Empathic numbing from violent media (Bushman & Anderson, 2009)

Via gamepolitics, Brad Bushman and Craig Anderson had conducted two sets of studies that showed how violent media numb individuals from helping others. Here’s the university press release and be sure to listen to the podcast. A little aside, I would like to give my finger to Wiley-Blackwell for poor customer service, again. Why make a press release in February when that paper isn’t going to be published until March? Can’t hold back the excitement?


Two studies tested the hypothesis that exposure to violent media reduces aid offered to people in pain. In Study 1, participants played a violent or nonviolent video game for 20 min. After game play, while completing a lengthy questionnaire, they heard a loud fight, in which one person was injured, outside the lab. Participants who played violent games took longer to help the injured victim, rated the fight as less serious, and were less likely to “hear” the fight in comparison to participants who played nonviolent games. In Study 2, violent- and nonviolent-movie attendees witnessed a young woman with an injured ankle struggle to pick up her crutches outside the theater either before or after the movie. Participants who had just watched a violent movie took longer to help than participants in the other three conditions. The findings from both studies suggest that violent media make people numb to the pain and suffering of others.

Since  they used a classical quasi-experimental paradigm from Bibb latané and John Darley, the bystander effect. My first impression is that they just added their knowledge on aggression and violent media into an existing experimental paradigm and saw the results.

Study 1

Participants: 320 college students

Measures: they used the video game rating sheet to assess the games (if there are differences on various factors like boringness or violent content, etc.), and participants indicated their favourite gaming genre. In addition, a lengthy bogus questionnaire.

Video games used: Carmageddon, Duke Nukem, Mortal Kombat, Future Cop are the violent games. Glider Pro, 3D Pinball, Austin Powers, and Tetra Madness are the non-violent games.


Participants are randomly assigned to one of the games and played for 20 minutes. They are then given the questionnaires and when they are doing the lengthy bogus questionnaire, the experimenter played an audio recording of a staged fight. The recording I believe is to standardize the event across all participants. If it’s not, then you might find one staged fight scene believable and another one not believable making some participant data unsuable or making data analysis a living hell for the poor research coders. In any case, they tested out the staged fight to see if it’s believable and they made some adjustments until it’s believable by all testers. The experimenter timed the participants on how long they would help out and some other detail that I leave out.


In general, there are no differences between participants in the violent (21%) or non-violent (25%) video game condition on whether they decided to help, it was noted that individuals who stated their favourite game (that involves fighting with hands or weapons) were among those who helped the least (11%) versus those whose favourite game is non-violent (26%).

When participants to help, there was a difference in that those in violent took longer: 73 seconds versus 16 seconds. They were less likely to report that they heard a fight (94% versus 99%) and they thought the fight as less serious ( average score of 5.91 versus 6.44 out of a scale of 10).

Looking at the effect size (cohen’s d only), according to wikipedia it’s small with the exception to the time it takes to help.

Study 2

This one is a field experiment and therefore has high external validity (IMO). Now imagine candid camera-style experiment, but this time from a movie theatre where they showed a violent (The Ruins) or non-violent (Nim’s Island) movie and you’ve got an actor whose obviously injured and dropped something and needs help. Do this before watching movie and after the movie. What you get are helping behaviour differences. Go back to the abstract for details of the procedure.

Participants: 162 moviegoers who did not suspect anything. Consent was not required because no personal information was asked of them. So researchers don’t know their age or socioeconomic status, but given that everyone goes to the movies, the experimenters are just doing the experiment with random people. It’s a give and take limitation between uncontrolled experiment versus tightly controlled experiment.

Results: no difference in helping behaviours before moviegoers watched a movie (violent or nonviolent). Personality factors are ruled out. Crowd effects were duly noted. There are differences in helping behaviours after moviegoers watched a movie, those who saw a violent movie took longer to help ( M= 6.89 seconds versus M =5.46 seconds).

That is all. I’ll be making some comments at gamepolitics to address whatever questions commentators may have.

Bushman, B. J., & Anderson, C. A. (2009). Comfortably numb: Desensitizing effects of violent media on helping others. Psychological Science, 20 (3), 273-277. doi: 10.1111/j.1467-9280.2009.02287.x.


8 thoughts on “Empathic numbing from violent media (Bushman & Anderson, 2009)

  1. There are too many differences between individual video games for the first test to be all that valid. Off the top of my head, all the violent games are about you playing as a person, you’re the guy killing aliens/ fighting / operating a sweet transformacar, where as Tetra madness, 3D pinball, and glider pro, you’re not playing AS a character doing these things, it’s a more abstract concept. Nukem and Future cop are long games with at least a vague story and individual levels, where as pinball and tetra madness (maybe glider pro too, as if I use a mac, psh) don’t have this, it’s more repetative than an action game. The games are pretty much all different genres making it a less than fair study.
    It was performed on college students, just like every final-year dissertation written under the sun, yet from such a minute fraction of the population the already clouded conclusion is generalised to everyone.
    The severity of the fight is measured qualitatively (i.e. based on opinions) which is far from ideal scientifically speaking, although I realise you can’t really make a device that measures how violent a fight is.

    With all that in light, I find it ridiculous that a professor thinks that ‘The findings from both studies suggest that violent media make people numb to the pain and suffering of others.’ is a fair statement to draw from this. At best they *vaguely support the hypothesis* that violent media makes you ‘numb to the pain and suffering of others’, but anyone who actually knows enough about games OR experiemental procedure should be able to see that the only thing the results ‘suggest’ is that the professor is a douche who by publishing this will weaken the already frail pseudo science that is psychology.

    Carmageddon, Duke Nukem, Mortal Kombat, Future Cop are the violent games. Glider Pro, 3D Pinball, Austin Powers, and Tetra Madness

  2. @ DStopian

    Do you have any suggestions for which non-violent games they should use?

    The college students only played the games for 20 minutes, so the differences in story or repetitiveness between the violent and non-violent would be insignificant.

    On the choice of games, this relates to the question of generalizability. It is true that they used different genres and had students assigned to each one of them. However, if you want to play a specific genre, Say FPS games. Then you would come to the problem of whether this study can relate to other gaming genres. Therefore, the researchers had to balance between specificity and generalizability.

    While it may seem far-fetched to conclude results from college students to the whole population. However, new research is based on previous research. As I mentioned earlier, they based their experiment from the bystander effect research (which that field of research was replicated in many different settings and people to the same general results). So it seems logical that it can be generalized to the whole population, but it seems the press release didn’t make that clear, so it’s understandable why you asked.

    As for your last paragraph, which research study then strengthens psychology or what do you regard as good research?

  3. Pingback: An international multi-study on helping in video games and its effects on prosocial behaviours (Gentile et al., in press) « VG Researcher - Psychology

  4. Pingback: Supernanny segment on violent video games « VG Researcher – Psychology

  5. I think this is a very interesting study. Over the past little while I have been doing some research about this particular topic and it amazes me on how many experiments have been done to prove that any sort of violent media effects people’s aggression. I think it would be interesting do to these studies in a school. That way you could compare all of these outcomes to a younger control group. I wonder if there would be a difference in that?

    • There are some academics that object that proving is an overstatement and that there are supporting evidence to the theory or some that the evidence is too weak. Christopher Ferguson, Henry Jenkins, Johnathan Freedman to name a few.

      How would it make a difference between older children and younger children?

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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