Media violence not linked to youth violence (Ferguson et al., 2009)

Via gamepolitics, Christopher Ferguson has published an article about the influence of violent video games, family, peers and depression.



To examine the multivariate nature of risk factors for youth violence including delinquent peer associations, exposure to domestic violence in the home, family conflict, neighborhood stress, antisocial personality traits, depression level, and exposure to television and video game violence.

Study design

A population of 603 predominantly Hispanic children (ages 10-14 years) and their parents or guardians responded to multiple behavioral measures. Outcomes included aggression and rule-breaking behavior on the Child Behavior Checklist (CBCL), as well as violent and nonviolent criminal activity and bullying behavior.


Delinquent peer influences, antisocial personality traits, depression, and parents/guardians who use psychological abuse in intimate relationships were consistent risk factors for youth violence and aggression. Neighborhood quality, parental use of domestic violence in intimate relationships, and exposure to violent television or video games were not predictive of youth violence and aggression.


Childhood depression, delinquent peer association, and parental use of psychological abuse may be particularly fruitful avenues for future prevention or intervention efforts.

As before, this post is a detailed summary to Ferguson’s succinct summary, but I will be omitting some parts like the authors’ structural equation modeling analysis which is too much.


Participants: 603 Hispanic youths in Texas. Average age is 12.35, ranges from 10 to 14. Equal gender ratio. Why Hispanic youths? I guess this sample is part of another study and they found something interesting among the measures used. However, the authors argued most research were conducted with White individuals and this can be used for comparative purposes.


Negative Life Events: the article doesn’t mention the number of items, but participants answer on a likert-type scale. The measure has several assessment subscales: neighbourhood problems, negative relations with adults, antisocial personality, family attachment, and delinquent peers.

Family Environment Scale: a 90 item true-false measure. It assesses styles of family interaction and communication. The family conflict subscale is used in this study.

Family Violence: Using the Conflict Tactics Scale which assesses good and bad behaviours occurring in marital or dating relationships (i.e. behaviours between spouses or dating partners). Two subscales were used in the study: Physical assaults and psychological aggression. The participants’ primary caregiver fill that one out, answered on a likert-type scale.

Media Violence Questionnaire: Participants are asked to name and rate their three favourite television shows and video games. They rate them on how often they play and its level of violence.

In comparison with the other measures in this study, it is the least accurate. Frankly, I think VG researchers should reconsider the validity of this measure. I administered this questionnaire in one of my projects, participants seem to have a hard time deciding (given our now huge repertoire of television programs and video games) and the answers I get are sometimes ridiculous. For example, one participant wrote down Duck Hunt and played it often… 10 years ago. WTF. Instead of asking their favourite video games, ask what they played in the last six months starting with the most recent or ask them which games they played the most to the least. It’s better that way to measure participants’ video game preferences and if used longitudinally, their fluctuations in taste.

Depression: The withdrawal/depression subscale in the Child Behavior Checklist (CBCL) Youth Self-Report. Answered on a likert-type scale.

Aggression: Used the CBCL, both participants (Youth version) and their primary caregiver (Parent version) fill out their version of the CBCL. The aggression and rule-breaking subscale is examined. Answered on a likert-type scale.

Bullying: The Olweus Bullying questionnaire. Answered on a likert-type scale.

Delinquent Behavior: Assessed through a subscale on the Negative Life Events questionnaire. Answered on a likert-type scale. It has two categories: nonviolent and violent criminal activities.


Analyses were done through hierarchical multiple regression.

Going by each dependent variable, listing the predictor by strength.

Aggression (child-report) is best predicted by depression and delinquent peers. Other significant factors included are having negative relations with adults, family conflict, anti-social personality, and psychological aggression between spouses or dating partners. When it comes to parent reports, the same predictors were significant, but much weaker. The two highest predictors were psychological aggression and negative relations with adults.

Rule-breaking (child-report): best predicted by delinquent peers and depression. Other significant factors are anti-social personality, negative relations with adults, being male, family conflict and psychological aggression. A negative relation is found for family attachement. Predictors from Parental reports of children’s rule-breaking were the same, but weaker. With the exceptions, family conflict is insignificant and physical assault between spouses or dating partner is significant.

Non-violent crime: only two predictors were found: depression and delinquent peers. However, it must be noted that only 92 children (15.4% of the sample) reported of committing a non-violent crime.

Violent crime: Delinquent peers is the only significant predictor. 74 children (12.3%) reported to have engaged in a violent crime.

Bullying behaviours: Strongest predictors are anti-social personality and delinquent peers. Other factors are psychological aggression, violent video games, family attachment, negative relations with adults, depression, and family conflict. The authors argued that the violent video game effects were negligible.

It should be noted the effect sizes are small, a small sample of relevant factors were examined which underlines the complexity of risk factors for violent and criminal behaviours.

Ferguson, C. J., San Miguel, C., & Hartley, R. D. (in press). A multivariate analysis of youth violence and aggression: The influence of family, peers, depression and media violence. Journal of Pediatrics


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