An international multi-study on helping in video games and its effects on prosocial behaviours (Gentile et al., 2009)

Healing others is a prosocial behaviour

Healing others is an in-game prosocial behaviour

Via Miller-McCune through google alert, a study on gaming prosocial behaviours and its effects on prosocial behaviours in real life is conducted internationally and is authored by the notables of psychology video game effects researchers (i.e. Gentile, Anderson, Huesmann, Bushman), their Japanese counterparts (where I can’t find their faculty webpage for my list) and the rest (i.e. grad students and names I don’t recognize).

I’ve managed to get the draft of the study, it was sufficient (and short) for me to post it. The study will be published soon in Personality and Social Psychology Bulletin.


Although dozens of studies have documented a relationship between violent video games and aggressive behaviors, very little attention has been paid to potential effects of prosocial games. Theoretically, games in which game characters help and support each other in nonviolent ways should increase both short-term and long-term prosocial behaviors. We report three studies conducted in three countries with three age groups to test this hypothesis. In the correlational study, Singaporean middle-school students who played more prosocial games behaved more prosocially. In the two longitudinal samples of Japanese children and adolescents, prosocial game play predicted later increases in prosocial behavior. In the experimental study, U.S. undergraduates randomly assigned to play prosocial games behaved more prosocially toward another student. These similar results across different methodologies, ages, and cultures provide robust evidence of a prosocial game content effect, and they provide support for the General Learning Model.

I’ve got 54 more and less than 2 weeks to go before the deadline. Update 02/04/09: here’s the university press release.

As mentioned earlier, the study is conducted internationally, that is in the United States, Singapore and Japan. Another notable point is that it is also a multi-study, that is the paper contains longitudinal, experimental and correlational studies, thereby covering three major types of psychological research methods and three studies.

Their theoretical basis for prosocial gaming behaviours is basically the same when it comes to violent gaming behaviours: learning (en gros). I guess they just flipped the learning coin and see what’s on the other side. They’ve noted that prosocial behaviours and aggressive behaviours are not mutually exclusive (i.e. it’s not black and white).

Study 1 (correlational study)

Participants: 727 Singaporean 7th (N = 446) and 8th graders (N = 281). Average age is 13 years. 73% of sample is male due to two schools being a boy’s school.


Video game habits: Participants are asked to list their three favourite video games; how many hours they played it; how often they help others in the game and how often they hurt or kill others.

Prosocial Orientation Questionnaire: 18-items questionnaire that assesses helping behaviours (subscale 1) and cooperation and sharing (subscale 2) in real life. It does not say how it’s answered, but I gather it could be a likert-scale questionnaire.

Children’s Empathic Attitudes Questionnaire: 16-items questionnaire that assesses trait empathy.

Normative Beliefs about Aggression Scale: 20-item used to assess aggressive cognitions (e.g. violent  thoughts or approving acts of aggression). To me, sounds like an attitude scale. I better review my social psychology…

Personal Strengths Inventory: used parts of the inventory, namely the emotional awareness subscale, which consisted of 5-items.

Hostile Attribution Bias: 6 stories of ambiguous provocative situations and they are asked to provide explanations of the situations. Also assesses for aggressive cognitions.


I don’t know what the beta stands for in the statistics. I don’t see the effects size. All I could recognize is the p value, which a certain professor in my Uni calls it meaningless. They controlled for sex, age, violence exposure, amount of play time in their analyses.

They found that prosocial gaming behaviours is statistically significantly related to prosocial behaviours in real life. Every prosocial scales were found to be significantly related to prosocial games. They also found the same when it comes to aggression, although it’s negatively related to helping behaviours and empathy. They speculated that the video games might have specific effects, instead of broad ones.

They’re aware of the dual correlation and something about multicollinearity did not influence their analyses.

Study 2 (longitudinal study)

Comparing with study 1 and study 3, this is the weakest link part of the paper because of the so few measures used. It seems that the problem came from the same data set from a previous study.

Participants: two samples: sample 1: 780 Japanese 5th graders (average age is 10.9), equal gender ratio. Sample 2: 1050 Japanese 8th  and 11th graders, average is 13 and 16 years, respectively. Equal gender ratio.


Prosocial video game exposure: asked how often they played games in the last month where they helped (a) characters helped others in distress, (b) scenes where friendships or affections between parents and children are shown. What about between the main characters? Like the friendship between Sora and the Disney characters? Answered on a 5-point scale.

Prosocial behaviours: asked whether they had done, in the last month,  “each of four helpful or prosocial behaviours” on a 5-point scale. That part of the paper needs to be expanded…


They gave those questionnaires twice with time span of 3 to 4 months in between. The questionnaires were given in classes. The authors noted that this is short for longitudinal study. A span of at least 6 months would be better.


They analysed the data using structural equation modeling. Another statistical technique that grad students get to learn. They’ve controlled the usual suspects (i.e. sex, age, etc.) and found statistical signifcances: prosocial video gaming and prosocial behaviours are significantly related to each other in both Time 1 and Time 2; they’re also related going from Time 1 to Time 2. So long-term exposure of prosocial gaming and taking into account of nice behaviours in real life leads to greater (?, not sure if i’m reading this right) prosocial behaviours and gaming. In addition, being nice to begin with can also lead to more prosocial video gaming which also leads to more prosocial behaviours in both gaming and in real life. A bi-directional influence, if I paraphrased the authors.

Study 3 (experimental study)

Participants: 161 American college students (64 men, 95 women, average is 19.2)


Tangram puzzles: this one is used to assess helping and hurting behaviours. Not calling it prosocial behaviours because the measure is a situational measure and does not correspond generally in real life or in the long-term. There are some missing details, so I’ll do my best.

Participants are asked to assign 11 tangram puzzles to a partner from a set of 30 puzzles ranging in three difficulty levels: 1o easy, 10 medium or 10 hard. They are told that if their partner completes 10 of 11 puzzles in 10 minutes, the partner would get a 10$ gift certificate. Participants were told to vary their puzzles in terms of difficulty to avoid people from choosing all easy puzzles. So the amount of easy or hard puzzles determine how much the participant would like to help another person getting a reward or how much they would like to hurt their partner from getting the reward. Interesting.

Video game rating: participants rate the games in terms of several criteria like how enjoyable, or excting or violent, etc. Answered on a 10-point scale.

Buss-Perry aggression questionnaire: 29-item questionnaire, answered on a 5-point scale. Measures for trait (personality) aggression.

Video games used

These games were, of course, evaluated by students (27, actually) prior to the experiment to see if these games would serve the study’s purpose.

Prosocial video games: Super Mario Sunshine and Chibi-Robo.

Neutral video games: Pure Pinball and Super Monkey Ball Deluxe.

Violent video games: Ty2 and Crash Twinsanity.


Participants are randomly assigned to play one of the games for 20 minutes. An interesting thing is that they played specific parts of the game, so probably parts that maximize exposure of the target variable of each game type. After that they performed the tangram puzzles task, then the video game rating and finally the Buss-Perry questionnaire. Participants were screened at debriefing to see whether they suspected of the study’s purpose, those that were excluded from analysis. Why? because if you are aware of the purpose of study, your behaviour and attitude will change and therefore screw up the data. Fortunately, excluding them did not change the results so they were kept in the analysis.


Their analyses looked  at the number of easy vs. hard puzzles given to their partner. They did not include the medium puzzles because it’s redundant said the authors. They also controlled trait aggression and game difficulty in their analyses.

If you look at the chart below, it shows the average number of helpful (easy puzzles) acts vs. hurtful acts (hard puzzles) given by the different video game group. Generally, there are more helpful behaviours than hurtful ones, a main effect. They’ve found interaction effects, in that those in the prosocial game were more helpful than the neutral or violent game group, the two latter groups did not differ from each other. Whereas the violent game group were more hurtful than the neutral or prosocial game group, the two latter groups did differ from each other.

Staring at the chart several times during testing of my participants (Andy knows and maybe Chris and Doug if you’re reading this post), I was wondering if they compared the helpful vs. hurtful behaviours within a game group, specifically the violent game group. Although it was indicated earlier that participants were generally more helpful, I can’t help but ask if those in the violent group were neither helpful nor hurtful since they give out about the same amount of helpful and hurtful behaviours. I was about to say neutral, but it’s more like they’re apathetic.


They conducted an alternative analysis, which is related to the student evaluation of the games, to see if the amount of in-game helpful or hurtful behaviours would have a role in the puzzle helpful/hurtful behaviours. It does in the expected direction, but that’s all I’ll write.


Aerith Gainsborough is one of the video game characters that characterizes prosocial behaviours

Aerith Gainsborough is one of the video game characters that displays a lot of prosocial behaviours and has a sweet and kind personality.

My first thought is role models of prosocial characters might make players more prosocial or the least more loving people than immature 13-year old boys.

A few notes about the study from the authors is that it covers the major types of psychological research methods, covers a diverse range of population and cultures and ages which leads to a great degree of generalization.

They demonstrated short-term and causal effects in study 3, correlations between prosocial vidoe gaming and prosocial behaviours in real life in study 2, and the bi-directionality of influences between prosocial gaming and prosocial behaviours, as the authors noted it doesn’t what started first, it doesn’t matter whether prosocial kids like prosocial games or prosocial games makes kids prosocial, both can happen.

They’ve also reported some limitations: self-report measures were used in study 1 and 2. So it’s a bit subjective on the part of the participants (social desirability bias, poor memory accuracy, etc.).

So the take home message that the authors has is a quote from John Wright (not sure which one in wikipedia): “The medium isn’t the message. The message is the message.”  (That sounds the opposite quote from Marshall McLuchlan) or it’s the context that matters and how it’s used.

So my take is that as games becomes more complex in terms of technology, literary and philosophy,  by exposing the many aspects of humanity (i.e. the question of good vs. evil, social responsibility, consequences of war politics, etc.). I guess it would adjust people’s worldview into something less of the black-and-white fantasy land. Gone are the heydays of shooting and points, games have grown up into interactive novels. However, some might say novels for teens, there’s still the need for games to grow up.

Gentile, D. A., Anderson, C. A., Yukawa, S., Ihori, N., Saleem, M., Ming, L. K., Liau, A. K., Khoo, A., Bushman, B. J., Huesmann, L. R. & Sakamoto, A. (2009). The effects of prosocial video games on prosocial behaviors: International evidence from correlational, longitudinal and experimental studies. Personality and Social Psychology Bulletin, 35 (6), 752-763.


12 thoughts on “An international multi-study on helping in video games and its effects on prosocial behaviours (Gentile et al., 2009)

  1. I will have to read over these details later, but I do have an initial ponderer/question.

    From glacing, it looks like these studies are trying to talk about violent/non-violent games, but it sounds like their methods are looking at cooperative/competitive and then testing, well, how cooperative or competitive the person is feeling afterwards.

    Did they test any violent/cooperative or non-violent/competitive combination? If not this research feels like a bit of a bait-and-switch.

    • It does seem like they might be just looking at cooperative/competitive gaming. It might just seem that coop games are often prosocial games and makes a good match gameplay wise, whereas violent games can fit with either coop, competitive or even combined. It’s quite possible that coop gaming could be an essential component for prosocial gaming.

      There are conceptual overlap between feeling cooperative or competitive with feeling prosocial or aggressive, respectively. Try to write a list of cooperative feelings and prosocial feelings, you would see that there are some things to go in both.

      When it comes to testing the combination of violent content and coop vs. competitive game, they haven’t tested that in this particular study. There are others that did something similar to what you described.

      Eastin (2007) looked at competitive vs. cooperative play in violent video games, although he did not looked into the non-violence aspect. He found some suggestive results that the competitiveness in violent video games is a factor.

      Another study that I’ve only briefly looked at is from Anderson & Carnagey (2009) looked at violent vs. non-violent play in a competitive environment (American Football). From the abstract, it seems that “violent content uniquely leads to increases in several aggression-related variables.”

      So if we infer from these studies, a non-violent cooperative game would seem to be the most prosocial game play, followed by non-violent competitive games.

  2. Pingback: Research: Violent Games Correlate with Hurting Others, Prosocial Games Correlate with Helping : ::: The Source of All News

  3. Pingback: Gaming parents, action videogames and civic engagement (Ferguson et al., 2010) « VG Researcher

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