German adolescents’ media violence use and their teacher-rated behaviours (Krahé & Möller, 2011)

Asuka Shikinami Langley playing her Wonderswan (Neon Genesis Evangelion)

Barbara Krahé (University of Potsdam) and Ingrid Möller (University of Postdam) have published an article in the Journal of Adolescence on teacher-rated aggression of German adolescents.


The relations between adolescents’ habitual usage of media violence and their tendency to engage in aggressive and prosocial behavior in a school setting were examined in a cross-sectional study with 1688 7th and 8th graders in Germany who completed measures of violent media exposure and normative acceptance of aggression. For each participant, ratings of prosocial and aggressive behavior were obtained from their class teacher. Media violence exposure was a unique predictor of teacher-rated aggression even when relevant covariates were considered, and it predicted prosocial behavior over and above gender. Path analyses confirmed a direct positive link from media violence usage to teacher-rated aggression for girls and boys, but no direct negative link to prosocial behavior was found. Indirect pathways were identified to higher aggressive and lower prosocial behavior via the acceptance of aggression as normative. Although there were significant gender differences in media violence exposure, aggression, and prosocial behavior, similar path models were identified for boys and girls.

Besides considerable videogame research in Germany, there is considerable research being conducted in Singapore. There is an MIT-affiliated laboratory in Singapore.


Participants: 1688 German 7th and 8th grade students from high schools in Berlin. Near Equal gender ratio, average age is 13.4 (SD = .90).


Habitual media violence exposure: Participants were asked to rate a list of genres for television, movies and videogames. Examples include action, cartoon/animation, mystery, military strategy, shooters, sports games, hospital series, etc. They rate them on a 5-point frequency scale (e.g. never to very often).

Media experts (i.e. sales persons and journalists) were asked to rate the violent content for the genre lists on a 5-point violence scale. The media experts were given a definition of “violent content” of which would help create a common conceptual ground between the experts. The definition is “the intentional harming of humans, other beings or human-like objects by one or more media characters/players.”

This is a pretty interesting way to ask participants’ media habits based on genres instead of specific shows. Their previous study (2009) used the same questionnaire because their study was a 30+ month longitudinal design, so titles at the beginning of their study would not match quite well with titles at a later date. However, this questionnaire presumed the participants are knowledgeable of the conventions underlying each genre or can categorize their shows, videogames and movies to the genre list the researchers provided. For television and movies, the genres are pretty well established thanks to several decades of cultivating tropes and conventions. As for videogames, there are some agreed-upon genres, but there are some videogames that can be categorized into several genres or cannot be classified at all. The questionnaire has ‘genre mix’ as a separate genre, which its ambiguity would pose some problems in determining the level of violence by the media experts.

They classified each genre into two categories: non-violent versus violent. Any genre that has an aggregated rating of higher than 2 is classified as violent. Over half of the genres in each medium are classified as violent. The correlations between the genres within each medium are significant, as measured by Cronbach’s alpha. The correlations between the violent genres in each medium are significantly correlated with each. So, they combined all genre scores into an aggregated media violence exposure variable for their analysis. So, participants’ frequency rating is multiplied with the average violence rating as a genre-specific violence exposure score. Then, all of the genre scores are averaged into a general media violence exposure score. The same is done for non-violent media exposure.

Normative acceptance of violence: this measure assessed participants’ normative beliefs about aggression. The participants read a hypothetical vignette where they were treated unfairly by another person. The participant was given seven possible reactions to the situation and was asked how acceptable it is for each reaction on a four-point acceptance scale. Two are physical aggression, 2 are verbal aggression, and 3 are relational aggression (e.g. spreading rumours, talking someone behind one’s back).

General media usage: four items about overall media usage like how often they watch tv or DVD, and videogames.

Aggressive and prosocial behaviours: Teachers rated their students on two items. How often such student helps others and how often such student behaves in an aggressive way towards others. The rating is on a 5-point frequency scale. IMO, I would ask how often such student behave in a friendly way towards others as a more equivalent question. Let’s not forget the cultural perceptions of masculinity and femininity, teachers would rate their students from such perceptions. They are more accepting of boys’ aggressiveness than girls and conversely they are more likely to observe and recall girls’ prosocial behaviours more so than boys, therefore they are likely to ignore some of the boys’ aggressive behaviours.

Academic achievement: report cards on three subjects: German, Maths, and English.


The zero-order correlations (i.e. without accounting of any variables) revealed a positive correlation between media violence exposure and teacher-ratings of aggression and a negative correlation for prosocial behaviour. The correlations for aggression are greater for girls (r = .17) than boys (r = .09), but not significantly different. Aggressive belief norm is positively correlated with media violence and aggression and negatively for prosocial. Media violence and non-violent media are positively correlated with each other. However, non-violent media seem to have some complicated correlations with teacher-rated aggression and prosocial behaviours in that the significant correlations differed by gender, but the directions are what you’d expect of non-violent media.

The authors did two stepwise regressions examining where media violence exposure becomes a non-significant variable when facing with multiple predictor variables like gender, grades, aggressive belief norms and such. The first stepwise regression found that media violence is still a significant predictor of teacher-rated aggression after taking into account of gender, grades, prosocial behaviour and aggressive belief norms. The second stepwise regression found that media violence lost its predictive significance for teacher-rated prosocial behaviour after gender and grades were entered into the model.

The authors continued with path analysis which examined the direct, indirect relationships between various factors of their study. In this path analysis, they started with media violence exposure with aggressive beliefs norms in the middle as a mediator/moderator with the teacher-ratings as the outcome/dependent variable.

They found for both genders that media violence exposure has a direct path that increases aggressive belief norms and aggressive behaviour. However, it is aggressive belief norms that have a greater direct impact with both aggressive behaviour (increase) and prosocial behaviour (decrease). The pathways between aggressive and prosocial behaviours are negatively correlated, but are not necessarily causally opposites. So, the path analysis showed that media violence has an indirect influence, as mediated by aggressive normative beliefs, on prosocial and aggressive behaviours.

The authors have some suggestions for future studies. This includes testing prosocial normative beliefs to clarify the relationship between media violence exposure and teacher-rated prosocial behaviour. Including more behavioural measures, like peer-, parent-, self- and observer-ratings of aggression and prosocial behaviours. The authors discussed that media violence may undermine prosocial normative beliefs as they present and legitimize aggression towards others. However, the study is correlational.

After an afternoon nap, I came to realize that this study gives theoretical support for  critical theorists who are on the other side of the communications field. Based on my recent reading of two books [1][2], both authors argued about videogames’ capacity in engaging players in reflective, constructive or cognitive engagement (IMO) to the content or meaning of a videogame. There’s even a book about militarism in videogames. This leads to the thought that violent depiction itself without context is meaningless to the cognitive construction of aggressive normative beliefs. The narrative or ludological aspects seem to deliver a greater influential impact on our beliefs. Some might point to cultivation theory, but I am not knowledgeable in that theory (yet). Musing on that thought, we could say that American culture seem to place violence a greater cultural and positive (maybe marginally) position than sex.

To continue this constructivist stream of consciousness, a media culture that effectively uses violence as a tool for morality would be in a better position to impact the audience’s normative beliefs than a media culture that just bans, minimizes or censors violence or sex because they don’t know how to communicate against its inherent message (i.e. media literacy and procedural literacy). Additionally, the commercial-oriented entertainment cycle between producer and audience would have an impact on how violence is depicted and used as narrative device. For example, how film studios and videogame publishers would merely use sex and violence, but disappoint an audience with a plain and overused story just to gain profits.  If the authors (videogames) can provoke reflectiveness or more concretely mindfulness in the audience, then I guess we would have critical thinking citizens. Think about the Metal Gear Solid series, there are violent and non-violent player actions. The narrative is intricate with intended authorial messages about war, nuclear weapons (see MGS 1), information technology (see MGS 2), and the Cold War (see MGS 3).

Krahé, B., & Möller, I. (2011). Links between self-reported media violence exposure and teacher ratings of aggression and prosocial behavior among german adolescents. Journal of Adolescence , 34 (2), 279-287. DOI:10.1016/j.adolescence.2010.05.003


One thought on “German adolescents’ media violence use and their teacher-rated behaviours (Krahé & Möller, 2011)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s