coefficients and standard errors for models predicting sexual risk, alcohol use, and violence

Pathways of influence Sexual risk Alcohol use Violence

B (SE) p value B (SE) p value B (SE) p value

Direct effects: predictors

TV exposure 0.05 (0.03) 0.08 0.02 (0.01) < 0.001 0.01 (0.004) = 0.01
Movie exposure 0.10 (0.03) = 0.001 0.04 (0.01) < 0.001 0.01 (0.004) = 0.001
Impulsivity 1.36 (0.21) < 0.001 0.43 (0.08) < 0.001 0.58 (0.07) < 0.001
Sensation seeking 0.16 (0.07) = 0.02 0.20 (0.03) < 0.001 0.01 (0.02) 0.72
Direct effects: covariates
Age 0.36 (0.05) < 0.001 0.14 (0.02) < 0.001 − 0.05 (0.02) = 0.006
Gender (Female =1) − 0.61 (0.11) < 0.001 − 0.05 (0.04) 0.28 − 0.17 (0.04) < 0.001
Race (Black = 1) 0.003 (0.13) = 0.98 − 0.39 (0.05) < 0.001 0.08 (0.04) = 0.06
Maternal Education 0.02 (0.05) = 0.66 0.002 (0.02) 0.91 − 0.04 (0.02) = 0.04
Parental Monitoring − 0.17 (0.07) = 0.02 − 0.15 (0.02) < 0.001 − 0.08 (0.03) = 0.001
TV time 0.01 (0.02) 0.38 0.01 (0.01) 0.13 0.001 (0.01) 0.84
Alcohol use 1.05 (0.07) < 0.001 0.19 (0.03) < 0.001
Indirect effects
Impulsivity 0.10 (0.04) = 0.007 0.06 (0.02) < 0.001 0.13 (0.03) < 0.001
Sensation seeking 0.05 (0.01) = 0.001 0.03 (0.01) < 0.001 0.05 (0.01) < 0.001
Moderated effects
Impulsivity × TV exposure – – – – 0.04 (0.01) < 0.001
Impulsivity × movie exposure – – – – 0.04 (0.01) < 0.001
Sensation seeking × TV exposure – – 0.01 (0.004) 0.01 0.01 (0.003) = 0.05
Sensation seeking × movie exposure – – 0.01 (0.01) = 0.05 0.01 (0.003) = 0.03
Race × sensation seeking – – − 0.17 (0.04) < 0.001 – –

Interaction effects were retained in the model only when significant. Indirect effects of impulsivity and sensation seeking reflect the total indirect effect,
involving mediated pathways through TVexposure and movie exposure. In the moderation models, the interactions of TVexposure and movie exposure
with impulsivity and sensation seeking were tested in separate models. Data rendered in bold represent effects are non-significant at p < 0.05

with high levels of impulsivity and SS as compared to those impulsivity and SS moderated the media exposure–risk be-
with lower levels. Specifically, the relation between media havior link such that the association between media exposure
(TV and movie) exposure to violence and involvement in and corresponding risk behavior would be stronger at higher
physical fights was significant only at high levels of impulsiv- levels of impulsivity and SS than at lower levels. We found
ity (i.e., values ≥ 0.4; variable range = 0–1) and SS (i.e., values evidence for moderation in case of alcohol use (for SS only)
≥ 3; variable range = 1–5). (See Figs. 2 and 3). No race differ- and violence (for both impulsivity and SS) but not for sexual
ences were observed. behaviors. In terms of race differences, we found that the
positive association between SS and alcohol use was signifi-
cant only in case of White youth and not for Black youth. Race
Discussion failed to moderate the direct effect of SS on sexual behaviors
when controlling for the effect of alcohol use. No other race
The purpose of this study was to examine the role of individ- differences were observed. Overall, our findings suggest that
ual differences in impulsivity and SS in understanding the prevention efforts aimed at reducing negative effects of risky
association between media risk exposure and adolescent risk media on adolescent risk behaviors could benefit from
behaviors. Specifically, we tested if adolescents high in im- targeting youth high in impulsivity and SS.
pulsivity and SS reported greater exposure to TV and movie Consistent with past studies (Bleakley et al. 2011a; Brown
risk content and whether that in turn was associated with and L’Engle 2009; Kim et al. 2006; Slater 2003), adolescents
greater involvement in risky behaviors. We found that both high in impulsivity and SS reported greater media risk expo-
impulsivity and SS were positively associated with media risk sure, which in turn was positively associated with involve-
exposure, which in turn was positively related to real-world ment in risk behaviors. The effects of impulsivity on sexual
risk-taking. We also tested if individual differences in activity and violence were more pronounced than SS. In case
Prev Sci

Fig. 1 a, b Interaction effects of

media (TV and movie) exposure
and sensation seeking (SS) on al-
cohol use

of alcohol use, the direct effects of impulsivity and SS were Violent behaviors on the other hand are not as normative and
comparable, likely because experimenting with alcohol use is are more likely to be influenced by a lack of self-control (i.e.,
a more normative type of adolescent risk-taking that is often inability to control aggressive tendencies) and/or a rewarding
associated with social rewards (Balsa et al. 2011). For sexual component (e.g., deviant peer group conformity). Studies find
behaviors and violence, lack of impulse control is expected to that chronic exposure to violent forms of media (e.g.,
play a larger role in predicting behavior than SS (Khurana videogames) is associated with aggressive/violent tendencies,
et al. 2012, 2015b). but consistent with our findings, these effects are stronger for
Both impulsivity and SS moderated the effect of media risk individuals with poor impulse control (Kronenberger et al.
exposure in case of violence. But for alcohol use, only SS 2005). Therefore, adolescents with impulse control difficulties
moderated the effect of TV alcohol exposure. It is possible and/or strong reward sensitivity may be more likely to act on
that because experimentation with alcohol use is socially ac- aggressive tendencies fueled by media violence exposure.
cepted and normative, adolescents high in SS may be more In terms of race differences, the direct effect of SS on al-
strongly influenced by media depictions of drinking behaviors cohol use was significant only in case of White youth. This
and engage in these behaviors for the rewarding component. may be because White youth are more likely to experiment
 Prev Sci

Fig. 2 a, b Interaction effects of

media (TV and movie) exposure
and impulsivity (IMP) on
violence

with alcohol and have greater access to alcohol than their (Jaeggi et al. 2011) and targeted drinking behaviors
Black counterparts (Blum et al. 2000). This finding also high- (Houben et al. 2011); however, no study to date has ex-
lights the fact that even though SS exhibits a universal peak amined effects of cognitive training on adolescent impul-
during adolescence (Khurana et al. 2018), the social contexts sivity. This is a ripe area for future research, as similar
of adolescents can play a significant role in determining which interventions with younger children have reported protec-
behaviors adolescents actually engage in to satisfy their ex- tive effects of cognitive training on self-control (Diamond
ploratory drives, as well as the health outcomes associated and Lee 2011). Other family-based interventions (Dishion
with these behaviors. and Stormshak 2007) and cognitive behavioral ap-
Adolescence is a period of rapid brain development, proaches (Heller et al. 2013) have also been used success-
providing a window of plasticity to train weaknesses in fully to reduce impulsivity in adolescents. Although these
impulse control. Several cognitive training interventions findings need to be replicated, these are promising strate-
have yielded promising results in improving executive gies to support adolescents who may be more vulnerable
functions related to top-down control over behavior to the detrimental effects of media.
Prev Sci

Fig. 3 a, b Interaction effects of

media (TV and movies) exposure
and sensation seeking (SS) on
violence

Parental monitoring of media use and content may be another Limitations

strategy to protect youth high in impulsivity and SS from being
exposed to risky content in the media (Collier et al. 2016). Our findings should be interpreted in light of the following lim-
Perceived parental monitoring, which is an indicator of the fam- itations. First, cross-sectional data is not ideal for testing media-
ily climate and parental involvement, operated as a significant tional processes that hypothesize specific temporal sequence of
protective influence in all our models. Parents could play an events (Maxwell and Cole 2007). Future research should exam-
important role by communicating with their child about media ine temporality of these relations and explore reciprocal effect
content and de-glamorizing it, so that youth high in SS and/or models. Previous studies have found that SS and impulsivity can
impulsivity are less likely to enact these behaviors. Policy-level predict exposure to risky media, and not the reverse, with the
changes can also help decrease risky media exposure for youth exception of videogaming literature where there is some evi-
by having stricter and clearer guidelines. For example, a change dence of excessive exposure to violence resulting in more impul-
in ratings systems (from age-based to content-based) could en- sive behaviors (Bushman and Huesmann 2006). Second, we
able parents in reducing exposure to risky media. only examined effects of TV and movie risk exposure; other
 Prev Sci

forms of interactive media, including videogames, internet, and References

social media, could not be explored. Their appeal to youth high
in impulsivity and SS as well as their effects on risk behaviors Allen, R. L., Dawson, M. C., & Brown, R. E. (1989). A schema-based
approach to modeling an African-American racial belief system.
can vary as compared to TVand movie risk exposure. Relatedly,
American Political Science Review, 83(2), 421–441. https://doi.
the high correlations between the media risk exposure variables org/10.2307/1962398.
(e.g., TValcohol exposure and TV sex exposure) made it difficult Anderson, C. A., & Bushman, B. J. (2001). Effects of violent video
to isolate the unique association of content-specific media risk games on aggressive behavior, aggressive cognition, aggressive af-
exposure and the corresponding behavioral outcome. Third, our fect, physiological arousal, and prosocial behavior: A meta-analytic
review of the scientific literature. Psychological Science, 12(5),
findings are based on a convenience sample of 14–17 years old 353–359. https://doi.org/10.1111/1467-9280.00366.
U.S. adolescents who were recruited online and cannot be gen- Balsa, A. I., Homer, J. F., French, M. T., & Norton, E. C. (2011). Alcohol
eralized to other populations. Finally, our findings based on self- use and popularity: Social payoffs from conforming to peers’ behav-
report data are susceptible to issues of recall bias and single ior. Journal of Research on Adolescence, 21(3), 559–568. https://
doi.org/10.1111/j.1532-7795.2010.00704.x.
informant bias.
Bjork, J. M., & Pardini, D. A. (2015). Who are those Brisk-taking
Our results provide support for both uses and gratifications adolescents^? Individual differences in developmental neuroimag-
paradigm and differential susceptibility hypothesis (for alco- ing research. Developmental Cognitive Neuroscience, 11, 56–64.
hol and violence). High impulsivity and SS can increase risk https://doi.org/10.1016/j.dcn.2014.07.008.
for problem behaviors through greater exposure to risky me- Bleakley, A., Ellithorpe, M. E., Hennessy, M., Khurana, A., Jamieson, P.,
& Weitz, I. (2017). Alcohol, sex, and screens: Modeling media
dia, and by operating as moderators, although the influence of influence on adolescent alcohol and sex co-occurrence. The
SS on some forms of risk-taking (e.g., experimentation with Journal of Sex Research, 0(0), 1–12. https://doi.org/10.1080/
alcohol use) seemed to be stronger for White youth than Black 00224499.2017.1279585.
youth. Understanding these processes longitudinally and in Bleakley, A., Fishbein, M., Hennessy, M., Jordan, A., Chernin, A., &
Stevens, R. (2008a). Developing respondent-based multi-media
the context of environmental influences (e.g., parental and
measures of exposure to sexual content. Communication Methods
peer influences) will be an important direction for future re- and Measures, 2(1–2), 43–64. https://doi.org/10.1080/
search. Impulsivity and SS are important characteristics to 19312450802063040.
consider when investigating how exposure to specific media Bleakley, A., Hennessy, M., & Fishbein, M. (2011a). A model of adoles-
content affects corresponding behavior, and they offer a pos- cents’ seeking of sexual content in their media choices. The Journal
of Sex Research, 48(4), 309–315. https://doi.org/10.1080/00224499.
sible point of intervention. For instance, by employing brief 2010.497985.
screeners for impulsivity and SS, health practitioners and ed- Bleakley, A., Hennessy, M., Fishbein, M., & Jordan, A. (2008b). It works
ucators could identify youth who may be at greater risk for both ways: The relationship between exposure to sexual content in
negative outcomes associated with media risk exposure. This the media and adolescent sexual behavior. Media Psychology, 11(4),
443–461. https://doi.org/10.1080/15213260802491986.
type of preventive approach would facilitate early identifica-
Bleakley, A., Hennessy, M., Fishbein, M., & Jordan, A. (2011b). Using
tion and intervention support for youth most in need. the integrative model to explain how exposure to sexual media con-
tent influences adolescent sexual behavior. Health Education &
Funding This study was funded by the National Institute of Child Health B e h a v i o r, 3 8 ( 5 ) , 5 3 0 – 5 4 0 . h t t p s : / / d o i . o r g / 1 0 . 11 7 7 /
and Human Development (NICHD) (Grant Number 1R21HD079615). 1090198110385775.
Its contents are solely the responsibility of the authors and do not neces- Bleakley, A., Jamieson, P. E., & Romer, D. (2012). Trends of sexual and
sarily represent the official views of the NICHD. violent content by gender in top-grossing U.S. films, 1950–2006.
Journal of Adolescent Health, 51(1), 73–79. https://doi.org/10.1016/
Compliance with Ethical Standards j.jadohealth.2012.02.006.
Bleakley, A., Romer, D., & Jamieson, P. E. (2014). Violent film charac-
ters’ portrayal of alcohol, sex, and tobacco-related behaviors.
Conflict of Interest The authors declare that they have no conflict of
Pediatrics, 133(1), 71–77. https://doi.org/10.1542/peds.2013-1922.
interest.
Blum, R. W., Beuhring, T., Shew, M. L., Bearinger, L. H., Sieving, R. E.,
& Resnick, M. D. (2000). The effects of race/ethnicity, income, and
Ethical Approval All procedures involving human participants were in family structure on adolescent risk behaviors. American Journal of
accordance with the ethical standards of the institutional and/or national Public Health, 90(12), 1879–1884.
research committee and with the 1964 Helsinki declaration and its later
Brown, J. D., & Bobkowski, P. S. (2011). Older and newer media:
amendments or comparable ethical standards. All procedures were ap-
Patterns of use and effects on adolescents’ health and well-being.
proved by the University of Pennsylvania Institutional Review Board.
Journal of Research on Adolescence, 21(1), 95–113. https://doi.org/
This article does not contain any studies with animals performed by any
10.1111/j.1532-7795.2010.00717.x.
of the authors.
Brown, J. D., & L’Engle, K. L. (2009). X-rated: Sexual attitudes and
behaviors associated with U.S. early adolescents’ exposure to sexu-
Informed Consent Informed consent was obtained from all participants ally explicit media. Communication Research, 36(1), 129–151.
included in the study. https://doi.org/10.1177/0093650208326465.
Brown, J. D., L’Engle, K. L., Pardun, C. J., Guo, G., Kenneavy, K., &
Publisher’s Note Springer Nature remains neutral with regard to jurisdic- Jackson, C. (2006). Sexy media matter: Exposure to sexual content
tional claims in published maps and institutional affiliations. in music, movies, television, and magazines predicts black and
Prev Sci

white adolescents’ sexual behavior. Pediatrics, 117(4), 1018–1027. Hennessy, M., Bleakley, A., Fishbein, M., & Jordan, A. (2009).
https://doi.org/10.1542/peds.2005-1406. Estimating the longitudinal association between adolescent sexual
Bushman, B. J., & Huesmann, L. R. (2006). Short-term and long-term behavior and exposure to sexual media content. The Journal of Sex
effects of violent media on aggression in children and adults. R e s e a rc h , 4 6 ( 6 ) , 5 8 6 – 5 9 6 . h t t p s : / / d o i . o r g / 1 0 . 1 0 8 0 /
Archives of Pediatrics & Adolescent Medicine, 160(4), 348–352. 00224490902898736.
https://doi.org/10.1001/archpedi.160.4.348. Houben, K., Wiers, R. W., & Jansen, A. (2011). Getting a grip on drinking
Centers for Disease Control and Prevention (2015). 1991–2015 high behavior training working memory to reduce alcohol abuse.
school youth risk behavior survey data. Psychological Science, 22(7), 968–975. https://doi.org/10.1177/
Chory, R. M., & Goodboy, A. K. (2010). Is basic personality related to 0956797611412392.
violent and non-violent video game play and preferences? Hoyle, R. H., Stephenson, M. T., Palmgreen, P., Lorch, E. P., & Donohew,
Cyberpsychology, Behavior and Social Networking, 14(4), 191– R. L. (2002). Reliability and validity of a brief measure of sensation
198. https://doi.org/10.1089/cyber.2010.0076. seeking. Personality and Individual Differences, 32(3), 401–414.
Collado, A., Felton, J. W., MacPherson, L., & Lejuez, C. W. (2014). https://doi.org/10.1016/S0191-8869(01)00032-0.
Longitudinal trajectories of sensation seeking, risk taking propensi- Iacono, W. G., Malone, S. M., & McGue, M. (2008). Behavioral disin-
ty, and impulsivity across early to middle adolescence. Addictive hibition and the development of early-onset addiction: Common and
Behaviors, 39(11), 1580–1588. https://doi.org/10.1016/j.addbeh. specific influences. Annual Review of Clinical Psychology, 4(1),
2014.01.024. 325–348. https://doi.org/10.1146/annurev.clinpsy.4.022007.
Collier, K. M., Coyne, S. M., Rasmussen, E. E., Hawkins, A. J., Padilla- 141157.
Walker, L. M., Erickson, S. E., & Memmott-Elison, M. K. (2016). Jaccard, J., & Turrisi, R. (2003). Interaction effects in multiple regression.
Does parental mediation of media influence child outcomes? A Thousand Oaks, CA: Sage.
meta-analysis on media time, aggression, substance use, and sexual Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Shah, P. (2011). Short- and
behavior. Developmental Psychology, 52(5), 798–812. https://doi. long-term benefits of cognitive training. Proceedings of the National
org/10.1037/dev0000108. Academy of Sciences. https://doi.org/10.1073/pnas.1103228108.
Collins, R. L., Elliott, M. N., Berry, S. H., Kanouse, D. E., Kunkel, D., Jamieson, P., & Romer, D. (2008). The changing portrayal of adolescents
Hunter, S. B., & Miu, A. (2004). Watching sex on television predicts in the media since 1950. Oxford University Press.
adolescent initiation of sexual behavior. Pediatrics, 114(3), e280–
Katz, E., Blumler, J., & Gurevitch, M. (1974). Utilization of mass com-
e289. https://doi.org/10.1542/peds.2003-1065-L.
munication by the individual. In J. Blumler & E. Katz (Eds.), The
Diamond, A., & Lee, K. (2011). Interventions shown to aid executive
uses of mass communication: Current perspectives on gratifications
function development in children 4 to 12 years old. Science,
research (pp. 19–32). Beverly Hills: Sage.
333(6045), 959–964. https://doi.org/10.1126/science.1204529.
Kerr, M., Stattin, H., & Burk, W. J. (2010). A reinterpretation of parental
Dishion, T. J., & Stormshak, E. A. (2007). The ecological family inter-
monitoring in longitudinal perspective. Journal of Research on
vention and therapy model. In T. J. Dishion & E. A. Stormshak
Adolescence, 20(1), 39–64. https://doi.org/10.1111/j.1532-7795.
(Eds.), Intervening in children’s lives: An ecological, family-
2009.00623.x.
centered approach to mental health care (pp. 49–67). Washington,
DC, US: American Psychological Association Retrieved from http:// Khurana, A., Romer, D., Betancourt, L. M., Brodsky, N. L., Giannetta, J.
psycnet.apa.org/index.cfm?fa=search.displayrecord&uid=2006- M., & Hurt, H. (2015a). Experimentation versus progression in ad-
11734-004. olescent drug use: A test of an emerging neurobehavioral imbalance
Ellithorpe, M. E., & Bleakley, A. (2016). Wanting to see people like me? model. Development and Psychopathology, 27(3), 901–913. https://
Racial and gender diversity in popular adolescent television. Journal doi.org/10.1017/S0954579414000765.
of Youth and Adolescence, 45(7), 1426–1437. https://doi.org/10. Khurana, A., Romer, D., Betancourt, L. M., Brodsky, N. L., Giannetta, J.
1007/s10964-016-0415-4. M., & Hurt, H. (2015b). Stronger working memory reduces sexual
Eysenck, S. B. G., Easting, G., & Pearson, P. R. (1984). Age norms for risk taking in adolescents, even after controlling for parental influ-
impulsiveness, venturesomeness and empathy in children. ences. Child Development, 86(4), 1125–1141. https://doi.org/10.
Personality and Individual Differences, 5(3), 315–321. https://doi. 1111/cdev.12383.
org/10.1016/0191-8869(84)90070-9. Khurana, A., Romer, D., Betancourt, L. M., & Hurt, H. (2017). Working
Gibbons, F. X., Kingsbury, J. H., Wills, T. A., Finneran, S. D., Dal Cin, S., memory ability and early drug use progression as predictors of ad-
& Gerrard, M. (2016). Impulsivity moderates the effects of movie olescent substance use disorders. Addiction. https://doi.org/10.1111/
alcohol portrayals on adolescents’ willingness to drink. Psychology add.13792.
of Addictive Behaviors, 30(3), 325–334. https://doi.org/10.1037/ Khurana, A., Romer, D., Betancourt, L., & Hurt, H. (2018). Modeling
adb0000178. trajectories of sensation seeking and impulsivity dimensions from
Hanewinkel, R., & Sargent, J. D. (2009). Longitudinal study of exposure early to late adolescence: Universal trends or distinct sub-groups?
to entertainment media and alcohol use among German adolescents. Journal of Youth and Adolescence.
Pediatrics, 123(3), 989–995. https://doi.org/10.1542/peds.2008- Khurana, A., Romer, D., Hurt, H., Betancourt, L., Brodsky, N. L., &
1465. Giannetta, J. M. (2012). Early adolescent sexual debut: The mediat-
Hayes, A. F., & Krippendorff, K. (2007). Answering the call for a stan- ing role of working memory ability, sensation seeking and impul-
dard reliability measure for coding data. Communication Methods sivity. Developmental Psychology, 48(5), 1416–1428. https://doi.
an d M eas u re s, 1( 1 ) , 7 7 –8 9 . h t t p s : / / do i . org / 1 0. 10 8 0/ org/10.1037/a0027491.
19312450709336664. Kim, J. L., Collins, R. L., Kanouse, D. E., Elliott, M. N., Berry, S. H.,
Heller, S., Pollack, H. A., Ander, R., & Ludwig, J. (2013). Preventing Hunter, S. B., et al. (2006). Sexual readiness, household policies,
youth violence and dropout: A randomized field experiment and other predictors of adolescents’ exposure to sexual content in
(Working Paper No. 19014). National Bureau of Economic mainstream entertainment television. Media Psychology, 8(4), 449–
Research. Retrieved from http://www.nber.org/papers/w19014 471. https://doi.org/10.1207/s1532785xmep0804_6.
Hennessy, M., Bleakley, A., Fishbein, M., & Jordan, A. (2008). Krcmar, M., & Kean, L. G. (2005). Uses and gratifications of media
Validating an index of adolescent sexual behavior using psychoso- violence: Personality correlates of viewing and liking violent genres.
cial theory and social trait correlates. AIDS and Behavior, 12(2), Media Psychology, 7(4), 399–420. https://doi.org/10.1207/
321–331. https://doi.org/10.1007/s10461-007-9272-1. S1532785XMEP0704_5.
 Prev Sci

Kronenberger, W. G., Mathews, V. P., Dunn, D. W., Wang, Y., Wood, E. Quarterly, 28(1), 38–47. https://doi.org/10.1111/j.1471-6402.2004.
A., Giauque, A. L., et al. (2005). Media violence exposure and 00121.x.
executive functioning in aggressive and control adolescents. Sheridan, E. (2006). Conservative implications of the irrelevance of rac-
Journal of Clinical Psychology, 61(6), 725–737. https://doi.org/10. ism in contemporary African American cinema. Journal of Black
1002/jclp.20022. Studies, 37(2), 177–192.
Lenhart, A. (2015). Teens, social media & technology overview 2015. Slater, M. D. (2003). Alienation, aggression, and sensation seeking as
Washington, DC: Pew Research Center. Retrieved from http:// predictors of adolescent use of violent film, computer, and website
www.pewinternet.org/2015/04/09/teens-social-media-technology- content. Journal of Communication, 53(1), 105–121. https://doi.org/
2015/ 10.1111/j.1460-2466.2003.tb03008.x.
Maxwell, S. E., & Cole, D. A. (2007). Bias in cross-sectional analyses of Slater, M. D., Henry, K. L., Swaim, R. C., & Cardador, J. M. (2004).
longitudinal mediation. Psychological Methods, 12(1), 23–44. Vulnerable teens, vulnerable times: How sensation seeking, alien-
https://doi.org/10.1037/1082-989X.12.1.23. ation, and victimization moderate the violent media content–
Moffitt, T. E., Arseneault, L., Belsky, D., Dickson, N., Hancox, R. J., aggressiveness relation. Communication Research, 31(6), 642–
Harrington, H., et al. (2011). A gradient of childhood self-control 668. https://doi.org/10.1177/0093650204269265.
predicts health, wealth, and public safety. Proceedings of the Strasburger, V. C., Jordan, A. B., & Donnerstein, E. (2010). Health effects
National Academy of Sciences, 108(7), 2693–2698. https://doi.org/ of media on children and adolescents. Pediatrics, 125(4), 756–767.
10.1073/pnas.1010076108. https://doi.org/10.1542/peds.2009-2563.
Patton, J. H., Stanford, M. S., & Barratt, E. S. (1995). Factor structure of
Valkenburg, P. M., & Peter, J. (2013). The differential susceptibility to
the Barratt impulsiveness scale. Journal of Clinical Psychology,
media effects model. Journal of Communication, 63(2), 221–243.
51(6), 768–774.
https://doi.org/10.1111/jcom.12024.
Raine, A., Reynolds, C., Venables, P. H., & Mednick, S. A. (2002).
Stimulation seeking and intelligence: A prospective longitudinal van Schuur, W. H. (2003). Mokken scale analysis: Between the Guttman
study. Journal of Personality and Social Psychology, 82(4), 663– scale and parametric item response theory. Political Analysis, 11(2),
674. 139–163.
Rideout, V. (2016). Measuring time spent with media: The common sense Whiteside, S. P., & Lynam, D. R. (2001). The Five Factor Model and
census of media use by us 8- to 18-year-olds. Journal of Children impulsivity: Using a structural model of personality to understand
and Media, 10(1), 138–144. https://doi.org/10.1080/17482798. impulsivity. Personality and Individual Differences, 30, 669–689.
2016.1129808. https://doi.org/10.1016/S0191-8869(00)00064-7.
Rideout, V., Foehr, U. G., & Roberts, D. F. (2010). Generation M2: Wills, T. A., Gibbons, F. X., Sargent, J. D., Gerrard, M., Lee, H.-R., & Dal
Media in the lives of 8 to 18 year olds. Menlo Park, CA: Henry J. Cin, S. (2010). Good self-control moderates the effect of mass media
Kaiser Family Foundation. Retrieved from http://kff.org/other/ on adolescent tobacco and alcohol use: Tests with studies of children
event/generation-m2-media-in-the-lives-of/. and adolescents. Health Psychology, 29(5), 539–549. https://doi.
Rubin, A. M. (2009). Uses and gratifications: An evolving perspective of org/10.1037/a0020818.
media effects. In R. L. Nabi & M. B. Oliver (Eds.), The SAGE Wills, T. A., Sargent, J. D., Gibbons, F. X., Forbes, M., Tappe, M., Cin, S.
handbook of media processes and effects (pp. 147–160). Thousand D., & Knight, R. (2012). Movie alcohol exposure and adolescent
Oaks, CA: SAGE. alcohol use in a multiethnic sample in Hawaii. Alcoholism: Clinical
Schooler, D., Monique Ward, L., Merriwether, A., & Caruthers, A. & Experimental Research, 36.
(2004). Who’s that girl: Television’s role in the body image devel- Zuckerman, M. (2006). Sensation seeking and risky behavior.
opment of young white and black women. Psychology of Women Washington D.C: APA Publications.