write an argument essay(800 words)

Article

Design Effectiveness Analysis of a Media Literacy Intervention to Reduce Violent Video Games Consumption Among Adolescents: The Relevance of Lifestyles Segmentation

Reynaldo Rivera 1 , David Santos

2 ,

Gaspar Brändle 3 , and Miguel Ángel M. Cárdaba

4

Abstract Background Exposure to media violence might have detrimental effects on psychological adjustment and is associated with aggression-related atti- tudes and behaviors. As a result, many media literacy programs were imple- mented to tackle that major public health issue. However, there is little evidence about their effectiveness. Evaluating design effectiveness,

1 InterMedia Social Innovation, Rome, Italy 2 Universidad Autónoma de Madrid, Madrid, Spain 3 Universidad de Murcia, Murcia, Spain

4 Centro Universitario Villanueva, Madrid, Spain

Corresponding Author:

Reynaldo Rivera, Via Torquato Taramelli 14, int. 2, Rome, 00197, Italy.

Email: [email protected]

Evaluation Review 2016, Vol. 40(2) 142-161

ª The Author(s) 2016 Reprints and permission:

sagepub.com/journalsPermissions.nav DOI: 10.1177/0193841X16666196

erx.sagepub.com

particularly regarding targeting process, would prevent adverse effects and improve the evaluation of evidence-based media literacy programs. Objectives The present research examined whether or not different rela- tional lifestyles may explain the different effects of an antiviolence interven- tion program. Research design Based on relational and lifestyles theory, the authors designed a randomized controlled trial and applied an analysis of variance 2 (treatment: experimental vs. control) � 4 (lifestyle classes emerged from data using latent class analysis: communicative vs. autono- mous vs. meta-reflexive vs. fractured). Subjects Seven hundred and thirty- five Italian students distributed in 47 classes participated anonymously in the research (51.3% females). Measures Participants completed a lifestyle questionnaire as well as their attitudes and behavioral intentions as the dependent measures. Results The results indicated that the program was effective in changing adolescents’ attitudes toward violence. However, behavioral intentions toward consumption of violent video games were moderated by lifestyles. Those with communicative relational lifestyles showed fewer intentions to consume violent video games, while a boom- erang effect was found among participants with problematic lifestyles. Conclusion Adolescents’ lifestyles played an important role in influencing the effectiveness of an intervention aimed at changing behavioral intentions toward the consumption of violent video games. For that reason, audience lifestyle segmentation analysis should be considered an essential technique for designing, evaluating, and improving media literacy programs.

Keywords adolescents, media literacy, lifestyles, antiviolence attitudes, violence, video games

Violence is an obstacle to positive youth development (PYD) and a major

public health issue (WHO, 1996) with long-term negative social repercussions

(WHO, 2008). For instance, United Nations International Children’s Emer-

gency Fund (2012) has reported that a high proportion of teenagers are affected

by physical violence and bullying. In developed countries, school violence has

become a prevalent social problem: In the United States, teenage students were

the victims of approximately 828,000 nonfatal assaults at school in 2010 (Kim

& Brown, 2014). Moreover, violent experiences during adolescence are asso-

ciated with severe social, emotional, cognitive, and physical damage and dis-

turbance in adulthood that can sometimes lead to substance abuse (Chapman

et al., 2011). Several studies have demonstrated that one of the factors

Rivera et al. 143

associated with violence among adolescents is exposure to media violence

(Anderson et al., 2010; Möller, Krahé, Busching, & Krause, 2012).

Considering the relevance of the issue and the scientific evidence, both

researchers (Möller et al., 2012; Patton et al., 2014) and policy makers (Bailey,

2011) have suggested the importance of implementing awareness-raising and

educational media literacy programs for reducing violent media consumption

and promoting negative attitudes toward deviant role models. These types of

strategies may help to facilitate PYD. Although several prevention strategies

have been implemented (e.g., Daphne III and Safer Internet Programs in the

European Union, violence prevention campaigns of the Centers for Disease

Control and Prevention in United States), very little is known about how these

media literacy school-based interventions affect behavioral change (Möller

et al., 2012; Smith, Schneider, Smith, & Ananiadou, 2004). Sometimes per-

suasive communication campaigns can be effective (e.g., Farrelly, Pechacek,

Thomas, & Nelson, 2008; Flynn et al., 2007). However, sometimes they can

be ineffective (e.g., Foxcroft, Lister-Sharp, & Lowe, 1997; Rhodes, Roskos-

Ewoldsen, Edison, & Bradford, 2008) or even backfire, resulting in the oppo-

site effect of what was expected, a phenomenon known as ‘‘boomerang

effect’’ (Brändle, Cárdaba, & Ruiz-San Román, 2011; Cárdaba, Briñol, Brän-

dle, & Ruiz San Román, 2016; Hart, 2013). To make a contribution to the

fields of PYD and adolescent interpersonal violence prevention strategies and

policies, the present study evaluated the efficacy of an Italian school-based

intervention program that focused on violent media consumption (Kirsh,

2010) using an a posteriori marketing research tool: audience segmentation

based on relational lifestyle variables. The novelty of this study resides in its

cross-disciplinary perspective and use of audience segmentation to evaluate

the effectiveness of a media literacy intervention. Drawing from an ecological

model (Livingstone, Haddon, & Görzig, 2012) that shows the importance of

social relationship variables, this research identified adolescent profiles that

should be considered for use in fine-tuning violence prevention messages. The

results of this study can therefore be used to provide researchers, decision

makers, and practitioners, who work with and for adolescents, with evidence

to develop effective, targeted campaigns, thereby facilitating adolescent

empowerment, healthy lifestyles, and positive development.

Adolescents Lifestyles in the Context of Reflexive Modernity

Lifestyles could be defined as a complex and dynamic integrated system of

preferences and attitudes, influenced by the socialization process, and

144 Evaluation Review 40(2)

translated into social routine decisions and actions that can be operationa-

lized in terms of a complex system of behaviors, orientations, resources, and

knowledge structures developed through experience that express personal

and social identity (Faggiano, 2007; Thirlaway & Upton, 2009).

Through exploratory qualitative studies on the relationships of young

people with their families and friends, Archer (2003, 2012) identified four

theoretical types of youth lifestyles based on adolescents’ different reflex-

ivity or ‘‘the regular exercise of the mental ability, shared by all normal

people, to consider themselves in relation to their (social) contexts and vice

versa’’ (Archer, 2012, p. 1): (1) communicative reflexives (young people

whose inner conversations and decisions include positive interactions with

the family and other agencies), (2) autonomous reflexives (adolescents who

make decisions in solitude, without too much parental support), (3) meta-

reflexives (with conflicted family relationships and frequent internal con-

versations: They do not accept messages channeled by agencies without

reflection), and (4) fractured reflexives (passive teenagers who, due to the

poor quality of their relationships, have difficulty both in conducting fruit-

ful conversations with others and purposeful internal conversations).

Research on consumer behavior (Garcia Ruiz, 2009) has shown that these

types of reflexivity can be used in an audience analysis with segmentation

purposes since they explain the reason why relational lifestyles have a huge

influence on the consumption decisions of young people.

Enabled by a marketing technique like lifestyle segmentation (Kahle &

Chiagouris, 2014), lifestyle theory (for a review, see Faggiano, 2007)

could provide useful insights for analyzing the effectiveness of media

literacy interventions. Interpersonal relationships might be effective in

moderating exposure to violent media. Market segmentation is a method

used for classifying individuals, on the basis of key discriminant or criter-

ion variables, into homogeneous segments that share relevant conditions

for an outcome of interest (McDonald & Dunbar, 2004). Audience seg-

mentation is vital for designing and evaluating social marketing strategies

(Briñol & Petty, 2015; Kotler & Lee, 2012). Although some studies have

made lifestyle analyses (Kahle & Chiagouris, 2014; Mathijssen, Janssen,

Bon-Martens, & Goor, 2012), few of them have included social relation-

ship variables, despite their influence in the social, emotional, and per-

sonality development of children (Ispa et al., 2013; Stacy, Newcomb, &

Bentler, 1991).

In line with previous literature, and based on a previous study of the

effectiveness of a segmentation strategy of alcohol users (Mathijssen et al.,

2012), we propose to cluster the adolescents who participated in a

Rivera et al. 145

school-based media literacy intervention (‘‘Stop Violence on Social

Media’’ project; for a review of the model underlying the intervention, see

Appendix) on the basis of their lifestyles in order to analyze its effective-

ness. Specifically, adolescents who have less experience making decisions

based on positive interactions and exchange of ideas with others (mainly

parents and other role models) may be less persuaded by an antiviolence

school-based intervention because they might have problems accepting

messages that come from external sources. The media literacy intervention,

being oriented toward behavior change, requires certain social competences

developed through an active process of socialization. In contrast, those

adolescents who can count on supportive role models to follow and help

develop healthy behaviors and habits should be more persuaded by the

program than the other groups of adolescents.

Objectives

The present research examined whether or not different relational lifestyles

may explain the different effects of a media literacy intervention program.

To examine the efficacy of the ‘‘Stop Violence on Social Media’’ project,

we analyzed attitudes and behavioral intentions toward violence and violent

video games shown by a sample of Italian middle and high school students

who participated in the program compared to a control group. When exam-

ining the latent structure of resulting data, and without using a predefined

classification strategy, participants were segmented on the basis of their

relational lifestyles. We hypothesized that some adolescents would be more

easily persuaded by the media literacy intervention than others depending

on their specific relational lifestyle. Therefore, we tested two hypotheses:

Hypothesis 1: The ‘‘Stop Violence on Social Media’’ project is a

useful tool to change teenagers’ attitudes toward violence.

Hypothesis 2: Adolescent relational lifestyles may moderate the effi-

cacy of the media literacy program although the specific direction is

not predicted.

Method

Participants and Design

Seven hundred and thirty-five (735) Italian students distributed in 47 classes

participated anonymously in the present research (48.7% males and 51.3%

146 Evaluation Review 40(2)

females). Schools were contacted and objectives for the intervention were

explained in detail. From those schools that agreed to participate (four

schools), informed consent was obtained both from schools’ authorities and

students’ parents. Ages ranged from 12 to 19 (Mage ¼ 14.04, SD ¼ 1.04). The 47 classrooms were randomly assigned either to an intervention pro-

gram or to a control group, that is, all the students from the same classroom

were assigned to one of the two conditions. Given that the randomization

procedure was done at the class level, standard errors are clustered on each

class. All participants were seated in front of a computer and completed the

questionnaire online by reporting their beliefs, attitudes, behavioral inten-

tions, and habits, so they could be classified later into one of four lifestyles

(i.e., communicative, autonomous, meta-reflexive, and fractured; Archer,

2012; Garcia Ruiz, 2009). The design was an analysis of variance

(ANOVA) 2 (treatment: experimental vs. control) � 4 (lifestyles: commu- nicative vs. autonomous vs. meta-reflexive vs. fractured). For the first time,

this research evaluated the impact of the ‘‘Safe Social Media’’ project

partially funded by the European Commission within the Daphne III

program.

Procedure

As part of a general program about lifestyles, participants were exposed

simultaneously to a program designed to promote antiviolence attitudes and

behaviors. The program focused on violence in the media. After receiving

the program, half of the participants reported their attitudes toward violence

and their behavioral intentions of consuming violent video games in the

next few days (experimental group). The other half of the participants

(control group) completed these same measures before receiving the inter-

vention. As mentioned above, all of the participants reported their beliefs,

attitudes, behavioral intentions, and habits and were classified into one of

four lifestyles (i.e., communicative, meta-reflexive, autonomous, and frac-

tured; see Appendix for a full description of independent and dependent

variables).

Independent Variables

Treatment. All participants received the media literacy program simultane- ously, but classes were randomly assigned to complete the dependent mea-

sures immediately before the program (control treatment group) or to

complete the dependent measures immediately after the program

Rivera et al. 147

(experimental treatment group). The change in the timing of the completion

of the dependent variable allows us to test the efficacy of the intervention

while, at the same time, permits that all the students benefit from receiving

the intervention unlike other research that employs a waiting-list procedure.

Lifestyles. To measure the different lifestyles of the participants, we selected attitudes, behaviors, and values that were more applicable to adolescents’

ways of living (Faggiano, 2007; Rivera & Santos, 2016). Based on the

above theoretical background, 38 items were retained from the total pool

of items in the questionnaire.

Dependent Variables

Attitudes toward violence. Participants’ attitudes toward violence were assessed using different prompts that asked about their perception of vio-

lence in terms of favorability.

Behavioral intentions toward consumption of violent video games. All the parti- cipants were asked about their behavioral intentions to consume violent

video games both in the near and distant future in terms of probability.

Results

Lifestyles

The 38 items were submitted to a latent class analysis (LCA). Regarding

theoretical and empirical concerns, we selected a four-class solution.

Although the most common model fit index is the w2, we could not compute it because there were too many cells where the observed frequency was

small (or zero). The model fit for this analysis met satisfactory fit based on

its entropy: The entropy reached in this analysis was .90, indicating high

certainty in classification. To assure that we selected the correct number of

classes, we compared different models using the Bayesian information

criterion (BIC) adjusted values: the four-class solution (BIC4 ¼ 51,048.14) fitted better than the three-class solution (BIC3 ¼ 68,151.03) and better than the five-class solution (BIC5 ¼ 68,867.93). We chose the model for which the adjusted BIC value was the lowest. The first class was

‘‘communicative lifestyle’’ (16.3%), the second class was ‘‘fractured life- style’’ (27.3%), the third was ‘‘autonomous lifestyle’’ (29.5%), and the fourth was ‘‘meta-reflexive lifestyle’’ (26.8%). The resulting four classes differed from each other in terms of critical variables. Figure 1 shows the

148 Evaluation Review 40(2)

probability distribution of some of the more critical variables in the LCA.

As we can see, there are differences between lifestyle classes. For example,

there was a higher probability that ‘‘communicative’’ adolescents would

engage in dialogue with parents about their feelings and seek the support of

their parents. ‘‘Communicative’’ adolescents were also more likely to report

a desire to do good things for their neighborhoods (i.e., civic values). In

addition, ‘‘communicative’’ adolescents had a lower probability for taking

drugs compared to fractured adolescents. The ‘‘fractured’’ adolescents had a

higher probability of taking drugs with respect to the other three classes

taken together and a lower probability both for doing good things for their

neighborhoods and for having respect for authority. The ‘‘autonomous’’

adolescents had a higher probability of respect for authority and the lowest

probability to consume drugs. The ‘‘meta-reflexive’’ adolescents had a

lower probability of seeking parents’ support compared to the other three

lifestyles.

Attitudes Toward Violence

We submitted the ‘‘attitude toward violence’’ index to an ANOVA 2 � 4. As expected (Hypothesis 1), this analysis revealed a main effect of treat-

ment: Participants who received the program rated violence as more

Figure 1. Probabilities for the maximum category (5) of some of the variables in the latent class analysis.

Rivera et al. 149

unfavorable (M ¼ 2.24, SD ¼ 1.51) than those who were in the control group (M ¼ 2.63, SD ¼ 1.85), F(1, 713) ¼ 11.574, p ¼ .001, Z2 ¼ .016. There was no effect of lifestyles, F(3, 713) ¼ 1.344, p ¼ .26, Z2 ¼ .006. The two-way interaction between treatment and lifestyle was nonsignificant,

F(3, 713) ¼ 1.780, p ¼ .15, Z2 ¼ .007.

Behavioral Intentions Toward Consumption of Violent Video Games

We submitted ‘‘behavioral intentions toward consumption of violent video

games’’ to an ANOVA 2 � 4. There was no effect of treatment, F(1, 722) ¼ .769, p ¼ .38, Z2 ¼ .001, nor lifestyles, F(3, 722) ¼ .040, p ¼ .99, Z2 < .001. As illustrated in Figure 2, however, the resulting two-way interaction

between lifestyles and treatment was marginally significant, F(3, 722) ¼ 2.311, p ¼ .07, Z2 ¼ .01, thus indicating that the treatment effectiveness was marginally different depending on the lifestyle (see Table 1 for mean,

SD, and cell sample sizes). We can decompose this interaction by compar-

ing lifestyle sectors between treatment groups. For the ‘‘communicative’’

type, those who were in the experimental group had fewer intentions to

consume violent video games (M ¼ 4.41, SD ¼ 1.93) than those who were in the control group (M ¼ 5.09, SD ¼ 2.12), F(1, 722) ¼ 3.248, p ¼ .07, Z2 ¼ .004, suggesting that the media literacy program was effective for them. For ‘‘fractured’’ types, we obtained the opposite pattern of results:

Those who were in the experimental group had more intentions to consume

violent video games (M ¼ 4.94, SD ¼ 1.90) than those who were in the

Figure 2. Behavioral intentions toward consumption of violent video games as a function of treatment and lifestyles.

150 Evaluation Review 40(2)

control group (M ¼ 4.46, SD ¼ 2.24), F(1, 722) ¼ 2.767, p ¼ .09, Z2 ¼ .004, suggesting a boomerang effect. For ‘‘autonomous’’ and ‘‘meta-reflexive’’ adolescents, we found a null effect. For ‘‘autonomous’’

type, no differences were found between experimental (M ¼ 4.52, SD ¼ 1.96) and control group (M ¼ 4.82, SD ¼ 1.99), F(1, 722) ¼ 1.178, p ¼ .28, Z2 ¼ .002. For ‘‘meta-reflexive’’ type, the difference between experimental (M ¼ 4.68, SD ¼ 2.04) and control group (M ¼ 4.73, SD ¼ 2.09) was also not significant, F(1, 722) ¼ .029, p ¼ .86, Z2 < .001.

Conclusion and Discussion

This research focused on relational lifestyles as a determinant in predicting

the effectiveness of a media literacy intervention. The results showed that

adolescents’ attitudes toward violence can be changed after participating in

an intervention program that promotes antiviolence attitudes and critical

media consumption, regardless of adolescents’ lifestyles. Those partici-

pants in the safe social media intervention were, on average, more likely

Table 1. Mean, Standard Deviation, and Cells Sample Size.

Lifestyle Treatment

Attitudes Toward Violence

Behavioral Intentions To Consume Violent Video

Games

nMean SD Mean SD

Communicative Control 2.740 1.942 5.089 2.117 64 Experimental 1.830 1.267 4.414 1.932 54 Total 2.319 1.719 4.780 2.054 118

Fractured Control 2.786 2.027 4.460 2.236 116 Experimental 2.234 1.541 4.941 1.902 85 Total 2.550 1.851 4.663 2.110 201

Autonomous Control 2.346 1.584 4.821 1.988 97 Experimental 2.208 1.507 4.520 1.956 118 Total 2.271 1.540 4.656 1.972 215

Meta-reflexive Control 2.634 1.806 4.732 1.955 87 Experimental 2.474 1.558 4.682 2.044 109 Total 2.545 1.670 4.704 1.999 196

Total Control 2.624 1.850 4.732 2.088 364 Experimental 2.236 1.505 4.650 1.967 366 Total 2.430 1.670 4.691 2.028 730

Rivera et al. 151

to form unfavorable attitudes toward violence compared to those in the

control group (Hypothesis 1).

With regard to the second hypothesis of the study, we only found tenta-

tive evidence supporting the idea that adolescents’ relational lifestyles

might influence the effectiveness of the media literacy intervention aimed

at changing behavioral intentions toward the consumption of violent video

games. Specifically, for those adolescents with a ‘‘communicative’’ life-

style, the program seemed effective in changing their intentions to consume

violent video games, but we need to point out that this evidence is, at best,

partial. That is, the ‘‘communicative’’ adolescents had fewer intentions to

consume violent video games after receiving the media literacy intervention

compared to those who were in the control group. However, for those

adolescents with a ‘‘fractured’’ lifestyle, the program seemed counterpro-

ductive in changing their behavioral intentions. That is, the ‘‘fractured’’

adolescents had higher intentions to consume violent video games after

receiving the media literacy intervention compared to those who were in

the control group, resulting in an undesired ‘‘boomerang effect.’’ For the

other two lifestyle segments (‘‘autonomous’’ and ‘‘meta-reflexive’’ types),

there was no effect on the intervention.

These results may suggest that further projects and studies are needed to

consider the complexity that exists when evaluating the effectiveness of

communication campaigns and media literacy programs. Campaign evalua-

tion is a complex task because, as this study shows, different results can be

obtained depending on the dependent variable that is analyzed. In the pres-

ent study, the media literacy program seemed effective when analyzing

attitudes toward violence but less effective when analyzing behavioral

intentions toward violent video games. Moreover, the effectiveness of an

intervention can be difficult to evaluate because even the same dependent

variable can be affected by other moderating variables. Including the rela-

tional lifestyles as predictors allowed this study to show some evidence

suggesting that the effect of treatment on behavioral intentions depended

partially on the type of lifestyle of the individual who received the

treatment.

It is not easy to know why some relational lifestyles (i.e., communica-

tive) facilitate the effectiveness of the intervention and others (i.e., frac-

tured) lead to a boomerang effect. Although some authors have proposed to

introduce lifestyles based on different reflexivity in consumption studies

(Garcia Ruiz, 2009; Garcia Ruiz & Rodriguez-Lluema, 2010), there is no

evidence about their influence on the way of reacting toward an interven-

tion. We did not have any strong a priori prediction regarding which

152 Evaluation Review 40(2)

segment would react positively or negatively to the intervention, but there

could be some post hoc explanations. It might be effective for those life-

styles characterized by good positive interactions with the family and

friends, while the contrary could be hypothesized for lifestyles character-

ized by negative interactions and drug consumption. For example, commu-

nicative adolescents—who can count on supportive role models to follow

and develop healthy behaviors and habits—may be more open to messages

that reinforce their own lifestyle, whereas fractured adolescents—who lack

a positive social network and have habits and behavior that hinder the

development of productive relationships—might react negatively to mes-

sages that threaten their own lifestyle.

Thus, taking into account relational lifestyles as a possible moderator

could be critical in creating effective media literacy interventions. Our

results suggest that not every individual should receive the same kind of

intervention because the intervention might have undesired and counter-

productive consequences for some individuals. To maximize the effective-

ness of antiviolence media literacy programs, communicative, fractured,

autonomous, and meta-reflexive adolescents might need to receive different

types of interventions (Briñol & Petty, 2006, 2009; Kahle & Chiagouris,

2014).

Therefore, this study provides new evidence suggesting that segmenta-

tion of the target audience could be a useful technique to reduce violence by

tailoring the type of message with important variables of the adolescents

such as their specific relational lifestyles. However, more research on the

role of lifestyles relational dimensions (experiences, social identity, sub-

jective culture, and motivations) in consumption patterns is needed (Living-

stone, Mascheroni, & Staksrud, 2015).

Of course, to determine the best strategy to avoid these undesirable

consequences of these interventions, more research is needed on the psy-

chological mechanisms underlying the assimilation and contrast effects of

different lifestyles. For example, if fractured adolescents reject the inter-

vention because they feel self-threatened by the message, then a self-

affirmation strategy (i.e., participants thinking about what values they

consider personally important) before delivery of the intervention should

increase its effectiveness (Briñol, Petty, Gallardo, & DeMarree, 2007;

Briñol, Petty, & Wagner, 2009; Cohen, Aronson, & Steele, 2000; Correll,

Spencer, & Zanna, 2004).

One important limitation of the present study is the absence of an active

control group in addition to the passive form used here. For instance, the

adolescents in the active control group should have attended a neutral

Rivera et al. 153

intervention outside the area of violence control that promoted habits such

as recycling to allow them to be involved in a program of a similar duration

to that of the experimental group. This would have permitted us to rule out a

number of alternative explanations for the resulting effects of the study

(e.g., passage of time, feelings of being special, feelings of being mis-

treated, etc.). Despite this, we chose a passive control group because access

to school was limited for a short period of time, making it difficult to offer

two different intervention programs.

Another limitation is the duration of the intervention. The intervention

was only 2 hrs for the same reason mentioned above (i.e., access to school

was limited for a short period of time). We might think that interventions

with longer durations should result in different outcomes; therefore, future

research with longer interventions than the one applied in this research is

needed.

In closing, although the results of this research does not allow to propose

a definitive typology for audience targeting, we think that our moderation-

by-lifestyles approach is useful for those who work with and for adolescents

and try to develop effective targeted programs and campaigns. First, as

previous segmentation research showed, this study presented some sugges-

tive evidence that an intervention program may not be equally effective for

all the adolescents. Usually, segmentation studies only take into account

variables at the individual level (e.g., personality traits or sociodemographic

characteristics) to divide groups (Ispa et al., 2013; Stacy et al., 1991).

Instead, we used multidimensionality variables such as lifestyles that

include variables at individual, structural, and societal levels (e.g., habits

of consumption, relationships with the community, friends, and parents,

etc.). Thus, in this study, we extended these individual factors to a new,

integrative concept such as the relational lifestyles of adolescents.

From a practical standpoint, our results suggest that lifestyle segmen-

tation and targeting strategies might benefit from considering not only

individual attitudes, interests, opinions, and values included in models

commonly used by social marketing and other disciplines but also more

complex factors like lifestyles developed through interpersonal relation-

ships. In consequence, the results of this study may be useful for the

development of new holistic segmentation strategies and refinements of

the instruments applied for their evaluation. Thus, future research should

explore new scales and questionnaires, as well as qualitative techniques,

to refine the clustering on lifestyles and make a clearer contribution for the

moderating role of lifestyles on the effectiveness of interventions and

social campaigns.

154 Evaluation Review 40(2)

Appendix

Variables Description

Independent variables Treatment. To avoid information contamination, all the participants

received the media literacy intervention simultaneously, but the experimen-

tal group completed the dependent measure after and the control group

before the intervention. This procedure permitted us to test whether the

intervention program was effective in changing antiviolence attitudes and

related behavioral intentions to consume violent video games. The inter-

vention program was designed to increase the adolescents’ awareness of the

negative consequences of the presence of violence in the media (i.e., TV

shows, films, and video games) and to change their violence consumption-

related attitudes and behaviors. The program consisted of three sessions,

each of 40 min duration. Every session was adapted to the motivations,

previous knowledge, and language of the adolescents so as to engage them

in listening to the speaker. Each session of the program was started by

providing the participants with a story about situations in which an adoles-

cent would need to make some decisions about consuming violence in the

media. The idea of these narratives was to make the participants identify

with the main character of the stories. The structure of the sessions was

composed of three modules based on media, TV, and video games with a

special emphasis on the latter, which was the focus of our research. The

theoretical model underlying this intervention is based on the recommen-

dations of Möller, Krahé, Busching, and Krause (2012), who recommend

the use of social learning theory (Bandura, 1977) and script theory (Hues-

mann, 1998). Both theories explain the processes by which exposure to

media violence leads to aggressive behavior and indicate how detrimental

effects of this exposure might be diminished. On the one hand, exposure to

media violence may trigger a mechanism of observational learning in which

aggressive behavior is developed through imitation. On the other hand,

script theory links observational learning from violent media characters

to the development of aggressive scripts.

Lifestyles. Thirty-eight items were used to assess participants’ lifestyles. The item-response format ranged from 1 ‘‘never’’ to 5 ‘‘very much.’’ Some

examples of these items were ‘‘When I hang out with my friends, I consume

alcohol’’ or ‘‘I smoke cigarettes,’’ ‘‘I have attended cultural activities such

as going to museums and theatres,’’ ‘‘It is important for me to do things to

improve my town or community,’’ ‘‘My parents know where I am going or

Rivera et al. 155

what I am doing during my leisure time,’’ ‘‘In my peer group, I can give my

opinion without fear because others will respect me,’’ and ‘‘I have talked to

my parents about drug and alcohol abuse.’’ 1

Then, in order to explore the

lifestyle sectors in this population, we carried out a LCA using the statistical

software Mplus, Version 4 (Muthén & Asparouhov, 2002). The number of

latent classes is unknown and cannot be directly estimated a priori. To

identify the model with the optimal number of classes, various models with

different numbers of latent classes must be estimated and compared with

each other (Wang & Wang, 2012). In this analysis, a cluster solution was

determined according to the following criteria. Firstly, the overall fit of the

tested latent class model was analyzed using the entropy, which indicated

the certainty in the classification. Secondly, the most economical model

was selected from those that fitted, that is, the model for which the BIC

(adjusted value) presents the lowest value. As a result of this analysis, we

were able to identify four different lifestyle factors. In order to classify these

lifestyles, we used both theoretical and empirical strategies. In order to

determine the salient characteristics of these four lifestyle sectors, we used

the probabilities of each category in each item. We based our views on

previous literature consistent with our findings (Faggiano, 2007; Garcı́a

Ruiz, 2009) that also suggested four lifestyle categories: communicative,

autonomous, meta-reflexive, and fractured. As a result of this process, we

were able to name the lifestyle sectors in a manner consistent both with

previous literature and with the findings from the LCA.

Dependent variables Attitudes toward violence. Participants were asked ‘‘how do you evaluate

violence?’’ in a series of 11 9-point semantic differential scales (i.e., bad–

good, dislikeable–likeable, not acceptable–acceptable, not recommended–

recommended, useless–useful, unpleasant–pleasant, negative–positive,

inconvenient–convenient, harmful–beneficial, undesirable–desirable, and

unnecessary–necessary). Ratings were highly intercorrelated (a ¼ .96), so they could be averaged to create a composite attitude index. Higher values

on this index indicated evaluations more favorable toward violence.

Behavioral intentions toward consumption of violent video games. This vari- able was measured using three 9-point scales anchored at ‘‘completely dis-

agree’’ to ‘‘completely agree.’’ The examples were ‘‘In the future, I would

like to play violent video games to a lesser degree,’’ ‘‘When I grow up, I will

forbid my children to play violent video games,’’ and ‘‘In the future, I would

like to play video games that teach values to the gamers.’’ These items were

156 Evaluation Review 40(2)

intercorrelated (a ¼ .69), therefore they were averaged to create a compo- site index of intentions to consume violent video games. We reversed these

items to allow for greater consistency between this dependent measure and

the ‘‘attitudes toward violence’’ index. Thus, higher values on this index

indicated more intentions to consume violent video games. We also

included an item about civic video games as a proxy to measure the inten-

tions of participants to consume violent video games plus their intentions to

refuse usage of antiviolence video games. The result of reversing the item

was that those who rated higher in it indicated greater intentions to consume

a larger proportion of violent video games.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research,

authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research,

authorship, and/or publication of this article: This work was supported by the Eur-

opean Union (Grant number JUST/2010/DAP3/AG/1111-30-CE-0397890/00-02)

and Intermedia Consulting A.C. (Grant number SSM2011-2013).

Note

1. Link to the full distribution of lifestyle items: https://drive.google.com/open?

id¼0B9Cz2PX_2KM3ZWZPbkZMMU8tUEk

References

Anderson, C. A., Shibuya, A., Ihori, N., Swing, E. L., Bushman, B. J., Sakamoto,

A., . . . Saleem, M. (2010). Violent video game effects on aggression, empathy,

and prosocial behavior in Eastern and Western Countries: A meta-analytic

review. Psychological Bulletin, 136, 151–173.

Archer, M. S. (2003). Structure, agency and the internal conversation. Cambridge,

England: Cambridge University Press.

Archer, M. S. (2012). The reflexive imperative in late modernity. Cambridge,

England: Cambridge University Press.

Bailey, R. (2011). Letting children be children: Report of an independent review of

the commercialisation and sexualisation of childhood. London, England: TSO.

Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioural change.

Psychological Review, 84, 191–215.

Rivera et al. 157

Brändle, G., Cárdaba, M. A., y Ruı́z San Román, J. A. (2011). Riesgo de aparición

del efecto boomerang en las comunicaciones contra la violencia. Comunicar,

XIX, 161–168.

Briñol, P., & Petty, R. E. (2006). Fundamental processes leading to attitude change:

Implications for cancer prevention communications. Journal of Communication,

56, 81–104.

Briñol, P., & Petty, R. E. (2009). Source factors in persuasion: A self-validation

approach. European Review of Social Psychology, 20, 49–96.

Briñol, P., & Petty, R. E. (2015). Elaboration and validation processes: Implications

for media attitude change. Media Psychology, 18, 267–291.

Briñol, P., Petty, R. E., Gallardo, I., & DeMarree, K. G. (2007). The effect of self-

affirmation in non threatening persuasion domains: Timing affects the process.

Personality and Social Psychology Bulletin, 33, 1533–1546.

Briñol, P., Petty, R. E., & Wagner, B. (2009). Body postures effects on self-

evaluation: A self-validation approach. European Journal of Social

Psychology, 39, 1053–1064.

Cárdaba, M. A., Briñol, P., Brändle, G., & Ruiz-SanRomán, J. A. (2016). The

moderating role of aggressiveness in response to campaigns and interventions

promoting anti-violence attitudes. Aggressive Behavior, 42, 471–482.

Chapman, D. P., Wheaton, A. G., Anda, R. F., Croft, J. B., Edwards, V. J., Liu,

Y., . . . Perry, G. S. (2011). Adverse childhood experiences and sleep distur-

bances in adults. Sleep Medicine, 12, 773–779.

Cohen, G. L., Aronson, J., & Steele, C. M. (2000). When beliefs yield to evidence:

Reducing biased evaluation by affirming the self. Personality and Social

Psychology Bulletin, 26, 1151–1164.

Correll, J., Spencer, S. J., & Zanna, M. P. (2004). An affirmed self and an open

mind: Self-affirmation and sensitivity to argument strength. Journal of

Experimental Social Psychology, 40, 350–356.

Faggiano, M. P. (2007). Stile Di Vita E Partecipazione Sociale Giovanile: Il Circolo

Virtuoso Teoria-Ricerca-Teoria. Milano, Italy: FrancoAngeli.

Farrelly, M. C., Pechacek, T. F., Thomas, K. Y., & Nelson, D. (2008). The impact of

tobacco control programs on adult smoking. American Journal of Public Health,

98, 304–309.

Flynn, B. S., Worden, J. K., Bunn, J. Y., Dorwaldt, A. L., Connolly, S. W., &

Ashikaga, T. (2007). Youth audience segmentation strategies for smoking-

prevention mass media campaigns based on message appeal. Health Education

& Behavior, 34, 578–593.

Foxcroft, D. R., Lister-Sharp, D. J., & Lowe, G. (1997). Alcohol misuse prevention

for young people: A systematic review reveals methodological concerns and lack

of reliable evidence of effectiveness. Addiction, 92, 531–537.

158 Evaluation Review 40(2)

Garcia Ruiz, P. (2009). El concepto de ‘‘reflexividad’’ en la sociologı́a del consumo:

algunas propuestas. RES. Revista Española de Sociologı́a, 12, 87–102.

Garcia Ruiz, P., & Rodriguez-Lluesma, C. (2010). ‘Reflexive consumers’: A rela-

tional approach to consumption as a social practice. In M. Archer, Conversations

about reflexivity (pp. 223–242). New York, NY: Routledge.

Hart, P. S. (2013). 18 Boomerang effects in risk communication. In J. Arvai & L.

Rivers III (Eds.), Effective risk communication (pp. 304–334). New York, NY:

Routledge.

Huesmann, L. R. (1998). The role of social information processing and cognitive

schema in the acquisition and maintenance of habitual aggressive behavior. In R.

G. Geen & E. Donnerstein (Eds.). Human aggression: Theories, research, and

implications for social policy (pp. 73–109). San Diego, CA: Academic Press.

Ispa, J. M., Csizmadia, A., Rudy, D., Fine, M. A., Krull, J. L., Bradley, R. H., &

Cabrera, N. (2013). Patterns of maternal directiveness by ethnicity among early

head start research participants. Parenting, 13, 58–75.

Kahle, L. R., & Chiagouris, L. (Eds.). (2014). Values, lifestyles, and psycho-

graphics. New York, NY: Psychology Press.

Kim, J. S., & Brown, S. M. (2014). School violence. In E. J. Mullen (Ed.), Oxford

bibliographies in social work. New York, NY: Oxford University Press. doi:10.

1093/obo/9780195389678-0179

Kirsh, S. (2010). Media and youth: A developmental perspective. Malden, MA:

Wiley-Blackwell.

Kotler, P., & Lee, N. (2012). Social marketing: Influencing behaviors for good. Los

Angeles, CA: Sage.

Livingstone, S., Haddon, L., & Görzig, A. (2012). Children, risk and safety on the

Internet: Research and policy challenges in comparative perspective. Bristol,

England: The Policy Press.

Livingstone, S., Mascheroni, G., & Staksrud, E. (2015). Developing a framework for

researching children’s online risks and opportunities in Europe. London,

England: EU Kids Online.

Mathijssen, J., Janssen, M., van Bon-Martens, M., & van de Goor, I. (2012).

Adolescents and alcohol: An explorative audience segmentation analysis.

BMC Public Health, 12. Retrieved from http://www.biomedcentral.com/1471-

2458/12/742

McDonald, M., & Dunbar, I. (2004). Market segmentation. West Sussex, England:

John Wiley.

Möller, I., Krahé, B., Busching, R., & Krause, C. (2012). Efficacy of an interven-

tion to reduce the use of media violence and aggression: An experimental

evaluation with adolescents in Germany. Journal of Youth and Adolescence,

41, 105–120.

Rivera et al. 159

Muthén, B., & Asparouhov, T. (2002). Latent variable analysis with categorical

outcomes: Multiple-group and growth modeling in Mplus. Mplus Web Notes,

4, 1–22.

Patton, D. U., Hong, J. S., Ranney, M., Patel, S., Kelley, C., Eschmann, R., &

Washington, T. (2014). Social media as a vector for youth violence: A review

of the literature. Computers in Human Behavior, 35, 548–553.

Rhodes, N., Roskos-Ewoldsen, D. R., Edison, A., & Bradford, M. B. (2008).

Attitude and norm accessibility affect processing of anti-smoking messages.

Health Psychology, 27, S224–S232.

Rivera, R., & Santos, D. (2016). Civic and political participation of children and

adolescents: A lifestyle analysis for positive youth developmental programs.

Children & Society, 30, 59–70.

Smith, J. D., Schneider, B. H., Smith, P. K., & Ananiadou, K. (2004). The effec-

tiveness of whole-school antibullying programs: A synthesis of evaluation

research. School Psychology Review, 33, 547–560.

Stacy, A. W., Newcomb, M. D., & Bentler, P. M. (1991). Social psychological

influences on sensation seeking from adolescence to adulthood. Personality and

Social Psychology Bulletin, 17, 701–708.

Thirlaway, K., & Upton, D. (2009). The psychology of lifestyle: Promoting healthy

behaviour. London, England: Taylor & Francis.

United Nations International Children’s Emergency Fund. (2012, April). Progress

for children. A report card on adolescents. Retrieved from http://www.unicef

.org/media/files/PFC2012_A_report_card_on_adolescents.pdf

World Health Organization. (1996). Violence: A public health priority. Geneva,

Switzerland: Author.

World Health Organization. (2008). Manual for estimating the economic costs

of injuries due to interpersonal and self-directed violence. Retrieved from

http://www.who.int/violence_injury_prevention/violence/activities/econom

ics/en/

Author Biographies

Reynaldo Rivera, PhD in Communication, University of Navarra, Spain; Msc in

Sociology, University of La Sapienza, Italy, is the director of InterMedia and the

Online Diploma in Human Rights and is the main researcher for Interaxion Project.

David Santos, master’s degree in Statistics, Universidad Complutense de Madrid

(Spain), is at the Psychology Department, Universidad Autónoma de Madrid. His

research interests focus on attitude change and how metacognitive processes influ-

ence subsequent judgments.

160 Evaluation Review 40(2)

Gaspar Brändle, PhD in Sociology, is at the Department of Sociology, Universidad

de Murcia, Spain. His research focuses on the study of communication processes,

mainly oriented to understand underlying causes and consequences of media vio-

lence on aggressive or antisocial behavior.

Miguel Ángel M. Cárdaba, PhD in Communication, University Complutense

Madrid, is at the Villanueva University.

Rivera et al. 161