Exposure to Degrading Versus Nondegrading Music Lyrics and Sexual Behavior

Among Youth
Steven C. Martino, Rebecca L. Collins, Marc N. Elliott, Amy Strachman, David E.
Kanouse and Sandra H. Berry
Pediatrics 2006;118;430-441
DOI: 10.1542/peds.2006-0131

This information is current as of March 13, 2007

The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://www.pediatrics.org/cgi/content/full/118/2/e430

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly

publication, it has been published continuously since 1948. PEDIATRICS is owned, published,
and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk
Grove Village, Illinois, 60007. Copyright © 2006 by the American Academy of Pediatrics. All
rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from www.pediatrics.org by on March 13, 2007

ARTICLE

Exposure to Degrading Versus Nondegrading Music

Lyrics and Sexual Behavior Among Youth
Steven C. Martino, PhDa, Rebecca L. Collins, PhDb, Marc N. Elliott, PhDb, Amy Strachman, MAc, David E. Kanouse, PhDb, Sandra H. Berry, MAa

aRAND, Pittsburgh, Pennsylvania; bRAND, Santa Monica, California; cUniversity of California, Los Angeles, California

The authors have indicated they have no financial relationships relevant to this article to disclose.

ABSTRACT
BACKGROUND. Early sexual activity is a significant problem in the United States. A
recent survey suggested that most sexually experienced teens wish they had
www.pediatrics.org/cgi/doi/10.1542/
waited longer to have intercourse; other data indicate that unplanned pregnancy peds.2006-0131
and sexually transmitted diseases are more common among those who begin doi:10.1542/peds.2006-0131
sexual activity earlier. Popular music may contribute to early sex. Music is an
Key Words
integral part of teens’ lives. The average youth listens to music 1.5 to 2.5 hours per adolescent sexual behavior, media effects,
day. Sexual themes are common in much of this music and range from romantic music lyrics, sexually explicit media
and playful to degrading and hostile. Although a previous longitudinal study has Abbreviation
STD—sexually transmitted disease
linked music video consumption and sexual risk behavior, no previous study has
Accepted for publication Apr 11, 2006
tested longitudinal associations between the content of music lyrics and subse­
Address correspondence to Steven C. Martino,
quent changes in sexual experience, such as intercourse initiation, nor has any PhD, RAND, 4570 5th Ave, Suite 600,
study explored whether exposure to different kinds of portrayals of sex has Pittsburgh, PA 15213. E-mail: martino@rand.
org.
different effects. PEDIATRICS (ISSN Numbers: Print, 0031-4005;
Online, 1098-4275). Copyright © 2006 by the
DESIGN AND PARTICIPANTS. We conducted a national longitudinal telephone survey of American Academy of Pediatrics
1461 adolescents. Participants were interviewed at baseline (T1), when they were
12 to 17 years old, and again 1 and 3 years later (T2 and T3). At all of the
interviews, participants reported their sexual experience and responded to mea­
sures of more than a dozen factors known to be associated with adolescent sexual
initiation. A total of 1242 participants reported on their sexual behavior at all 3
time points; a subsample of 938 were identified as virgins before music exposure
for certain analyses. Participants also indicated how frequently they listened to
each of more than a dozen musical artists representing a variety of musical genres.
Data on listening habits were combined with results of an analysis of the sexual
content of each artist’s songs to create measures of exposure to 2 kinds of sexual
content: degrading and nondegrading.
OUTCOME MEASURES. We measured initiation of intercourse and advancement in non­
coital sexual activity level over a 2-year period.
RESULTS. Multivariate regression analyses indicated that youth who listened to more
degrading sexual content at T2 were more likely to subsequently initiate inter­
course and to progress to more advanced levels of noncoital sexual activity, even

e430 MARTINO et al
Downloaded from www.pediatrics.org by on March 13, 2007
after controlling for 18 respondent characteristics that hour a day listening to music, and a quarter of them
might otherwise explain these relationships. In contrast, listen in excess of 3 hours per day. Some have argued
exposure to nondegrading sexual content was unrelated that, because popular music is such a large part of ado­
to changes in participants’ sexual behavior. lescents’ everyday experience, youth cannot be under­
stood without a serious consideration of how music fits
CONCLUSION. Listening to music with degrading sexual lyr­ into their lives.7 From music, adolescents gain informa­
ics is related to advances in a range of sexual activities tion about society, social and gender roles, and expected
among adolescents, whereas this does not seem to be behavior, and they use music to facilitate friendships and
true of other sexual lyrics. This result is consistent with social interactions and to help them create a personal
sexual-script theory and suggests that cultural messages identity.7–9 It is reasonable to expect, therefore, that the
about expected sexual behavior among males and fe­ messages conveyed in popular music have significant
males may underlie the effect. Reducing the amount of implications for adolescent socialization and behavior.
degrading sexual content in popular music or reducing References to relationships, romance, and sexual be­
young people’s exposure to music with this type of havior are commonplace in the music that is most pop­
content could help delay the onset of sexual behavior. ular with teens, with �40% of popular songs in the
mid-1990s containing such references.7 A recent analysis
of the content of television shows, movies, magazines,

B Y THE TIME they reach 12th grade, 3 of 5 young

people in the United States have engaged in sexual
intercourse.1 Sex is part of a healthy life, and developing
newspapers, and music popular among teens demon­
strated that sexual content is much more prevalent in
popular music lyrics than in any other medium.10 De­
an interest in sex is natural as younger teens undergo spite their sexual nature, some researchers have argued
hormonal and other physical changes and older teens that the words of popular songs matter very little, sug­
begin to take on young adult roles. Although there is no gesting that youth do not necessarily understand or even
clear age at which sexual activity becomes appropriate, it pay attention to lyrical content.11–13 Although listening to
is clear that for some youth, intercourse initiation hap­ music may often be only a secondary activity for many
pens too soon. A recent national survey suggested that youth, the sexual references in many popular songs may
two thirds of sexually experienced teens wish they had be difficult for them to ignore, because the language
waited longer to have intercourse.2 Moreover, early sex­ used to describe sex has become increasingly direct.7,14
ual initiation creates the potential for significant health Consider, for example, these lyrics by the rap music
risks. One in 5 sexually active high school seniors has artist Lil’ Kim: “When it comes to sex don’t test my skills,
had �4 sexual partners and more than one third failed �cause my head game will have you head over heels.
to use a condom the last time they had sexual inter- Guys wanna wife me and give me the ring. I’ll do it
course.1 Nearly 900 000 teenagers become pregnant in anywhere, anyhow; I’m down for anything.”15 The in­
the United States each year (1 of every 5 sexually active terest in sex expressed in these lyrics is unlikely to be lost
teen girls).3 An estimated 1 in 4 sexually active teens gets on many teens.
a sexually transmitted disease (STD) every year (�4 There is strong theoretical justification for the notion
million teen cases of STDs annually) and half of all new that listening to sexual lyrics may influence adolescents’
human immunodeficiency virus infections in the United sexual behavior. According to social cognitive theo-
States occur among young people.4 Unplanned pregnan­ ry,16–18 people learn how to perform new behaviors by
cies and STDs are more common among those who observing others and will imitate the behaviors they
begin sexual activity earlier,5 probably because some have observed insofar as those behaviors are perceived
youth begin activity before they are prepared for the to have functional value. This theory would predict that
responsibilities entailed or because they begin having sex listening to musicians sing about having sex with no
as part of relationships/circumstances that are less likely unfavorable consequences will lead teens to perceive
to foster responsible behavior. Although much is known this behavior as appropriate and desirable, thereby in­
about the predictors of intercourse initiation, there re­ creasing the likelihood that they will imitate the behav­
main wide gaps in our knowledge and in our ability to ior. The likelihood of imitation increases when the
predict other sexual behavior, including sexual risk be­ model is perceived as attractive or similar to the self.16
havior. Highly popular music artists can, therefore, serve as
There is good reason to believe that music may have especially potent role models for teens.
an important influence on adolescents’ sexual behavior. Some versions of social learning theory argue that
On average, American youth listen to music from 1.5 to what people learn from media role models are
2.5 hours a day, which does not include the amount of “scripts.”19–21 A script typically includes information
time they are exposed to music via music videos.6 Music about what events are likely to occur in a specific sce­
media grow in importance as youth become older. Sixty nario, how a person should behave in that scenario, and
percent of teens aged 15 to 18 years report spending �1 what the likely outcomes of their behavior will be. Sex-

PEDIATRICS Volume 118, Number 2, August 2006 e431

Downloaded from www.pediatrics.org by on March 13, 2007
ual scripts establish norms and expectations about how portrayals of men as sex driven and women as sexual
and with whom to be sexual, why and when to have objects are unknown, endorsement of these types of
sex, and what the appropriate setting and sequence of beliefs is associated with an earlier age of first sexual
events are. Through its dominant themes and character­ activity and with more and riskier sexual experiences
izations, music may provide young people insights into among both males and females.34–36 Thus, portrayals of
particular sexual scripts, shaping their attitudes and as­ sexuality that objectify and degrade either gender may
sumptions about sexual relationships and creating a no­ affect sexual decision-making and behavior differently
tion of what is expected and normative.22 These atti­ from portrayals that do not.
tudes, expectations, and norms may, in turn, play a role
in adolescents’ sexual decision-making and behavior.
The only study to examine the association between Current Study
exposure to sexual content in music and adolescents’ This study tests the prospective relationship between
sexual intentions and activity was reported recently by exposure to degrading and nondegrading sexual music
Pardun et al.10 In this study, Pardun et al10 analyzed the content and subsequent changes in adolescent sexual
content of popular music lyrics for references to sexual behavior. We collected longitudinal data on adolescents’
development, romantic relationships, and sexual behav­ media use and sexual behavior, conducting 3 surveys.
ior and found that the more teens listened to music that Music listening was assessed in the second of these 3
contained such references, the more likely they were to surveys and used to predict initiation of sexual activities
be sexually active and to anticipate future sexual activ­ in the following 2 years among participants, taking into
ity. In that study, higher levels of exposure to sexual account sexual experiences before the second survey.
music content might have led to increased sexual activ­ This design allowed for the possibility that sexual expe­
ity and future intentions to be sexually active; however, rience influences teens’ selection of sexual music con­
a plausible alternative interpretation is that teens who tent and accounted for that possibility in testing for the
were sexually active or anticipating sexual activity in the effects of music on sexual behavior. The design also
near future chose to listen to more sexual music than allowed us to take into account baseline characteristics
their sexually inactive peers.23–25 This alternative expla­ that might contribute both to sexual behavior and ex­
nation could not be ruled out, because the researchers posure to sexual content in music, including several
did not know the sequence of these events. individual and family characteristics. By controlling for
Another limitation of the study by Pardun et al10 is these possible confounding variables, we were able to
that it did not consider the specific nature of the sexual more confidently attribute any differences in teens’ sex­
content. Because media depictions of sexuality contain a ual behavior to their exposure to sexual content in mu­
diversity of messages about sex that may affect teens’ sic.
sexual decision-making and behavior differently (per­ Because media researchers have often stressed the
haps through the communication of distinct sexual importance of investigating subgroup differences in the
scripts), to treat all sexual content the same may be to effects of media,25,37 we tested for gender and racial
overlook potentially important distinctions.
group differences in the association between exposure to
One depiction of courtship and sexual relationships
sexual music lyrics and teens’ sexual behavior. Music
that is common in youth-oriented media (and that pro­
preferences and interpretation of musical content differ
vides a very specific sexual script) features sex-driven
by gender and race,6,38–40 as do some effects of media use
males competing with one another for females who are
on teen’s sexual attitudes and behaviors.33,37,41,42 Con­
viewed as sexual objects or conquests whose value is
ducting gender group comparisons may be especially
based on their physical appearance.21,26 This portrayal of
important when examining the effects of degrading sex­
men as sexually insatiable and women as sexual objects
ual content, because this content is typically gender
is particularly prevalent in music videos,27–29 the frequent
specific in its messages about expected sexual behavior.
viewing of which is associated with stronger endorse­
ment of women as sexual objects30 and more traditional
gender role attitudes.30–32 Media researchers have spec­ METHODS
ulated that these types of portrayals may promote sexual
self-objectification (distancing oneself from one’s own Procedure
desires and actions), thereby promoting early sexual ac­ We conducted a national telephone survey in spring
tivity as objectified youth ignore their anxieties sur­ 2001 (T1) and reinterviewed the same group 1 and 3
rounding sexual initiation.30,33 Repeated exposure to years later, in the springs of 2002 and 2004 (T2 and T3,
these portrayals may also lead to the internalization of respectively). The survey measured media use, sexual
perceived gender norms that prescribe sexually degrad­ knowledge, attitudes, and behavior and a large set of
ing behavior as central to male and female gender roles. demographic and psychosocial variables known to pre­
Although the behavioral effects of exposure to media dict sexual behavior or media use.

e432 MARTINO et al
Downloaded from www.pediatrics.org by on March 13, 2007
Sample Recruitment attrition weights, which were combined with the final
Our sample for this study was recruited from a pur­ baseline weights to produce longitudinal weights. All of
chased national list of households with a high probability the analyses used these weights, appropriately account­
of containing a member aged 12 to 17 years. This list was ing for their effects on SEs.43
based on residential telephone listings, supplemented After applying these weights, 47% of T3 respondents
with other sources of information. The sample frame were girls, 68% were white, 14% were black, 12% were
was stratified by census tract race/ethnicity to produce Hispanic, and 6% either identified themselves as having
nationally representative proportions of minority and other racial/ethnic backgrounds or did not provide in­
non-Hispanic white youth. We mailed parents in these formation about race/ethnicity (�1% of cases missing
households an explanation of the study in advance and race/ethnicity). At least 1 parent had a college degree for
obtained verbal consent via telephone from a parent or 33% of the longitudinal sample; 59% had a parent who
legal guardian just before conducting an interview with had been otherwise educated beyond high school.
a randomly selected teen from the household. Teens
provided verbal assent. Our refusal rate at baseline was Measures
36%, similar to other studies of less sensitive topics. The
vast majority of those adults who refused consent cited Sexual Behavior
time constraints rather than a concern with the sexual Questions assessed behavior with someone of the oppo­
content of the survey. site sex. Intercourse experience at each time point was
measured with the item, “Have you ever had sex with a
Baseline Sample Characteristics and Weights boy/girl? By sex we mean when a boy puts his penis in
Without weights, the baseline sample of 2003 teens had a girl’s vagina” (yes/no). At T2, those with intercourse
demographic characteristics similar to those of all teens experience also reported the month and year of their
in the United States but included somewhat fewer His­ first intercourse experience. We used this date to deter­
panics and youth with highly educated parents. A mul­ mine the relevant analysis sample: youth who had not
tivariate logistic regression predicting nonresponse at had sex by September 2001 (the beginning of the refer­
baseline from information provided by the supplier of ence period for music listening; see below). At all 3 of the
our sample and a brief nonresponse interview with par­ interviews, we also measured lifetime levels of noncoital
ents identified higher response rates: (1) in census tracts experience with a scale developed for this study,44 based
with higher proportions of blacks, (2) among house­ on a measure used by Miller et al.45 Adolescents indi­
holds where a teen aged 12 to 14 years was present but cated whether they had ever (1) kissed, (2) “made-out
not randomly selected, and (3) when girls of any age or (kissed for a long time),” (3) touched a breast/had their
boys aged �14 years were randomly selected for sam­ breast touched, (4) touched genitals/had their genitals
pling. We created nonresponse weights inversely pro­ touched, and (5) given oral sex or received oral sex.
portional to the probability of enrollment indicated by Items 1 and 3 were asked of all youth; others were asked
this regression equation. After applying these weights, only if the response was “yes” to the item listed imme­
there were still small departures from the 1999 Current diately before it. Participants received a score of 1 to 5
Population Survey, which we corrected with poststrati­ reflecting the highest level of noncoital behavior expe­
fication weights. These nonresponse and poststratifica­ rienced; adolescents who reported none of the noncoital
tion weights were combined to form final baseline behaviors were included in the lowest category, along
weights. with those who had only kissed.

Sample Attrition and Longitudinal Weights Exposure to Sexual Content in Music

At T2, attrition was 12%. Extensive multivariate mod­ At T2, participants indicated how often (never, some­
eling with rich baseline response data found no evidence times, or a lot) they listened to each of 16 music artists
of selective attrition.41 Seventy-three percent of the on CDs, tapes, over the Internet, or on the radio “since
baseline sample (N � 1461) participated in the T3 sur­ school started last fall.” These 16 artists represented a
vey. Multivariate logistic regression modeling of attrition variety of music genres (eg, teen pop, rap, and alterna­
from baseline to T3 revealed some selective attrition. tive rock) and included male and female performers.
Overall, attrition was higher among all races for teens They were chosen from lists of top billboard artists. We
�14 at baseline, boys, and those whose parents had selected those who were featured in teen magazines
greater educational attainment. Among blacks, attrition around the time of the survey and/or those who partic­
was also higher among those with the least sexual ac­ ipated in teen-oriented music and entertainment award
tivity at baseline and was lower among those who, as of shows near the time of survey administration. We cre­
baseline, had not engaged in intercourse but had en­ ated measures of exposure to sexual content in music by
gaged in genital noncoital sexual activity. Results from linking information from our survey to information
this modeling were used to generate inverse-probability about the sexual content of each artist’s most recently

PEDIATRICS Volume 118, Number 2, August 2006 e433

Downloaded from www.pediatrics.org by on March 13, 2007
released album (before our T2 survey). Two raters inde­ proportion of songs with any sexual references). We
pendently coded the lyrics, obtained from Internet Web derived the 2 exposure measures (exposure to degrading
sites, of all songs (N � 193) from each of the 16 albums. sexual content in music and exposure to nondegrading
The unit of analysis was the song. Raters first judged sexual content in music) by weighting each content
whether a song contained 1 or more references to sexual score by self-reported frequency of listening to each
behavior (implicit or explicit references to intercourse, artist and summing across all 16 of the artists. Both
oral sex, or other sexual acts). For each song deemed to exposure measures were standardized to a mean of 0
contain �1 sexual reference, raters then judged whether and an SD of 1.
the song contained only nondegrading references to sex
or contained �1 degrading sexual reference. Thus, these Total Time Spent Listening to Music
classifications of content were mutually exclusive, and At T2, participants reported the total amount of time
the degrading/nondegrading designation accounted for they spend listening to music on CDs, tapes, over the
all of the instances of sexual content. Examples of non­ Internet, or on the radio during a typical week (0 � no
degrading sexual lyrics from our study include: “When time to 5 � �10 hours). Because of this statistical con­
my eyes open I wanna see your face/Spendin’ my days trol, our sexual content variables should be viewed as
in your sweet embrace/Just one night with you could set reflecting the effect of a high proportion of sexual con­
me free/I get next to you and I get dizzy, dizzy/You make tent relative to other content in one’s musical diet, re­
me think of things to come/I’m dreamin’ day and night gardless of the total amount of listening.
of making love,” from Ninety-Eight Degrees, “Dizzy.”46
Examples of degrading sexual lyrics include: “Half the Covariates
ho’s hate me, half them love me/The ones that hate All of the covariates were measured as part of the base­
me/Only hate me ‘cause they ain’t fucked me/And they line interview. Gender and race/ethnicity (dummy
say I’m lucky/Do you think I’ve got time/To fuck all coded as white versus black, Hispanic, or other) were
these ho’s?” from Ja Rule, “Livin’ It Up.”47 self-reported. Respondent age, calculated from date of
For each song containing a sexual reference, raters birth, and baseline interview date was measured contin­
also judged whether the reference was implicit or ex­ uously in years.
plicit and whether the reference was about casual or We included several indicators of social environment
committed sex. These dimensions strongly overlapped known to predict initiation of coitus. Teens who re­
with the degrading/nondegrading distinction. Of songs ported living with both of their parents were contrasted
with degrading sexual lyrics, 71% were judged to con­ with all others. Parent education was measured as
tain explicit references to sex, and 96% were judged to schooling completed by the more highly educated parent
be about casual sex. In contrast, 51% of songs with (1 � less than high school to 6 � completed graduate or
nondegrading sexual lyrics contained explicit references professional degree). Parental monitoring was measured
to sex and only 19% were judged to be about casual sex. with a 5-item scale (eg, “When you are away from
In our predictive analyses, we chose to focus on the home, your parents know where you are and who you
degrading/nondegrading dimension because it was of are with”; items rated from 1 � strongly agree to 5 �
central theoretical concern and because it seemed to strongly disagree) developed to predict adolescent risk
encompass all 3 characteristics (ie, the definition of de­ behavior (� � .70).48 An additional measure tapped
grading sexual lyrics strongly implies that they also be parental prescriptive norms by asking perceived parental
explicit and address casual sex). response if the respondent had sex in the following year
To establish interrater reliability for classifying the (1 � disapprove a lot to 5 � approve a lot). Because
type of content in a song, raters double-coded one third responses were bimodal, we recoded the item to dichot­
(n � 63) of all of the songs. These songs were selected via omously reflect parents’ disapproval (responses of 1 or
stratified random sampling, with artists as strata. Inter­ 2) versus approval or neutrality (responses of 3, 4, or 5).
rater reliability was satisfactory (Cohen’s � ranged from A single item assessed whether the respondent’s friends
0.74 to 0.92). were primarily older, younger, or about the respon-
For each artist studied, the proportion of songs con­ dent’s age and was dichotomized to indicate “older”
taining degrading sexual lyrics was calculated by divid­ versus all other responses. We measured perceived
ing the number of the artist’s songs that contained any friends’ approval of sex by asking respondents, “How
degrading sexual references by the total number of songs would your friends feel if you had sexual intercourse in
on the artist’s album, regardless of their sexual content. the next year (1 � disapprove a lot to 5 � approve a
Similarly, the proportion of songs containing nonde­ lot)?”
grading sexual lyrics was calculated by dividing the We also assessed a number of other personal charac­
number of the artist’s songs that contained only nonde­ teristics that are known correlates of adolescent sexual
grading sexual references by the total number of songs behavior. Mental health (� � .68) was assessed with the
on the artist’s album (these 2 indices sum to indicate the Mental Health Inventory, a well-validated 5-item scale

e434 MARTINO et al
Downloaded from www.pediatrics.org by on March 13, 2007
tapping affective state over the previous 4 weeks.49 Re­ these analyses (analogous to what we did with the anal­
spondents’ self-reported grades in school at baseline yses predicting intercourse initiation). We did, however,
were used as an indicator of academic performance (1 � control for the level of noncoital behavior at the time of
mostly As to 5 � mostly Fs). Deviant behavior was the baseline survey.
measured with 6 items drawn from previous studies of We began by testing simple associations between our
adolescent risk behavior (� � .65 in this sample).50 Par­ T2 music variables and intercourse initiation by T3 and
ticipants indicated how many times in the past 12 advancement in the level of noncoital behavior between
months they had: been sent out of class; broken into a T1 and T3. Because teens who listen to more sexual
house, school, or place of business; skipped school; content in music also listen to more music overall, the
cheated on a test; damaged something on purpose; or music variables are best understood in the context of one
stolen something (1 � not at all, 4 � �10 times). To another. We, thus, examined all 3 of the music variables
measure religiosity, we asked participants to indicate on (exposure to degrading and nondegrading sexual con­
a 4-point scale their agreement with the statement, “Re­ tent and total amount of music listening) simultaneously
ligion is very important in my life.” Sensation seeking, a in these tests. We also tested whether other respondent
strong predictor of sexual initiation in a previous study characteristics might explain any relationship between
of this sample,41 was measured with 3 items from the listening to sexual content and sexual behavior by ex­
Zuckerman scale (� � .58).51 amining bivariate associations between these character­
Finally, we included several indicators of adolescent istics and exposure to sexual content in music at T2,
interest in sex or sexual readiness before music listening intercourse initiation by T3, and advancement in the
to control for the possibility that youth who are consid­ level of noncoital behavior between T1 and T3.
ering coital or noncoital activities that they have not yet We then tested multiple-group path models to look
enacted may listen to more sex-oriented music. These (separately) for gender and racial group differences in
indicators included baseline level of noncoital sexual the relationships between exposure to sexual content in
activity, intentions to have sex in the next year (1 � not music and sexual behavior while controlling for the
at all likely to 5 � extremely likely), expected negative effects of all of the other variables. We were limited to
consequences of having sex, and sex self-efficacy. Ex­ comparing whites and nonwhites in our models that
pected negative consequences (� � .60) were measured examined racial group differences, because our sample
with a 3-item scale (eg, “If you had sex, you would feel sizes for minority subgroups were too small to obtain
guilty afterward”) drawn from a previous study.42 Sex reliable estimates of group differences. To test for be-
self-efficacy was measured with a single item: “How tween-group differences, we first constrained each
likely is it that you would be able to talk with a boy (if model parameter separately to be equal across groups.
female respondent) or girl (if male respondent) about We then calculated the difference between �2 statistics
whether or not you should have sex?” This item was for this constrained model and an unconstrained model
drawn from a scale used in a previous study in which it that allowed for gender or racial group differences and
was shown to have the highest factor loading.42 tested each �2 difference for significance. A significant
effect indicates that the influence of the predictor differs
Analyses for boys/girls or whites/nonwhites. In examining be-
To be included in the current analyses, respondents had tween-group differences in the path coefficients involv­
to have completed interviews at each of the 3 survey ing the music variables, we constrained/freed all of the
waves (N � 1390). Respondents also had to have valid music parameters at once because of their conceptual
sexual behavior and music exposure data at all 3 of the interdependence.
time points. One hundred forty-eight respondents re­ All of the analyses were conducted in Mplus 3.12
quested that we skip questions about sexual behavior using maximum likelihood for parameter estimation.43
(an option presented during the interview) at �1 of the To be certain that our estimates were robust to violations
3 surveys. These respondents had missing data for sexual of the assumption of multivariate normality, we esti­
behavior. To control for sexual behavior before music mated SEs using a sandwich estimator and tested the
exposure, the analysis sample for predicting intercourse significance of coefficients with the Yuan-Bentler T2*
was restricted to those youth who had never had inter­ test statistic.52
course (“virgins”) before the reference period for the
music items (September 2001; see above). Thus, the Missing Data Imputation
final sample for these analyses was 938. Tests predicting Ten covariates were missing in small numbers of cases
noncoital behavior included all of the respondents who among those retained through the third wave (0.5%–
provided data on sexual behavior at all 3 of the survey 3.0% for 4 variables and �0.5% for 6 variables). To
waves (N � 1242). Because we did not have dates for avoid bias that listwise deletion of cases with missing
noncoital behavior, we were not able to control for level data might introduce in our results, we imputed missing
of noncoital behavior just before music exposure in data on these predictors.53 Imputation was based on

PEDIATRICS Volume 118, Number 2, August 2006 e435

Downloaded from www.pediatrics.org by on March 13, 2007
random draws corresponding to model-based predicted exposure to music in general was also related to teens’
probabilities. sexual behavior. Teens who spent more time listening to
music were more likely than those who spent less to
RESULTS initiate intercourse and to progress in their noncoital
In our weighted longitudinal sample, 17% had ever had activity.
intercourse at baseline, 29% at the first follow-up assess­ Simple associations between other respondent char­
ment, and 53% at the second follow-up assessment. acteristics and both intercourse initiation and advance­
Rates of intercourse observed in this sample are similar ment of noncoital behavior are also shown in Table 2. As
to those found in other national surveys with household can be seen, almost all of these factors were significantly
sampling frames (eg, the National Survey of Family related to advances in sexual behavior. Most were also
Growth54). associated with exposure to degrading and nondegrad­
ing sexual music content at T2, although more charac­
Content Analysis teristics were associated with degrading than with non­
Table 1 shows the musical genre of each of the 16 artists degrading music exposure. Some of these variables could,
whose lyrics we analyzed, the number of songs analyzed
therefore, account for the observed relationships between
per artist, the percentages of each artist’s songs contain­
exposure to sexual music content and later sexual behav­
ing any sexual references, and the percentage of each
ior. To test for music effects independent of these variables
artist’s songs containing sexually degrading lyrics. As the
and to increase the sensitivity of the analyses predicting
table shows, all but 3 artists had 1 or more songs that
sexual behavior, we entered all of these variables into our
made reference to sexual behavior. Sexual lyrics were
multivariate models as covariates.
found in music by artists of all of the genres but were
We first tested multiple-group models predicting in­
most prevalent in the songs of rap and rap metal artists.
The percentage of songs with sexual lyrics ranged from tercourse initiation and noncoital activity by gender and
0% to 71% across the 16 artists. The percentage of songs by race but found no differences in the associations
that contained sexually degrading lyrics ranged from 0% between music exposure and sexual advancement for
to 70% across artists, with the highest concentration of boys versus girls or whites versus nonwhites. Thus, we
degrading sexual lyrics observed among the songs of rap report results of a single-group multiple regression anal­
and rap metal artists. ysis (with race and gender included among the covari­
Before controlling for other variables, greater expo­ ates) in Table 3. As this table shows, the more teens
sure to music with degrading sexual content at T2 was listened to degrading sexual music content, the more
strongly and positively related to initiation of intercourse likely they were to subsequently initiate intercourse and
by T3 and advancement of noncoital activity between progress in their noncoital activity. These music effects
baseline and T3 (see Table 2). In contrast, exposure to held, even though 18 other predictors of sexual behavior
music with nondegrading sexual content was nega­ were taken into account. Exposure to nondegrading sex­
tively associated with intercourse initiation and unasso­ ual music was unrelated to changes in sexual behavior in
ciated with advancement of noncoital activity. Greater these multivariate models.

TABLE 1 Musical Genre, Number of Songs, and Sexual Content of Songs by Each Artist Included in the
Content Analysis
Artist Musical Genre Number of Songs % of Songs With % of Songs With Degrading
No. per Album Sexual Content Sexual Content
1 Hard rock 12 50 0
2 Alternative rock 15 33 0
3 Alternative rock 11 0 0
4 Rap-rock 14 64 43
5 Rap-rock 11 63 45
6 Rap 13 70 70
7 Rap 17 71 59
8 Rap 13 0 0
9 Rap-metal 14 21 14
10 R&B 12 42 17
11 Country 12 0 0
12 Country 13 8 0
13 Teen pop 12 33 0
14 Teen pop 13 23 0
15 Teen pop 12 42 0
16 Teen pop 13 23 0

e436 MARTINO et al
Downloaded from www.pediatrics.org by on March 13, 2007
 TABLE 2 Associations Between Candidate Predictors and Exposure to Sexual Lyrics and Sexual Behavior
Predictor Variable Exposure to Exposure to Intercourse Level of Noncoital
Degrading Sexual Nondegrading Initiation by Sex at Time 3
Lyrics at Time 2a Sexual Lyrics at Time 3 (Controlling for Time
(n � 1390) Time 2a (n � 1390) (n � 981) 1 Level) (n � 1242)
� P � P � P � P
Music exposure variables
Time 2 exposure to degrading sexual lyrics .36 �.001 .20 �.001
Time 2 exposure to nondegrading sexual lyrics �.14 .02 �.05 .13
Time 2 total time spent listening to music .21 �.001 .10 .002
Covariates
Age, y �.01 .84 .01 .79 .29 �.001 .04 .21
Female gender �.03 .44 .24 �.001 �.11 .02 �.10 .002
Black (vs non-Hispanic white) .24 �.001 �.05 .16 .10 .07 .01 .70
Hispanic (vs non-Hispanic white) .13 �.001 �.01 .91 .06 .36 �.02 .60
Other race (vs non-Hispanic white) .06 .05 .03 .44 .09 .05 .02 .51
Lives with both parents �.18 �.001 �.10 .02 �.11 .07 �.04 .19
High parent education �.18 �.001 �.05 .20 �.09 .10 �.06 .03
Time 1 parental monitoring �.07 .04 .07 .06 �.30 �.001 �.13 �.001
Time 1 perceived parent disapproval of sex �.10 .003 �.01 .86 �.06 .13 .03 .17
Time 1 has mostly older friends .04 .32 .02 .61 .12 .02 .00 .96
Time 1 perceived friend approval of sex .17 �.001 �.08 .02 .39 �.001 .14 �.001
Time 1 good mental health �.06 .07 �.08 .02 �.09 .07 �.02 .40
Time 1 low school grades .17 �.001 �.01 .73 .24 �.001 .08 .003
Time 1 deviant behavior .15 �.001 �.10 .004 .35 �.001 .11 �.001
Time 1 religiosity .03 .38 .12 �.001 �.14 .002 �.11 �.001
Time 1 sensation seeking .14 �.001 .04 .25 .26 �.001 .16 �.001
Time 1 level of noncoital activity .22 �.001b .04 .27b .47 �.001c .57 �.001
Time 1 intentions to have sex .26 �.001 .00 .98 .44 �.001 .12 .002
Time 1 expected negative consequences �.17 �.001 .09 .02 �.27 �.001 �.12 �.001
Time 1 sex self-efficacy .11 .003 .02 .52 .38 �.001 .17 �.001
a Tests for exposure to degrading sexual lyrics control for exposure to nondegrading sexual lyrics and total music exposure. Tests for nondegrading sexual lyrics control for exposure to degrading

sexual lyrics and total music exposure.

b Because of nonimputed missing values on time 1 noncoital activity, n for this analysis was 1298.

c Because of nonimputed missing values on time 1 noncoital activity, n for this analysis was 938.

DISCUSSION sexually degrading. In our coding scheme, lyrics classi­

To our knowledge, this is the first study to test longitu­ fied as degrading depicted sexually insatiable men pur­
dinal associations between exposure to sexual content in suing women valued only as sex objects. These types of
music lyrics and changes in sexual behavior. Our results portrayals objectify and degrade women in ways that are
provide evidence that such exposure is related to ad­ obvious but do the same to men by depicting them as
vances in a range of sexual activities among adolescents, sex-driven studs whose individual desires are subsumed
including intercourse and noncoital behavior. These as­ in their gender role. Adolescents who listen to a lot of
sociations were evident among male and female listeners music containing these objectifying and limiting charac­
and among whites and nonwhites and held even after terizations of sexuality progress more quickly in their
accounting for a wide variety of other personal and sexual behavior, regardless of their race or gender.
social factors correlated with adolescent sexual behavior
Our results suggest that the relationship between ex­
and with exposure to sexual lyrics. These findings con­
posure and behavior may be causal in nature, because
tribute to the emerging body of evidence of the role of a
we controlled for teens’ previous sexual experience, as
variety of media in the sexual socialization of youth.41,55
well as factors like parental monitoring, religiosity, and
Our study also identified an important limitation to
deviance; however, our correlational data do not allow
music effects. Exposure to sexual lyrics was not related
to changes in sexual behavior when those lyrics were us to make causal inferences with certainty. If the rela­
not sexually degrading. Although there was a significant tionship is causal, listening to music with degrading sex­
bivariate relationship between exposure to nondegrad­ ual content may have important public health and other
ing sexual lyrics and intercourse initiation, this relation­ societal consequences, because those who initiate sex
ship disappeared when covariates, including exposure to early have more STDs and unplanned pregnancies.5 It is
degrading lyrics, were controlled. This suggests quite important to point out, however, that at the time of the
strongly that the influence of sexual music content on third survey, about half of our sample had become legal
teens’ sexual development is specific to content that is adults (18 –20 years); initiation of intercourse in this

PEDIATRICS Volume 118, Number 2, August 2006 e437

Downloaded from www.pediatrics.org by on March 13, 2007
 TABLE 3 Multivariate Regression Analyses Predicting Intercourse Initiation and Advancement in
Noncoital Sexual Behavior
Predictor Variable Intercourse Initiation Level of Noncoital Sex
by Time 3 (n � 938) at Time 3 (n � 1242)
� P � P
Time 2 exposure to degrading sexual lyrics .23 �.001 .18 �.001
Time 2 exposure to nondegrading sexual lyrics �.06 .29 .00 .89
Time 2 total time spent listening to music .09 .04 .06 .05

Age, y .09 .08 .04 .27

Female gender .03 .59 �.06 .08
Black (vs non-Hispanic white) .03 .52 �.02 .57
Hispanic (vs non-Hispanic white) �.01 .86 �.05 .09
Other race (vs non-Hispanic white) .09 .04 .01 .77
Lives with both parents �.02 .64 �.01 .66
High parent education �.05 .27 �.05 .06
Time 1 parental monitoring �.04 .39 �.08 .005
Time 1 perceived parent disapproval of sex .04 .33 .05 .02
Time 1 has mostly older friends .07 .07 �.02 .59
Time 1 perceived friend approval of sex .12 .03 .04 .29
Time 1 good mental health .03 .57 .01 .60
Time 1 low school grades .12 .02 .02 .51
Time 1 deviant behavior .05 .32 .02 .45
Time 1 religiosity �.06 .14 �.09 �.001
Time 1 sensation seeking .03 .59 .09 .003
Time 1 level of noncoital activity .21 �.001 .36 �.001
Time 1 intentions to have sex .08 .17 �.02 .73
Time 1 expected negative consequences .02 .70 .00 .93
Time 1 sex self-efficacy .14 .01 .12 �.001

group would not be considered early according to US “boys being boys” and dismiss girls’ sexual preferences
norms and might be considered healthy. and desires as inconsequential. Our research is unable to
That the effect of sexual music content on adoles­ test for these effects but does suggest that degrading
cents’ sexual behavior is specific to degrading lyrics sug­ sexual lyrics do more than “go in one ear and out the
gests something about the process by which this effect other.”
occurs. Musicians who incorporate this type of sexual We have explained the effect of degrading sexual
imagery in their songs are not simply modeling an in­ lyrics as operating through the acquisition of a specific
terest in healthy sexual behavior for their listeners; they sexual script in which sexually aggressive men treat
are communicating something specific about what are women as valued only for the sexual pleasure that they
appropriate sexual roles for men and women. These can provide. There are, however, other characteristics of
lyrics are likely to promote acceptance of women as degrading sexual lyrics that may contribute to their ef­
sexual objects and men as pursuers of sexual con- fects. For example, degrading sexual lyrics are more
quest.21,33 Teens who are repeatedly exposed to and ac­ likely than nondegrading sexual lyrics to be explicit and
cept these messages may come to see these as appropri­ to focus on casual rather than committed sex. In the
ate characterizations for themselves and enact these future, researchers may want to distinguish the effects of
stereotyped gender roles in their sexual behavior. De­ these dimensions to determine whether multiple pro­
spite the fact that degrading sexual lyrics are particularly cesses operate to determine teen sexual behavior.
demeaning in their treatment of women, they affect Lyrical content may be only part of what drives the
adolescent boys and girls similarly. This could have wor­ associations found in this study. For example, teens who
risome implications for what both genders come to ex­ listen to music by artists who use degrading sexual im­
pect from their own sexual partners and experiences. It agery in their songs probably also watch music videos by
may be that girls who are repeatedly exposed to these these artists, in which case the effect of these songs is
messages expect to take a submissive role in their sexual likely to be greatly enhanced.31,32,56 Adding a visual por­
relationships and to be treated with disrespect by their trayal of sex may reinforce sexual lyrics not only by
partners. If so, these expectations may have lasting ef­ increasing the number of sexual cues in the message57
fects on their relationship choices, a possibility that war­ but also by aiding the interpretation of the meaning of a
rants further investigation. Boys, on the other hand, song and clarifying ambiguous lyrics.58 Other than lis­
may come to interpret reckless male sexual behavior as tening to music and watching videos by music artists,

e438 MARTINO et al
Downloaded from www.pediatrics.org by on March 13, 2007
teens may also read about them in music magazines, references to sexual behavior, as has been done in some
attend live performances, and otherwise expose them­ recent media content analyses.10,62 Future research
selves to the messages that these artists portray. The should undertake a more detailed analysis of music sex­
music and the artists form a consistent package that ual content to confirm that our findings are robust to this
many youth identify with strongly.7,23,59,60 Not all youth limitation. A related shortcoming is that we only asked
will identify with these artists, however, and not all participants about 16 musical artists. Although these
youth will interpret their lyrics in the same way. Inter­ artists were popular with teens at the time of our survey
pretation probably depends on many factors, such as and represent the musical genres that teens listen to
age, race, gender, social background, where youth are in most commonly,6 a measure that included a broader
their sexual development, and how involved they are in sample of artists, including non-English-speaking artists,
the music to which they listen.25,37 Future work is would likely be more sensitive and may have shown
needed to elucidate the meanings assigned to degrading stronger effects for music exposure than those reported
sexual lyrics by male and female listeners and how these here.
may explain the associations that we have observed. We were also limited by our inability to distinguish
We also observed an association between time spent among racial and ethnic minorities in our multiple­
listening to music in general and changes in sexual be­ group comparisons. Research has shown that patterns of
havior. The more time teens spent listening to music, the media use and interpretation of media content may dif­
more likely they were to advance in their noncoital fer across racial/ethnic subgroups. The vast majority of
sexual behavior and to initiate intercourse. This was true nonwhites in our sample were black and Hispanic youth.
although the sexual content of the music was controlled Both of these groups had significantly more exposure to
for statistically. It may be that listening to popular music, degrading sexual content than did whites, and neither
regardless of its content, results in heightened physio­ group’s exposure to nondegrading sexual content dif­
logic arousal that, through a process of excitation trans- fered from that of white youth. These findings are con­
fer,61 incites sexual behavior among teens. Alternatively, sistent with research showing that black and Hispanic
time spent listening to music may be a proxy for a youth listen to more rap music,7,63,64 the genre with the
covariate that we did not measure in our survey, such as highest concentration of degrading content in our anal­
use of leisure time. It is likely that youth who spend a ysis. Although black and Hispanic youths’ amounts of
great deal of time listening to music also spend more exposure to sexual content, relative to whites’, were
time at parties and clubs and less time working, study­ similar, their interpretation of this content may never­
ing, interacting with their families, or watching TV. This theless have differed. Future research should attempt to
difference may create greater opportunities to meet po­ discriminate among other minority groups to determine
tential sexual partners and/or pressure to engage in sex­ whether the effects observed in our study fail to hold
ual activity. A third alternative was alluded to earlier in among certain racial and ethnic groups.
our recommendation for caution regarding interpreta­ Finally, we were not able to fully control for previous
tions of the degrading music lyric effect. It may be some­ sexual experience when predicting advances in nonco­
thing about popular musical artists other than their lyrics ital behavior. Because we did not know the precise dates
that produces a correlation between music listening and on which teens advanced from one level of noncoital
sexual advancement. The artists may project sexual im­ behavior to the next, we were limited to controlling for
ages, lifestyles, or otherwise adult behavior that youth noncoital behavior at baseline, �6 months before the
identify with and emulate, and this in turn may result in reference period for music exposure. Because adoles­
progressing sexual activity. cents’ music choices are likely to be shaped, in part, by
This underscores an important limitation to our their level of sexual experience,23–25 we may have over­
study. Although we accounted for many individual and estimated the relationship between exposure to degrad­
environmental factors that might otherwise explain the ing sexual content and advances in noncoital behavior.
relationship between exposure to sexual music content Nonetheless, the analyses presented go beyond a simple
and adolescent sexual behavior, including sexual inter­ cross-sectional design, and our prediction of intercourse
est, readiness, and behavior before music exposure, it is is not affected by ambiguity in timing.
possible that we have overlooked some variable that These limitations notwithstanding, our findings sug­
may account for the association. A second limitation gest a need for intervention. Reducing the amount of
concerns the level of detail in our coding scheme. We degrading sexual content in popular music, or reducing
coded 2 general categories, sex and sexual degradation, young people’s exposure to music with this type of
that comprise a set of more specific attitudes and behav­ content, could delay initiation of intercourse and related
iors that we did not attempt to distinguish (eg, oral sex, activities. This, in turn, may reduce sexual risk behavior
anal sex, intercourse, and behavior toward committed and sexual regret. Intervention possibilities include
versus casual partners). We also coded for sexual content reaching out to parents of adolescents, to teens, and to
at the level of the song rather than coding for discrete the recording industry. Parents could be encouraged to

PEDIATRICS Volume 118, Number 2, August 2006 e439

Downloaded from www.pediatrics.org by on March 13, 2007
monitor the type of music to which their children are J, ed. Popular Music and Communication. Beverly Hills, CA: Sage;
exposed, set limits on what they can purchase and listen 1987:140 –174
10. Pardun CJ, L’Engle KL, Brown JD. Linking exposure to
to, and be careful not to listen to sexually degrading
outcomes: early adolescents’ consumption of sexual content in
music when children are around. Parents could also be six media. Mass Commun Soc. 2005;8:75–91
encouraged to discuss the sexual content of music with 11. Greenfield PM, Bruzzone L, Koyamatsu K, et al. What is rock
their children, offering their own perspectives on the music doing to the minds of our youth? A first experimental
sexual themes to which their children are exposed. look at the effects of rock music lyrics and music videos. J Early
Through media education, teens could be made aware of Adolesc. 1987;7:315–329
12. Leming JS. Rock music and the socialization of moral values in
the ways in which sex is depicted and perhaps distorted early adolescence. Youth Soc. 1987;18:363–383
in the music to which they are exposed and develop 13. Rosenbaum J, Prinsky L. Sex, violence, and rock ‘n’ roll:
skills for listening to and thinking about the sexual mes­ youth’s perceptions of popular music. Pop Music Soc. 1987;11:
sages of music in a more critical way. Finally, the record­ 79 – 89
ing industry could be made aware of the potential neg­ 14. Strausburger VC. Adolescents and the Media: Medical and Psycho-
logical Impact. Thousand Oaks, CA: Sage; 1995
ative impact of sexually degrading music. Additional
15. Lil’ Kim. Magic stick. On La Bella Mafia [CD]. New York, NY:
research, as well as feedback from stakeholders, is Atlantic Records; 2003
needed to determine which of these strategies is most 16. Bandura A. Social Foundations of Thought and Action: A Social
appropriate and likely to meet with success. Future re­ Cognitive Theory. Englewood Cliffs, NJ: Prentice Hall; 1986
search should also move beyond the examination of 17. Bushman BJ, Huesmann LR. Effects of televised violence on
intercourse initiation to investigate directly the psycho­ aggression. In: Singer D, Singer J, eds. Handbook of Children and
the Media. Thousand Oaks, CA: Sage; 2001:223–254
logical and public health consequences implied by early
18. Bandura A. Social cognitive theory of mass communication. In:
sexual initiation (eg, sexual regret, number of sexual Bryant J, Zillman D, eds. Media Effects: Advances in Theory and
partners, unplanned pregnancies, and sexually transmit­ Research. Hillsdale, NJ: Erlbaum; 1994:61–90
ted infections). Such research would provide important 19. McCormick NB. Sexual scripts: social and therapeutic implica­
evidence regarding the connection between adolescent tions. Sex Marital Ther. 1987;2:3–27
20. Unger R, Crawford M. Women and Gender: A Feminist Psychology.
sexual behavior and exposure to degrading sexual music
New York, NY: McGraw-Hill; 1992
content while also suggesting ways to mitigate adverse 21. Ward LM. Talking about sex: common themes about sexuality
outcomes. in prime-time television programs children and adolescents
view most. J Youth Adolesc. 1995;24:595– 615
ACKNOWLEDGMENT 22. Gerbner G, Gross L, Morgan M, Signorielli N. Living with
television: the dynamics of the cultivation process. In: Bryant J,
This research was supported by National Institute of
Zillman D, eds. Perspectives on Media Effects. Hillsdale, NJ: Law­
Child Health and Human Development grant R01-HD rence Erlbaum Associates; 1986:17– 40
38090 to Dr Collins. 23. Arnett JJ. Adolescents’ uses of media for self-socialization. J
Youth Adolesc. 1995;24:519 –533
REFERENCES 24. Brown JD, Halpern CT, L’Engle KL. Mass media as a sexual
super peer for early maturing girls. J Adolesc Health. 2005;36:
1. Centers for Disease Control and Prevention. Youth risk behav­
420 – 427
ior surveillance - United States, 2003. MMWR Morb Mortal Wkly
Rep. 2004;53:SS-2. 25. Steele J, Brown JD. Adolescent room culture: studying media
2. National Campaign to Prevent Teen Pregnancy. With One Voice in the context of everyday life. J Youth Adolesc. 1995;24:
2003: America’s Adults and Teens Sound Off about Teen Pregnancy. 551–576
Washington, DC: National Campaign to Prevent Teen 26. Gow J. Reconsidering gender roles on MTV: depictions in the
Pregnancy; 2003 most popular music videos of the early 1990s. Commun Rep.
3. Henshaw SK. U.S. Teenage Pregnancy Statistics with Comparative 1995;9:151–161
Statistics for Women Aged 20 –24. New York, NY: Alan Guttma­ 27. Andsager JL, Roe K. Country music video in country’s year of
cher Institute; 2004 the woman. J Commun. 1999;49:69 – 82
4. Panchaud C, Singh S, Feivelson D, Darroch JE. Sexually trans­ 28. Seidman SA. An investigation of sex-role stereotyping in music
mitted diseases among adolescents in developed countries. Fam videos. J Broadcast Electron. 1992;36:209 –216
Plann Perspect. 2000;32:24 –32, 45 29. Vincent RC. Clio’s consciousness raise? Portrayal of women in
5. Koyle, P, Jensen, L, Olsen, J. Comparison of sexual behaviors rock videos, re-examined. Journalism Quart. 1989;66:155–160
among adolescents having an early, middle and late first inter­ 30. Ward LM. Does television exposure affect emerging adults’
course experience. Youth Soc. 1989;20:461– 476 attitudes and assumptions about sexual relationships? Corre­
6. Roberts DF, Foehr UG, Rideout V. Generation M: Media in the lational and experimental confirmation. J Youth Adolesc. 2002;
Lives of 8 –18 Year-Olds. Menlo Park, CA: Henry J Kaiser 31:1–15
Foundation; 2003. Available at: www.kff.org/entmedia/ 31. Strouse JS, Buerkel-Rothfuss NL. Media exposure and the
entmedia030905pkg.cfm. Accessed December 15, 2005 sexual attitudes and behaviors of college students. J Sex Educ
7. Christenson PG, Roberts DF. It’s Not Only Rock and Roll: Popular Ther. 1987;13:43–51
Music in the Lives of Adolescents. Cresskill, NJ: Hampton Press; 32. Strouse JS, Buerkel-Rothfuss NL, Long E. Gender and family as
1998 moderators of the relationship between music video exposure
8. Lull J. On the communicative properties of music. Commun Res. and sexual permissiveness. Adolescence. 1995;30:505–521
1985;12:363–372 33. Tolman DL, Higgins T. How being a good girl can be bad for
9. Lull J Listeners communicative uses of popular music. In: Lull girls. In: Maglin NB, Perry D, eds. Good Girls/Bad Girls: Women,

e440 MARTINO et al
Downloaded from www.pediatrics.org by on March 13, 2007
 Sex, Violence, and Power in the 1990s. New Brunswick, NJ: Rut­ 49. Ware JE Jr, Sherbourne CD. The MOS 36-item short-form
gers University Press; 1996:205–225 health survey (SF-36). I. Conceptual framework and item se­
34. Fox G. Sex role attitudes as predictors of contraceptive use lection. Med Care. 1992;30:473– 483
among unmarried university students. Sex Roles. 1977;3: 50. Collins RL, Ellickson PL, Bell RM. Simultaneous polydrug use
265–283 among teens: prevalence and predictors. J Subst Abuse. 1998;
35. MacCorquodale P. Gender roles and premarital contraception. 10:233–253
J Marriage Fam. 1984;46:57– 63 51. Zuckerman M. Notes and shorter communications: item revi­
36. Pleck JH, Sonenstein FL, Ku LC. Masculinity ideology: its im­ sions in the Sensation Seeking Scale Form V (SSS-V). Pers Indiv
pact on adolescent males’ heterosexual relationships. J Soc Differ. 1996;20:515
Issues. 1993;49:11–29 52. Yuan KH, Marshall LL, Bentler PM. A unified approach to
37. Ward LM. Understanding the role of entertainment media in exploratory factor analysis with missing data, nonnormal data,
the sexual socialization of American youth: a review of empir­ and in the presence of outliers. Psychometrika. 2002;7:95–122
ical research. Dev Rev. 2003;23:347–388 53. Little RJA, Rubin DB. Statistical Analysis With Missing Data. 2nd
38. Brown JD, Schulze L. The effects of race, gender, and fandom ed. New York, NY: Wiley; 2002
on audience interpretation of Madonna’s music videos. J Com- 54. Santelli JS, Lindberg LD, Abma J, McNeely CS, Resnick M.
mun. 1990;40:88 –102 Adolescent sexual behavior: estimates and trends from four
39. Roberts DF. Media and youth: access, exposure, and privatiza­ nationally representative surveys. Fam Plann Perspect. 2000;32:
tion. J Adolesc Health. 2000;27(suppl):8 –14 156 –165, 194
40. Ward LM, Gorvine B, Cytron A. Would that really happen? 55. Brown JD, L’Engle KL, Pardun CJ, Guo G, Kenneavy K, Jack­
Adolescents’ perceptions of sexual relationships according to son C. Sexy media matter: exposure to sexual content in music,
prime-time television. In Brown J, Walsh-Childers K, Steele J, movies, television, and magazines predicts black and white
eds. Sexual Teens, Sexual Media. Hillsdale, NJ: Erlbaum; 2001: adolescents’ sexual behavior. Pediatrics. 2006;117:1018 –1027
95–123 56. Wingood GM, DiClemente RJ, Bernhardt JM, et al. A prospec­
41. Collins RL, Elliott MN, Berry SH, et al. Watching sex on TV tive study of exposure to rap music videos and African Amer­
predicts adolescent initiation of sexual behavior. Pediatrics. ican female adolescents’ health. Am J Public Health. 2003;93:
2004;114(1). Available at: www.pediatrics.org/cgi/content/ 437– 439
full/114/1/e1 57. Strasburger VC, Hendren RO. Rock music and rock music
42. Martino SC, Collins RL, Kanouse DE, Elliott M, Berry SH. videos. Pediatric Annals. 1995;24:97–103
Social cognitive mediators of the relationship between expo­ 58. Sun SW, Lull J. The adolescent audience for music videos and
sure to television’s sexual content and adolescents’ sexual be­ why they watch. J Commun. 1986;36:115–125
havior. J Pers Soc Psychol. 2005;89:914 –924 59. Roe K. The school and music in adolescent socialization. In:
43. Muthén LK, Muthén BO. Mplus User’s Guide. Los Angeles, CA: Lull J, ed. Popular Music and Communication. Beverly Hills, CA:
Muthén & Muthén; 2004 Sage; 1987:212–230
44. Kanouse DE, Collins RL, Elliott MN, Berry SH, Hunter SB, Miu 60. Roe K. Adolescents’ use of socially disvalued media: towards a
A. Non-coital and coital sexual behaviors of adolescents in the theory of media delinquency. J Youth Adolesc. 1995;24:
U.S. Paper presented at: Annual Meeting of the Population Asso- 617– 631
ciation of America; May 2, 2003; Minneapolis, MN 61. Zillmann D. Transfer of excitation in emotional behavior. In:
45. Miller K, Clark L, Wendell DA, et al. Adolescent heterosexual Cacioppo JT, Petty RE, eds. Social Psychophysiology: A Sourcebook.
experience: a new typology. J Adolesc Health. 1997;20:179 –186 New York, NY: Guilford; 1983:215–240
46. Ninety-Eight Degrees. Dizzy. On Revelation [CD]. Los Angeles, 62. Kunkel D, Biely E, Eyal K, Cope-Farrar K, Donnerstein E,
CA: Universal Records; 2000 Fandrick R. Sex on TV 3: A Biennial Report to the Kaiser Family
47. Ja Rule. Livin’ it up. On Pain is Love [CD]. New York, NY: Def Foundation. Menlo Park, CA: Henry J Kaiser Foundation; 2003
Jam Recordings; 2001 63. Emerson RA. “Where my girls at?”: negotiating black woman­
48. Kosterman R, Hawkins J, Guo J, Catalano R, Abbott R. The hood in music videos. Gender Soc. 2002;16:115–135
dynamics of alcohol and marijuana initiation: patterns and 64. Tapper J, Thorson E, Black D. Variations in music videos as a
predictors of first use in adolescence. Am J Public Health. 2000; function of their musical genre. J Broadcast Electron. 1994;38:
90:360 –366 103–113

PEDIATRICS Volume 118, Number 2, August 2006 e441

Downloaded from www.pediatrics.org by on March 13, 2007
Exposure to Degrading Versus Nondegrading Music Lyrics and Sexual Behavior
Among Youth
Steven C. Martino, Rebecca L. Collins, Marc N. Elliott, Amy Strachman, David E.
Kanouse and Sandra H. Berry
Pediatrics 2006;118;430-441
DOI: 10.1542/peds.2006-0131
This information is current as of March 13, 2007

Updated Information
including high-resolution figures, can be found at:
& Services
http://www.pediatrics.org/cgi/content/full/118/2/e430
References
This article cites 23 articles, 4 of which you can access for free
at:
http://www.pediatrics.org/cgi/content/full/118/2/e430#BIBL
Citations
This article has been cited by 2 HighWire-hosted articles:
http://www.pediatrics.org/cgi/content/full/118/2/e430#otherarticl
es
Post-Publication
5 P3Rs have been posted to this article:
Peer Reviews (P3Rs)
http://www.pediatrics.org/cgi/eletters/118/2/e430
Subspecialty Collections
This article, along with others on similar topics, appears in the
following collection(s):
Office Practice
http://www.pediatrics.org/cgi/collection/office_practice
Permissions & Licensing Information about reproducing this article in parts (figures,
tables) or in its entirety can be found online at:
http://www.pediatrics.org/misc/Permissions.shtml
Reprints Information about ordering reprints can be found online:
http://www.pediatrics.org/misc/reprints.shtml

Downloaded from www.pediatrics.org by on March 13, 2007