Promoting sexual health in high school: A feasibility study of a web-based media literacy education program

Abstract

Introduction:

Comprehensive sexual health education (SHE) is an effective strategy for improving adolescent sexual health. However, few of these programs address media influence on sexual cognitions and behaviors. Also, more research is needed on using web-based instruction for SHE.

Methods:

Seventeen classes (N=331 students) in one high school in the United States were enrolled in a pre-post randomized controlled trial to assess the feasibility of Media Aware, a web-based SHE program that uses a media literacy education (MLE) approach.

Results:

Compared to a delayed-intervention group, students who received Media Aware had significant reductions in their perceived realism of and similarity to media messaging, improved cognitive elaboration of media messages, more realistic perceptions of teen sex norms and risky sex norms, increased efficacy and intention to act as a bystander to potential sexual assault, increased intent to communicate before sex, and increased efficacy to use contraception/protection. These students reported being less willing to hook up, being less willing to have unprotected sex (for males), and positive feedback on their experiences using a web-based program.

Conclusions:

This study provides evidence that web-based MLE sexual health programming is a feasible and acceptable strategy for improving media-related and sexual health outcomes among adolescents.

Introduction

Adolescent Sexual Health

Nearly half of high schoolers have had sexual contact of some kind (Kann et al., 2018). However, many are not practicing safe sexual behaviors, like using contraception or another form of protection (Kann et al., 2018). Among industrialized countries, the United States (U.S.) has one of the highest teen birth rates (Sedgh, Finer, Bankole, Eilers, & Singh, 2015), and adolescents and young adults account for half of new sexually transmitted infection (STI) cases in the U.S. (Centers for Disease Control and Prevention, 2018). Of the more than half of high school students with relationship experience, 7% have experienced dating violence, and 8% have been physically hurt by a partner (Kann et al., 2018). Eight percent report being forced to have sexual intercourse (Kann et al., 2018). Dating violence and sexual assault are also risk factors for future unintended pregnancy and STIs (Silverman, Raj, & Clements, 2004), making these topics critical to address as a part of sexual health programming.

Sexual Health Education in the United States

School-based, comprehensive sexual health education (SHE) has been widely shown to be effective in delaying sexual debut, reducing the number of sexual partners, and increasing condom use (Kirby & Laris, 2009; Kirby, Laris, & Rolleri, 2006). However, only 24 states and DC mandate that SHE be provided in public schools, and of those only 17 require content to be medically accurate (Guttmacher Institute, 2019). SHE programs in U.S. public schools are funded at the federal and state level, but implemented at the local level (Hall, McDermott Sales, Komro, & Santelli, 2016). Thus, school-based adolescent SHE programming content and implementation are not standardized. Federal funding for abstinence-only curricula—rebranded as “sexual risk avoidance” programs—has increased (Boyer, 2018) despite evidence that these programs are ineffective at reducing sexual risk behaviors (Santelli et al., 2017). Research has estimated that the millions of federal dollars spent on abstinence-only programs has had no effect in reducing the number of adolescent births (Fox, Himmelstein, Khalid, & Howell, 2019). School-based SHE programming, when provided, is also often incomplete, focusing narrowly on pregnancy and disease prevention (Guttmacher Institute, 2019), despite the call to include information on gender, sexual assault, consent, and healthy relationships (Haberland & Rogow, 2015; Haberland, 2015).

Delivery of Sexual Health Programming

School-based SHE is provided primarily by teachers who are not necessarily trained to teach those health topics (Borawski et al., 2015; Centers for Disease Control, 2016). Teachers may skip topics they feel are controversial (Eisenberg, Madsen, Oliphant, & Sieving, 2013). In contrast, teachers who do receive professional development in SHE cover a broader range of sexual health topics and spend more time on the subject (Clayton, Brener, Barrios, Jayne, & Jones, 2017). Unfortunately, less than half of school districts have funds for teachers to receive professional training in human sexuality, HIV/other STIs, and pregnancy prevention (Centers for Disease Control, 2016).

Schools have increasingly turned to web-based instruction to prepare students for a world dominated by technology (Zheng, Warschauer, Lin, & Chang, 2016). This kind of instruction is often perceived as more engaging than traditional classroom-based instruction as students are able to self-direct and self-pace their learning (Varier et al., 2017). Health teachers are often required to teach many topics within a limited amount of class time. Self-paced learning can often be completed more quickly as teachers have less classroom behavior issues to manage. Notably for SHE, web-based programs allow for standardization of information, which reduces teacher bias and decreases the need for extensive teacher training and affords privacy to learn about sensitive topics.

Media influence

Media influence is one overarching, but often overlooked, factor that affects adolescent sexual health. Media use among teens is pervasive – almost all teens have access to a smartphone and approximately half report that they are online nearly constantly (Pew Research Center, 2018). Such high rates of media use means that teens are exposed to an incredible number of sexual references. One content analysis found that 85% of movies and 82% of television shows include sexual content (Ward, Erickson, Lippman, & Giaccardi, 2016). Estimates vary across studies, but a sizable minority of adolescents intentionally use pornography and even more have unintentionally been exposed (Peter & Valkenburg, 2016). Notably, media content regarding sex and relationships is often unhealthy (Hust, Brown, & L’Engle, 2008) or inaccurate (Brown, Halpern, & L’Engle, 2005; Hust et al., 2008). For example, in a content analysis of movies (1950–2006), 98% of sexual acts made no mention of contraception (O’Hara, Gibbons, Gerrard, Li, & Sargent, 2012). In contrast, healthy sexual messages in media are exceedingly rare (Hust et al., 2008; Ward & Friedman, 2006). Among popular adolescent television shows, only three percent had messages of sexual risk and responsibility (Signorielli & Bievenour, 2015) and nine percent of sexual acts in movies had healthy messaging (e.g., use of condoms) (Pardun, L’Engle, & Brown, 2005).

Researchers posit that media may act as a “super peer” that can influence adolescent sexual behavior (Brown et al., 2005; L’Engle, Brown, & Kenneavy, 2006). A meta-analysis of decades of research reveals an association between non-explicit sexual media exposure (i.e., verbal/visual references to sexual relationships, dating, or sexual acts in mainstream media) and sexual attitudes and behavior, which was found to be particularly strong during adolescence (Coyne et al., 2019). A systematic literature review of the impacts of sexually explicit media (e.g., pornography) and sexually violent media found that exposure to both are positively related to more accepting attitudes toward dating and sexual violence, as well as actual and anticipated dating violence and sexual violence victimization and perpetration (Rodenhizer & Edwards, 2017).

Media literacy education

Media literacy education (MLE) is a promising strategy for addressing negative media influence on health. Media literacy is generally defined as the ability to access, analyze, and produce media messages (Aufderheide, 1993). MLE has been shown to affect attitudes, normative beliefs, and behavioral intentions related to a number of health behaviors including alcohol use (Hindmarsh, Jones, & Kervin, 2015); tobacco use (Primack, Douglas, Land, Miller, & Fine, 2014; Scull, Kupersmidt, & Weatherholt, 2017); and disordered eating (McLean, Wertheim, Masters, & Paxton, 2017). MLE has also been applied with success to sexual health promotion. Evaluations of both abstinence-focused (Pinkleton, Austin, Chen, & Cohen, 2013; Pinkleton, Austin, Chen, & Cohen, 2012; Pinkleton, Austin, Cohen, Chen, & Fitzgerald, 2008), comprehensive (Scull, Kupersmidt, Malik, & Morgan-Lopez, 2018; Scull, Malik, & Kupersmidt, 2014); and pornography-specific (Rothman et al., 2018) classroom-based SHE for adolescents using an MLE approach have revealed many positive impacts on adolescent outcomes related to sexual health. However, relatively less is known about the effectiveness of this kind of intervention using a web-based format.

The Intervention

Media Aware - High School is a web-based, comprehensive SHE program for high school students that uses an MLE approach. It is a developmentally-appropriate adaptation of Media Aware - Young Adults, aimed at ages 18+ shown to reduce risky sexual behavior and improve other sexual health outcomes (Scull et al., 2018). Media Aware is designed to provide high school students with sexual health knowledge, media literacy skills, and healthy decision-making skills regarding sexual activity and relationships. Media Aware consists of four highly interactive, self-paced modules, each designed to be completed within one traditional class period.

The program uses text-based narration, streaming media examples, videos, animations, and interactivities to present course content. There are many opportunities for students to analyze media messages and receive automatic feedback on their responses. This may include medically accurate and developmentally appropriate health information and resources and/or peer videos that includes a group of diverse high school students discussing their answers to the media message deconstructions or thoughts surrounding the health topics at hand.

Module 1:

Students set personal goals related to school success and health. Students identify media as influential sources of health information in their lives and are provided with tools to critically analyze media messages. Students examine gender role stereotypes, the realism and implied messages in media, and healthy and unhealthy relationships.

Module 2:

Students identify the ways in which media may promote and/or normalize substance use, sexual assault, and dating violence and learn more advanced message analysis to redress inaccurate norms. Students are introduced to the concept of consent, are taught to recognize when consent is needed, when consent cannot be given, and learn and practice bystander intervention skills.

Module 3:

Students analyze media messages with a focus on how media messages frequently omit information or provide misinformation about the use of contraception/protection. Students learn about FDA-approved methods of contraception and methods of protection against sexually transmitted infections, differentiating between perfect and typical use.

Module 4:

Students analyze media messages with a focus on how infrequently communication is portrayed prior to sexual activity. Students learn skills needed to have effective sexual health conversations with trusted adults, romantic partners, and medical professionals – including sexual refusal and contraception negotiation skills. Students get skills-based practice in applying techniques, reflect on personal goals, and create a plan for the future.

Theoretical frameworks and hypotheses

The development of Media Aware was informed by theoretical frameworks of media message processing [i.e., the Message Interpretation Process (MIP) model; Austin & Johnson, 1997a], and behavior change [i.e., Theories of Reasoned Action and Planned Behavior (TRA/TPB); Ajzen & Fishbein, 1975; Ajzen, 1991]. The present study examines the feasibility of using web-based MLE sexual health promotion programming for improving adolescents’ short-term sexual health and media-related outcomes. The results of this study can inform the development of future sexual health promotion programs.

The MIP model addresses the processes associated with the interpretation of media messages such that perceived realism of, similarity to, and desirability of media messages contribute to the level of identification with the message. The model posits that the level of identification with the message predicts behaviors. While the MIP model posits that perceived desirability of media messages, which is the affective component of the model, also impacts behaviors, there are complex questions about how desirability functions in the model. Research suggests that decreasing desirability is not necessary for MLE program effectiveness and, in fact, some MLE programs have been shown to increase desirability while also favorably impacting health outcomes – “the double-edged desirability hypothesis” (Austin et al., 2015). Therefore, the Media Aware was designed to address the cognitive aspects of the model, specifically decreasing perceived realism of and similarity to unhealthy sexual media messages, while also enhancing skepticism of and critical thinking (e.g., cognitive elaboration) about media messages about sex and substance use. Accordingly, the following hypothesis was tested:

Hypothesis 1:

Compared to students in the delayed-intervention group, students in the intervention group (Media Aware) will have more positive, self-reported logic-based media-related outcomes at posttest, including: (1) lower perceived realism of media messages, (2) lower perceived similarity to media messages, (3) more cognitive elaboration when processing a media message, and (4) higher media skepticism about sex and substance use messages.

The MIP model is consistent with the TRA/TPB, which has empirical support for predicting safe sex behaviors (Albarracín, Johnson, Fishbein, & Muellerleile, 2001). The TRA/TPB posits that behavioral beliefs and perceived behavior control influence behavioral intentions which, in turn, impact behaviors. The basic premise of MLE is that teaching critical thinking skills creates an active filter through which media images are processed, which is expected to change adolescents’ cognitions about sexual practices, beliefs about normative sexual practices, and self-efficacy to engage in healthy behaviors. These changes result in increased intentions for healthy behaviors and less willingness for risk behaviors. Accordingly, the following hypothesis was tested:

Hypothesis 2:

Compared to students in the delayed-intervention group, students in the intervention group (Media Aware) will have more positive, self-reported predictors of sexual health outcomes at posttest, including: (1) healthier normative beliefs about the frequency of teen sexual activity and teen risky sexual activity, (2) healthier normative beliefs regarding the acceptance of rape, dating violence, and gender role stereotypes, (3) more self-efficacy and more intentions to communicate about sexual health, refuse sexual activity, use contraception, and act as a bystander to prevent sexual assault and (4) less willingness to hook up and less willingness to engage in unprotected sexual activity.

Given the challenges associated with consistent, accurate, and unbiased delivery of teacher-led SHE in the classroom and the potential for student discomfort in discussing sexual health content in a teacher-led setting, web-based programs are likely to provide a more comfortable and engaging approach to SHE for students. Therefore, the following hypothesis was tested:

Hypothesis 3:

Students will report positive feedback about completing a web-based SHE program.

Methods

Participants

A two-armed (intervention, delayed-intervention), randomized, pretest-posttest study evaluated the feasibility of Media Aware for improving media-related and sexual health outcomes in high school students. One large U.S. high school with four health education teachers and 17 ninth-grade health education classrooms participated. Health teachers and their classrooms were randomized to condition: intervention (n = 2 teachers; 8 classes; 212 students) and delayed-intervention (n = 2 teachers; 9 classes; 212 students). Students were eligible to participate if they had permission to receive SHE and were fluent in English.

The final sample included 331 participants with slightly more male students than female students (see Table 1). A few individuals identified as gender non-binary (represented as missing data in the gender analyses). Fifty-six percent of the sample identified as white and not of Hispanic, Latino, or Spanish descent. The average age of participants was 14.53 years (SD=.57).

Table 1.

Sample demographic characteristics and baseline equivalence tests from pretest data collection.

	Total (N = 331)a	Control (n = 165)	Intervention (n = 166)
	n (%)	n (%)	n (%)	df	χ2	p-value
Gender				1	0.39	0.53
Male	168 (55.08)	81 (53.29)	87 (56.86)
Female	137 (44.92)	71 (46.71)	66 (43.14)
Missing	26
Age in years				4	4.21	0.38
13	2 (0.63)	2 (1.27)	0 (0.00)
14	152 (48.10)	80 (50.63)	72 (45.57)
15	154 (48.73)	72 (45.57)	82 (51.90)
16	7 (2.22)	3 (1.90)	4 (2.53)
17	1 (0.32)	1 (0.63)	0 (0.00)
Missing	15
Ethnicity				1	0.23	0.63
Hispanic	53 (16.93)	25 (15.92)	28 (17.95)
Non-Hispanic	260 (83.07)	132 (84.08)	128 (82.05)
Missing	18
Race				5	4.40	0.49
American Indian	1 (0.34)	0 (0.00)	1 (0.68)
Asian	46 (15.65)	22 (14.86)	24 (16.44)
Black	24 (8.16)	9 (6.08)	15 (10.27)
Pacific Islander/Native Hawaiian	3 (1.02)	1 (0.68)	2 (1.37)
White	173 (58.84)	94 (63.51)	79 (54.11)
More than one race	47 (15.99)	22 (14.86)	25 (17.12)
Missing	37
Sexual Experience				2	0.31	0.85
Yes	46 (14.98)	23 (15.23)	23 (14.74)
No	258 (84.04)	127 (84.11)	131 (83.97)
Unsure	3 (0.98)	1 (0.66)	2 (1.28)
Missing	24

^aEleven students were absent for the pretest data collection.

Procedure

All procedures were reviewed and approved by an Institutional Review Board. Teachers received $30/class if their classes returned at least 80% of their informed consent forms regardless of participation decision. See Figure 1 for a breakdown of the intervention and delayed-intervention groups.

Figure 1.

CONSORT Flow Diagram

Researchers administered the web-based pretest in classrooms. In the week between pre- and posttest, students in the intervention group completed Media Aware at their own pace during class. Teachers in the delayed-intervention group were instructed not to teach sexual health or media literacy topics during the same time period. Approximately one week after the pretest, researchers administered the web-based posttest to all participating students. After posttest, delayed-intervention teachers were allowed to use Media Aware with their classes. Students received a small incentive (i.e., pen, sticky flags) for completing each questionnaire. Students that completed both questionnaires were entered into a classroom drawing for a $25 gift card. The school received a $250 incentive for participation.

Measures

The pretest questionnaire assessed demographic characteristics (i.e., age, gender, school performance, socioeconomic status, race, ethnicity, and parental education), past sexual behavior, and current dating status. The pretest and posttest included media-related and sexual health outcomes. Response scales, sample items, alphas, and sources for these constructs can be found in Table 2.

Table 2.

Media-related and sexual health outcome measures

Category	Measures	#	scale	Sample item	α
Media-related Outcomes	Realisma	5	1-strongly disagree, 2=disagree, 3=agree, 4=strongly agree	Teens in the media…do things that average teens do.	0.80
	Similaritya	3	1-strongly disagree, 2=disagree, 3=agree, 4=strongly agree	How much do you agree or disagree with the following statements? The things I do in my life are similar to what I see teens do in the media.	0.79
	Cognitive elaborationb	3	1=not much at all, 2=a little; 3=a good amount, 4= a lot	(When presented with an advertisement) – Please think about the advertisement you just saw. How much did you think about this advertisement?	0.86
	Skepticism of media messages about sex and substance usec	3	1-strongly disagree, 2=disagree, 3=agree, 4=strongly agree	How much do you agree or disagree with the following statements? Media are dishonest about what happens if people have sex.	0.76
Sexual Health Outcomes	Teen sex descriptive normsd	6	0=none, 100=all	What percentage of teens between 0% (no teens) and 100% (all teens) …are having vaginal sex?	--
	Risky sex descriptive normsd	4	0=none, 100=all	What percentage of teens between 0% (no teens) and 100% (all teens) …hook up and engage in sexual behavior, but not have oral, anal, or vaginal sex with someone they are not in a relationship with?	--
	Dating violence normse	4	1-strongly disagree, 2=disagree, 3=agree, 4=strongly agree	How much do you agree or disagree with the following statements? It is OK for people to hit their girlfriends/boyfriends if they did something to make them mad.	0.74
	Gender role normse	6	1-strongly disagree, 2=disagree, 3=agree, 4=strongly agree	How much do you agree or disagree with the following statements? Raising children is primarily a woman’s responsibility.	0.87
	Rape myth acceptancef	5	1-strongly disagree, 2=disagree, 3=agree, 4=strongly agree	How much do you agree or disagree with the following statements? It shouldn’t be considered rape if a guy is drunk and didn’t realize what he was doing.	0.88
	Bystander efficacyg	5	0 (can’t do), 10 (quite uncertain), 20, 30, 40, 50 (moderately certain), 60, 70, 80, 90, 100 (very certain)	I could talk to a friend who I suspected is in an abusive relationship.	0.85
	Bystander intentionsh	4	1=not at all likely, 2=unlikely, 3=likely, 4=extremely likely	Approach a friend if I thought they were in an abusive relationship and let them know that I am here to help.	0.81
	Sexual health knowledgec	13	Multiple choice or true/false; range correct 0–13	You can tell if someone has an STI by looking at him/her. (true/false)	--
	Communication intentionsc	5	1=not at all likely, 2=unlikely, 3=likely, 4=extremely likely	Before deciding to have sex, how likely would you be to…talk with a boy/girlfriend about HIV/AIDS and other STIs	0.84
	Teen sexual intentionsi	5	1=not at all likely, 2=unlikely, 3=likely, 4=extremely likely	How likely is it that you will have any type of sexual contact with another person (oral sex, anal sex, vaginal sex, or genital-to-genital contact) in the next year?
	Contraception/protection intentionsj	3	1=not at all likely, 2=unlikely, 3=likely, 4=extremely likely	If you were to have vaginal or anal sex, how likely would you be to use a condom?	0.71
	Contraception/protection efficacy k	3	1-strongly disagree, 2=disagree, 3=agree, 4=strongly agree	How much do you agree or disagree with the following statements? If I decided to have sex, I could use a condom correctly or explain to my partner how to use a condom correctly.	0.74
	Communication efficacyk	7	1-strongly disagree, 2=disagree, 3=agree, 4=strongly agree	How much do you agree or disagree with the following statements? I could talk with a boy/girlfriend about using condoms for STI protection.	0.90
	Refusal efficacyk	5	1-strongly disagree, 2=disagree, 3=agree, 4=strongly agree	How much do you agree or disagree with the following statements? I could say no to someone who is pressuring me to have sex.	0.89
	Willingness to hook upl	1	1=very willing, 2=unwilling, 3=willing, 4=very willing	Suppose you were with a boyfriend/girlfriend. He/she wants to hook-up, but you are not sure that you want to. In this situation, how willing would you be to go ahead and hook-up anyway?	--
	Willingness for unprotected sexl	1	1=very willing, 2=unwilling, 3=willing, 4=very willing	Suppose you were with a boyfriend/girlfriend. He/she wants to have sex, but neither of you have any form of protection. In this situation, how willing would you be to go ahead and have sex anyway?	--

^a.Adapted from (Pinkleton, Austin, Cohen, Chen, & Fitzgerald, 2008; Scull, Malik, & Kupersmidt, 2014)

^b.Adapted from (Shiv, Britton, & Payne, 2004)

^c.Adapted from (Scull et al., 2014)

^d.Adapted from (Scull, Kupersmidt, Malik, & Keefe, 2018)

^e.Adapted from (Foshee et al., 2005)

^f.Adapted from (Payne, Lonsway, & Fitzgerald, 1999)

^g.Adapted from (Banyard, Plante, & Moynihan, 2005)

^h.Adapted from (Banyard & Cross, 2008)

^i.Adapted from (L’Engle, Brown, & Kenneavy, 2006)

^j.Adapted from (Jemmott & Jemmott, 1991)

^k.Adapted from (Soet, Dudley, & Dilorio, 1999)

^l.Adapted from (Gibbons, Gerrard, Blanton, & Russell, 1998)

The intervention group provided program feedback on the posttest. Participants were asked how much they agree or disagree with the following statements: 1) I liked that I could do this program on a computer; 2) I liked this program better than a teacher teaching sexual health education in the classroom; 3) I felt less embarrassed taking this program than having a teacher teaching sexual health; and 4) This is a good program for teens to learn about sexual health. The response scale was 4 points (Strongly disagree to Strongly Agree). Students were also prompted to respond to, “What did you like best about Media Aware?”

Analysis

Summary scores were calculated for scaled constructs, and psychometric analyses examined internal consistency of measures (see Table 2). Overall, there were low levels of missing data (<7%). Descriptive statistics were calculated for pretest outcomes (see Table 3). Pretest correlations between outcomes were examined (see Table 4). Correlations for outcomes ranged from a minimum of −0.00 and a maximum of 0.57. All significant relationships between outcomes were as expected except for small correlations between bystander efficacy and perceived similarity (.16); bystander efficacy and teen sex descriptive norms (.14); and bystander intentions and teen sex descriptive norms (.12). Overall, the pattern of interscale correlations suggested that multicollinearity would not be an issue.

Table 3.

Descriptive statistics for outcome variables at pretest.

	Intervention		Control		Total
Measure	M	SD	M	SD	M	SD	min	max
Media-related outcomes
Realism	2.19	0.50	2.16	0.51	2.18	0.50	1.00	4.00
Similarity	2.04	0.66	2.08	0.54	2.06	0.60	1.00	4.00
Cognitive elaboration	2.08	0.69	2.16	0.72	2.12	0.71	1.00	4.00
Skepticism	3.08	0.39	3.05	0.43	3.06	0.41	1.50	4.00
Sexual health outcomes
Teen sex descriptive norms	39.31	20.15	38.43	19.00	38.87	19.55	0.00	100.00
Risky sex descriptive norms	32.34	19.14	31.49	19.32	31.91	19.20	0.00	100.00
Dating violence norms	1.46	0.49	1.36	0.44	1.41	0.47	1.00	3.00
Gender role norms	1.49	0.50	1.45	0.45	1.47	0.48	1.00	3.17
Rape myths	1.31	0.46	1.32	0.44	1.31	0.45	1.00	2.80
Bystander efficacy	70.98	20.25	74.35	20.14	72.66	20.23	10.00	100.00
Bystander intentions	3.08	0.50	3.15	0.55	3.12	0.53	1.00	4.00
Knowledge	11.09	1.44	11.36	1.34	11.23	1.40	5.00	13.00
Communication intentions	3.03	0.59	2.99	0.58	3.01	0.58	1.00	4.00
Teen sexual intentions	1.94	0.80	2.04	0.78	1.99	0.79	1.00	4.00
Contraception/protection intentions	3.28	0.62	3.16	0.67	3.22	0.65	1.00	4.00
Contraception/protection efficacy	3.21	0.56	3.16	0.58	3.19	0.57	1.33	4.00
Communication efficacy	3.43	0.49	3.39	0.56	3.41	0.53	1.00	4.00
Refusal efficacy	3.39	0.52	3.39	0.55	3.39	0.53	1.00	4.00
Willingness to hook up	2.20	0.80	2.19	0.72	2.19	0.76	1.00	4.00
Willingness for unprotected sex	1.59	0.78	1.66	0.72	1.63	0.75	1.00	4.00

Table 4.

Correlation table for the outcome variables

		1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20
1	Realism	1.00
2	Similarity	0.32***	1.00
3	Cognitive Elaboration	−0.05	−0.02	1.00
4	Skepticism	−0.05	−0.03	0.04	1.00
5	Teen sex descriptive norms	0.13*	0.11*	−0.16*	−0.03	1.00
6	Risky sex descriptive norms	0.05	0.09	−0.06	−0.12*	0.69***	1.00
7	Dating violence norms	−0.03	−0.01	−0.07	−0.15*	−0.10	−0.06	1.00
8	Gender role norms	0.04	0.09	−0.10	−0.23***	−0.003	−.0.00	0.24***	1.00
9	Rape myth acceptance	−0.001	−0.01	−0.09	−0.34***	0.05	0.11	0.35***	0.57***	1.00
10	Bystander efficacy	0.09	0.16*	0.34	0.01	0.14*	0.18*	−0.24***	−0.19**	−0.21**	1.00
11	Bystander intentions	0.03	0.09	0.08	0.16**	0.12*	0.18*	−0.27***	−0.34***	−0.32***	0.42***	1.00
12	Knowledge	−0.01	0.08	−0.01	0.28***	0.03	−0.09	−0.19**	−0.28***	−0.32***	0.13*	0.27***	1.00
13	Communication intentions	−0.002	−0.12*	0.12*	−0.01	−0.05	−0.01	−0.01	−0.12*	−0.10	0.20**	0.27***	0.01	1.00
14	Sexual intentions	0.15*	0.44***	−0.05	−0.04	0.33***	−0.06	−0.06	0.19**	0.05	0.10	0.11	0.10	−0.28***	1.00
15	Contraception intentions	−0.03	−0.14*	0.11	0.08	−0.19**	−0.05	−0.05	−0.21**	−0.23***	0.12*	0.22***	0.12*	0.47***	−0.37***	1.00
16	Contraception efficacy	−0.03	0.05	0.07	−0.02	0.01	−0.05	−0.05	−0.01	−0.11*	0.26***	0.16*	0.09	0.34***	0.06	0.21**	1.00
17	Negotiation efficacy	0.01	−0.05	0.07	0.09	0.05	−0.13*	−0.13*	−0.23***	−0.22**	0.32***	0.34***	0.19**	0.47***	−0.10	0.35***	0.39***	1.00
18	Refusal efficacy	0.04	−0.15*	0.03	0.25***	−0.05	−0.19**	−0.19**	−0.31***	−0.31***	0.28***	0.28***	0.08	0.35***	−0.30***	0.33***	0.29***	0.34***	1.00
19	Willingness hookup	0.02	0.22**	−0.11	−0.15*	0.07	0.17*	0.17*	0.26***	0.15*	−0.09	−0.20**	−0.07	−0.31***	0.41***	−0.28***	−0.09	−0.12*	−0.44***	1.00
20	Willingness unprotected	−0.06	0.18*	−0.04	−0.23***	0.14*	0.03	0.03	0.31***	0.24***	−0.06	−0.16*	−0.16*	−0.31***	0.46***	−0.39***	−0.10	−0.27***	−0.41***	0.42***	1.00

Note:

^*p<.05.

^**p<.001.

^***p<.0001.

Chi-squared analyses did not reveal differences between groups with respect to demographic characteristics (Table 1). Examination of intervention dosage found that students spent an average time of 91 minutes using Media Aware (SD = 43 min). A series of single factor, random effects ANCOVA models were estimated for each outcome in intent-to-treat analyses. Hierarchical linear models (random intercept at the teacher level using SAS MIXED) were used to account for conditional non-independence. The model for the mean of each outcome contained fixed effects representing the influence of (1) the adolescent’s pre-intervention outcome score, (2) condition, (3) gender, and (4) the interaction between gender and condition. The results are reported as adjusted posttest mean scores.

Students provided both quantitative and qualitative feedback about their experiences with Media Aware. The percent of respondents agreeing with each statement (Agree or Strongly Agree) was calculated. Two coders coded open-ended responses about what students liked best about the program. Each coder scored the responses as to whether they referenced any of the following categories (yes/no): program was web-based; privacy; self-paced format; increased comfort/reduced embarrassment. Cohen’s kappas were calculated for each category (mean κ=.89; range .83–.94).

Results

Table 5 summarizes the findings from the outcome analyses. Findings revealed are main effects of condition with the exception of willingness for unprotected sex, which resulted in a significant condition by gender interaction. Importantly, all program effects represent a desired positive outcome for adolescents.

Table 5.

Adjusted posttest means, standard errors, F-statistics, effect size (Cohen’s d), and p-values for the outcome analyses

Category	Outcome	Control	Intervention	F-stat	Cohen’s d	p-value
		M (SE)	M (SE)
Media	Realism	2.19 (.06)	1.92 (.06)	(1,282) = 11.70***	.38	<.001
	Similarity	2.09 (.04)	1.94 (.04)	(1,278) = 7.63**	.31	0.01
	Cognitive elaboration	1.98 (.07)	2.19 (.07)	(1,282) = 3.81*	.24	0.05
	Skepticism	2.90 (.05)	3.01 (.05)	(1,280) = 2.37		0.12
Sexual Health	Teen sex descriptive norms	39.00 (1.17)	30.05 (1.21)	(1,260) = 28.26***	.67	<.001
	Risky sex subjective norms	30.15 (1.30)	26.17 (1.32)	(1,258) = 4.59*	.27	0.03
	Dating violence norms	1.39 (.05)	1.47 (.05)	(1,281) = 1.13		0.29
	Gender role norms	1.56 (.03)	1.50 (.03)	(1,281) = 1.79		0.18
	Rape myths	1.43 (.04)	1.37 (.04)	(1,278) = 1.11		0.29
	Bystander efficacy	72.74 (1.14)	76.97 (1.14)	(1,279) = 6.89**	.31	0.01
	Bystander intentions	3.08 (.03)	3.20 (.03)	(1,281) = 7.30**	.33	0.01
	Knowledge	11.39 (.09)	11.85 (.10)	(1,267) = 11.75***	.42	<.001
	Communication intentions	3.01 (.04)	3.14 (.04)	(1,266) = 6.31*	.28	0.01
	Teen sexual intentions	2.02 (.04)	1.96 (.04)	(1,273) = 1.15		0.28
	Contraception/protection intentions	3.26 (.04)	3.32 (.04)	(1,271) = .94		0.33
	Contraception/protection efficacy	3.11 (.05)	3.28 (.05)	(1,274) = 5.25*	.27	0.02
	Communication efficacy	3.23 (.03)	3.28 (.03)	(1,277) = 1.09		0.30
	Refusal efficacy	3.40 (.03)	3.47 (.03)	(1,281) = 2.12		0.15
	Willingness to hook up	2.24 (.05)	1.98 (.05)	(1,269) = 13.44***	.44	<.001
	Willingness for unprotected sex (males only)	1.70 (.04)	1.58 (.05)	(1,271) = 4.53*	.32	0.03

Note.

^***p<.001

^**p<.01

^*p<.05

Media-related Outcomes

Hypothesis 1 was supported in part, as students who received Media Aware reported many positive media-related outcomes at posttest compared with students in the delayed-intervention group. They reported significantly lower perceived realism of media messages; less perceived similarity to media messages; and higher levels of cognitive elaboration when thinking about an advertisement. However, there were no reported differences between the two groups with respect to skepticism of media messages about sex and substance use.

Sexual Health Outcomes

Hypothesis 2 was supported in part, as students who received Media Aware reported many positive sexual health-related outcomes at posttest compared with students in the delayed-intervention group. Students in the intervention group were significantly less likely to believe that teen sexual activity and teen risky sexual activity was normative. They reported more efficacy for and intention to intervene if they were a bystander to a potential sexual assault. They also scored higher on sexual health knowledge and reported higher efficacy for using protection, if they decided to be sexually active. These students reported more intention to communicate about sexual health with parents, partners, and medical professionals. Finally, they reported being less willing to hook up when they did not necessarily want to and, males specifically, reported less willingness to have unprotected sex. However, students in both groups reported about the same level of acceptance of dating violence, gender role stereotypes, and rape, the same level of efficacy for sexual health communication or refusing sexual activity, and about equal intentions for sexual activity and use of protection, if they decided to be sexually active.

Program Feedback

Hypothesis 3 was supported; intervention students reported positive feelings about Media Aware. Students reported affirmatively that they liked that they could do the program on a computer (83%); they liked this program better than having a teacher provide sexual health instruction (73%); felt less embarrassed taking this program than having a teacher deliver sexual health instruction (73%); and felt that Media Aware was a good program for teens to learn about sexual health (82%).

About 93% of intervention students responded with an answer to the open-ended question asking what they liked best about the program. Nearly half of students (45%) mentioned at least one aspect related to web-based program format. Specifically, students were most likely to mention that what they liked best was the web-based format in general (20%); self-paced content (18%); privacy afforded (12%); and comfort afforded (5%). Some examples of student responses include: 1) that we did it on computers instead of with a teacher; 2) it’s online so I can go at my own pace; 3) it was a lot better to learn about it privately rather than having open discussions about it in class; and 4) it would’ve been awkward to talk to someone else. I like that it is anonymous.

Discussion

Research has illuminated the potentially harmful role that media messages can play in adolescents’ sexual health as well as the protective role of comprehensive SHE. However, there are few evidence-based resources that address both factors. This study details the findings from a feasibility study of Media Aware, a web-based comprehensive SHE program that uses an MLE approach to improve sexual health outcomes for high school students. The study used a rigorous randomized control trial (RCT) design with a delayed-intervention group ensuring that all students in the study had the opportunity to receive the SHE program. Results from the analyses provide initial evidence that this strategy is a promising approach for U.S. high school SHE as Media Aware was shown to positively affect both media-related and sexual health outcomes in adolescents, in addition to being well-received by students.

Media messages can function as sex educators for youth, informing expectations of what is normal. Therefore, it is important to increase youths’ critical thinking skills about media messages. Results from this study show that completing Media Aware resulted in positive impacts on these media-related outcomes, which can serve as protective factors against the potentially negative effects of unhealthy sexual and romantic media messages. Per the MIP model, perceived realism of and similarity to media messages predict identification with the messages and, in turn, behaviors. For example, if an adolescent perceives a media message depicting unprotected sex without consequence as not realistic and they perceive themselves not to be similar to the people in media messages, they are less likely to identify with the message, and, in turn, would be less likely to engage in unprotected sex. Similarly, higher levels of cognitive elaboration are indicative of more careful message processing; when effortful processing results in predominately negative thoughts (e.g., STIs can be a consequence of having unprotected sex), the message is more likely to be rejected and its persuasive effect attenuated (Petty & Cacioppo, 1986). MLE may favorably impact sexual health outcomes by, in part, encouraging students to carefully and critically process sexual media messages and evaluate the accuracy and realism of the information provided in the media message. This careful processing of sexual media messages, paired with accurate sexual health knowledge and beliefs, may also attenuate or override the affect-based impact of media messages on sexual health outcomes. Skepticism of media messages may also encourage careful and critical processing of media messages. In this study, skepticism of media messages about sex and substance use was not found to be significantly impacted by the program. High levels of skepticism were reported across both conditions at pretest; this potential ceiling effect may limit the ability to identify potential changes in skepticism All three skepticism items were loaded toward perceiving media messages as “dishonest,” which may have predisposed youth to report more skepticism to start. A more balanced set of skepticism items should be considered for future research. This study employed intent-to-treat analyses, so program effects were likely underestimated, and complier average causal effect analyses may reveal significant program effects on skepticism. Despite this, these results present convincing evidence that Media Aware can enhance the message processing skills of high school students.

Adolescents’ normative beliefs about sexual activity are linked to their future sexual behaviors and have been found to be more predictive than peer pressure (van de Bongardt, Reitz, Sandfort, & Dekovic, 2015). Teens who believe that their peers are sexually active and engaging in risky behavior are more likely to engage in risky sexual activity themselves (Prinstein, 2003; van de Bongardt et al., 2015). The results showed that Media Aware lowered descriptive normative beliefs (i.e., estimates of the frequency) of teen sexual activity in general, and risky sexual activity, specifically. However, students’ subjective normative beliefs (i.e., perceived behavioral expectations) about rape, strict gender roles, and dating violence did not change as a result of the program. This could be due to participants’ low baseline endorsement of these beliefs in general. U.S. cultural conversations about gender, consent, and dating violence have been spotlighted recently, and high schoolers today may be less likely to personally accept strict gender roles, rape myths, and dating violence than previous generations. However, adolescent sexual assault is still a concern. Bystander intervention has become an increasingly common strategy for preventing sexual assault. Self-efficacy and intentions for acting as a bystander to prevent sexual assault are both found to influence future bystander behavior (McMahon et al., 2015). Students who completed Media Aware felt both more efficacy for and greater intentions to intervene as a bystander to a potential sexual assault compared with students in the delayed-intervention group. These findings are promising but additional work remains to be done in this area.

Sexual health communication with partners and parents and the use of protection are key to safe sexual behaviors. Such abilities and intentions influence future communication and behaviors (DiClemente et al., 2001; Stone & Ingham, 2002). Media Aware increased students’ intentions to communicate with a parent, medical professional, or partner before engaging in sexual activity. Furthermore, students reported more efficacy to use contraception/protection if they decide to engage in sexual activity. The program did not impact students’ efficacy to communicate about sexual health, efficacy to refuse sexual activity, or intentions to use contraception/protection in the future. At pretest, students in both groups endorsed very high levels for each of these constructs suggesting a ceiling effect that allows little room for improvement during the study timeframe. Students who completed Media Aware reported less willingness to hook up in situations where they did not necessarily want to. Boys, specifically, reported less willingness to engage in unprotected sex. Changes in willingness as a measure of subsequent behaviors is significant because previous studies found that willingness is a better predictor of future behaviors than intentions (Gerrard, Gibbons, Houlihan, Stock, & Pomery, 2008). Overall, the findings related to sexual health communication and safe sexual behaviors are promising indicators that students who receive Media Aware will engage in fewer risky sexual behaviors in the future.

Finally, a concern of critics of comprehensive sexual education is that learning about sexual health will encourage youth to become sexually active. Students who received Media Aware learned many important sexual health facts in the program, as evidenced by higher levels of sexual health knowledge and positive impacts on normative beliefs, efficacy, and intentions related to sexual health. However, students receiving Media Aware did not report greater intentions to engage in sexual activity in the future compared with the delayed-intervention group. This is evidence that Media Aware increased sexual health knowledge without encouraging teen sexual activity.

Researchers have suggested that SHE that is considered engaging and relevant is more likely to have an impact (Byers, Sears, & Foster, 2013; Wight, 1999). This program was received positively by students. Students particularly enjoyed that the web-based program afforded them privacy, comfort, and the ability to learn at their own pace. Furthermore, the web-based nature of the program allowed students to receive consistent, medically-accurate content, free from potential teacher bias. These findings suggest the web-based format of Media Aware is a feasible approach to comprehensive SHE.

Limitations

There are a few limitations to consider. This study was conducted within one school, and these students may not be representative of the U.S. adolescent population at large. This also raises the the possibility of contagion between students in different conditions making it more difficult to detect significant effects. Because there were only a few participants who identified as gender non-binary, analyses were only able to examine binary gender effects of the program. Future evaluations should include a larger sample size that would allow for more complete analyses considering gender identity and sexual orientation. While the program was designed to be non-heteronormative (e.g., inclusion of gender-neutral names, pronouns, and images), it does not does explicitly discuss sexual orientation or gender identity. It is essential that sexual health programming discuss these topics in a factual, affirming, and inclusive manner; therefore, future work should expand program content. Involving youth in acts of self-reflection through media message production activities has been emphasized as a way to enhance and sustain positive gains from MLE (Greene, 2013). However, Media Aware does not include these activities. Future research should examine ways that media message production could impact media- and health-related outcomes. This line of research would benefit from a program evaluation study with variables measured at a minimum of three time points to allow a mediator analysis to determine the mechanisms through which MLE impacts media-related and sexual health outcomes. Likewise, this study examined changes within a short time period, but a longer-term study would enable the examination of whether Media Aware impacted outcomes like sexual debut and safe sex behaviors as well as determine whether findings emerge or degrade over time. Finally, this version of Media Aware did not have an outlet for students to have questions answered as they completed the program, identifying an important direction for future development work.

Conclusion

Comprehensive SHE can improve adolescent health outcomes. However, SHE in the United States is varied – sometimes biased, inadequate, inaccurate, or missing completely. Web-based SHE offers an opportunity to provide relevant and engaging content in a potentially less embarrassing format than teacher-led instruction. Of note, media exposure has been shown to influence adolescents’ sexual socialization and risky sexual behaviors, as well as negatively impacting attitudes around dating violence, rape, and consent. However, SHE for high school students does not typically address media influence on sexual health. This study found that Media Aware, a web-based MLE sexual health program is a feasible intervention to improve the sexual health of high school students. Students showed improvements in many media-related and sexual health outcomes as a result of the program and reported having a very favorable experience using the program. The results of this study have implications for future work aimed at engaging high school students in SHE and addressing media influence, an often ignored, unhealthy impact on adolescent sexual health.