A new study just published in Psychological Bulletin has reviewed studies on the effects of violent video games and concludes that they cause a small but reliable increase in aggressive behaviour and anti-social thinking.
The study, led by psychologist Craig Anderson, is a type of meta-analysis which attempts to mathematically aggregate the results of past studies to see what the overall effect is.
There have been several similar studies in recent years which have come to different conclusions, based on whether the results have been thought to have been affected by publication bias or not. In other words, while the published studies suggest there is a small reliable effect of video games on aggression some reviews have suggested this is because fewer of the studies that don’t find a link actually get published.
This new study aimed to include unpublished studies and also looked at studies from both Western cultures, like the US and Europe, and Eastern cultures, such as Japan, to see if social environment influences any potential link.
The review included both observational studies, which look at what happens in the ‘real world’ but can’t tell us whether gaming causes aggression (it could be just that more aggressive people play more violent games), and experimental studies which can determine cause, because participants are randomly assigned to groups and given either violent or non-violent games, but are a little more removed from everyday life.
The researchers examined whether violent video games led to changes in aggressive behaviour, thoughts and emotions, and for changes in empathy and helpful behaviour to others.
Overall, the analysis concluded that violent video gaming causes a small but reliable increase in aggression and possibly a reduction in helpful behaviour and empathy.
The results on empathy were the weakest, however, as only study was an experiment and the researchers lumped together research that used questionnaires and which tested bodily desensitisation (whether people bodily react less to emotional events when they re-experience them) which is not a good measure of someone’s mental state.
One interesting aspect of the analysis is that the researchers looked at a number of game characteristics to see if they had an effect; for example, whether the people were playing in first- or third-person, whether the violence was towards human or non-human characters; and found that none of this made much difference.
What this suggests is that the effect is not due to non-specific priming, a psychological effect whereby experiencing one type of concept makes closely related concepts and actions more accessible and more likely. In this case, the fictional violence is assumed to make aggressive thoughts and actions more easily triggered.
It must be said that the overall effect was quite small. For the statistically inclined, the correlation was r = .19 for all studies and r = .24 when they looked only at the most rigorous research. This means that violent video games accounted for between about 3.6% and 5.8% of the total change in aggressiveness.
Interestingly, despite the fact that Japan, for example, is more culturally adverse to aggression than Western countries, the effects seems to be equally as apparent on either side of the world.
The journal published a discussion based on the study, including a criticism by psychologists Christopher Ferguson and John Kilburn, who have published previous analyses suggesting that the violence effect is down to publication bias.
The discussion focusses on various technical issues which are well answered by the original authors, although perhaps the most significant points of disagreement focus on two areas.
The first is that this new analyses only focused on simple relations, and didn’t take into account whether other factors could be having an influence. For example, a previous study suggested that when pre-existing emotional, family and social problems are accounted for, the aggression increasing effects of video games disappears.
The second concerns how important this small effect is. On an individual level a small change may be undetectable amid the to-and-fro of everyday life, but at the level of the population it could conceivably increase the number of aggressive incidents, although these are often the hardest effects to track.
Link to PubMed entry for study.
pdf of full text.