Video games cause violence

For almost as long as there have been video games, there have been people arguing that they are bad for you. There also seems to be a wealth of experimental evidence behind it (Hasan et al., 2013, to name just one of many). But there have been suggestions that these negative outcomes are oversold.

2DJimmyN/DeviantArt

Problems with the literature:

One of the strongest pieces of evidence for the negative effects of video games is a meta-analysis by Anderson et al. (2010). They found strong evidence that “exposure to violent video games is a causal risk factor for increased aggressive behaviour, aggressive cognition, and aggressive affect and for decreased empathy and prosocial behaviour”. However, there were immediate questions about the methodology in this meta-analysis. Ferguson & Kilburn (2010) commented that many studies do not relate well to aggression and the authors do not consider the impact of unstandardised aggression measures (differences between studies in how they measured aggressive behaviour), among other things. They comment that the studies analysed in Anderson et al. (2010) only show weak evidence for their conclusion. A more recent reanalysis by Hilgard, Engelhardt, and Rouder (2016) used more advanced tools to adjust for research bias and found that the short-term effects of game play on aggressive feelings and behaviour were badly overestimated by bias. The adjustments recommended by Hilgard et al. (2016) were mostly substantially lower than those performed by Anderson et al. (2010), with some being smaller adjustments. In some studies, the result was adjusted to zero e.g. aggressive affect. This does not completely eliminate the original findings but I feel we should adjust our estimate of the strength of the causal association downwards.

Ferguson (2007a) conducted a meta-analysis to see the relationship between violent video games and aggression demonstrated in a lab environment. He found that there was strong evidence of publication bias, showing that the inclusion of unpublished or suppressed studies made the result of the meta-analysis non-significant and/or trivial. It was also shown that age of the participant was a significant moderator effect which research showing a positive association e.g. Anderson & Dill (2000), being one of the most cited, did not control for. Anderson & Dill (2000) also didn’t control for family background which may lead to a preference for aggressive video games as well as aggressive behaviour.

Variety when there shouldn’t be:

Most of these studies use a tool called the “Competitive Reactive Time Task”, also called the “Taylor Aggression Paradigm” (TAP). It is supposed to measure aggressive behaviour in the laboratory. But there are a huge number of ways that it is used: 120 studies were found to have used TAP but there were 147 different ways it has been used (at the time of writing) e.g. some use electric shocks, others use noise blasts. This makes comparing results between studies difficult and raises questions about why there is so much variety. Are researchers using a variety of methods and reporting the measure that produces a positive result? For more detail, check out this excellent website by Malte Elson, 2016. An analysis by Elson et al. (2014) found that not only was there a wide variety of ways the TAP was being implemented, there were many different data analysis strategies. This raises further questions about researchers looking for positive results and publishing only those methods that produce them. Elson et al. analysed data for 3 studies and found that, depending on which strategy was used, p-values and effect sizes showed very different results (and sometimes even reversed the effect). This further undermines the credibility of research using the TAP. Ferguson (2007a) points out that studies use TAP but “none of these indices of “aggression” [have] been linked with actual criminally violent behavior” (these “indices” being giving noise blasts to other participants and other methods). This further questions the ability of studies using TAP to show us how violent video games make people more violent; even if they make them more aggressive in the TAP, that might not correspond with an increase in real world aggression.

There have also been concerns about the validity of this tool to measure aggression in the laboratory. Tedeschi & Quigley (1996) first identified the problem of demand characteristics for this method (participants acting as they think the researcher wants them to) as this would undermine the relevance to the real world of findings using this tool. They developed their criticism (Tedeschi & Quigley, 2000) to include difficulties of measuring the concept of aggression using the TAP and how researchers have selected certain evidence to support the convergent validity of the TAP whilst ignoring evidence that doesn’t.

For research that demonstrates a link between video game violence and actual aggressive behaviour, Ferguson (2007b) found a very weak relationship between video game violence and aggressive behaviour. However he found more evidence of publication bias which, once corrected for, reduced the relationship to almost zero. The link between video games and aggression in real life has been further questioned by Ferguson (2014). His paper suggested that as video games sales increases, youth violence decreases which suggests (on a macro-scale) that there isn’t a clear relationship between video game violence and real-world violence.

Minimal impact:

Etchells et al. (2016) performed an analysis of 1,815 children’s video game exposure at age 8/9 and rates of conduct disorder (CD) and depression at 15. They found a very weak association between video game exposure and CD when controlling for confounders (e.g. sex, bullying, peer problems, etc.) but only 26 met the criteria for CD. The measure just reached significance: p=0.05 fully adjusted (Table 2). They found no association between video game exposure and depression when controlling for confounders (Table 2). There was also only a weak association between children playing more violent games (shoot-em-ups vs competitive games) and an increased risk of CD (Table 4). With the weak evidence and the fact they do not consider the impact of other media exposure, I agree with their conclusion that assuming this association is causal is “inappropriate”.

A meta-analysis found that video games have a minimal impact on increased aggression, reduced prosocial behaviour, reduced academic performance, depressive symptoms, and attention deficit symptoms (Ferguson, 2015). This was found after controlling for publication bias. The study was re-analysed by Furuya-Kanamori & Doi (2016) and replicated the statistically significant but very small impact video-games have on several outcomes. A longitudinal study by Ferguson et al. (2012) found that video game exposure did not relate to any of the negative outcomes analysed, whilst exposure to family violence and antisocial personality traits did predict aggressive behaviour. However this study used self-report measures of aggressive behaviour and video game exposure so caution should be encouraged.

Drummond & Sauer (2014) analysed data from 192,000 students across 22 countries who took part in the Programme for International Student Assessment (PISA) and found video-game use had a negligible impact on academic success in Science, Mathematics, and Reading. Their large sample size, psychometrically valid tests, and natural environment were positives of the study whilst the reliance on self-report for the frequency of video-game playing is a negative. McCarthy et al. (2016) found playing video-games (either violent or nonviolent) had no impact on participant’s aggressive inclinations. This study was preregistered (for a detailed description of what study preregistration is and why it’s a good thing, click here) so we can be more confident in the results and that they aren’t the result of p-hacking. These studies together indicate that video-games have less of an impact than implied by a quick read of the literature.

It’s (always) more complicated:

A Bayesian reanalysis of many of the studies and meta-analyses in the violent video-game literature was conducted by Hilgard, Engelhardt, Bartholow, & Rouder (2016). They found the evidence against the effect of violent video-games varied hugely: some strongly supported the null hypothesis (violent video-games had no impact on the dependent variable), some finding weak evidence for the null, and some may have even found evidence for the alternative hypothesis. Their results also suggest many of the studies using “matched” non-violent and violent video-games weren’t in fact well matched. This means comparisons between the types of games cannot rule out confounding variables to explain differences between them. They recommend larger sample sizes and Bayesian analysis for future studies to improve the strength of the evidence.

Freedom?

Does this mean that you can do whatever you want, play for 7 hours a day, 6 days a week, and everything be rosy? No. Some research has found that playing more than 3 hours a day is linked with higher levels of hyperactivity and conduct issues (Przybylski & Mishkin, 2015) and lower levels of prosocial behaviour (Przybylski, 2014). But this same research found a positive impact of playing roughly 1 hour a day on hyperactivity and conduct issues (Przybylski & Mishkin, 2015) and prosocial behaviour (Przybylski, 2014). The idea of an ideal amount of time to spend playing video games was corroborated by Przybylski & Weinstein (2017). They conduced a preregistered study on just over 120,000 children and examined the link between video game use and mental health. They found that “moderate use of digital technology is not intrinsically harmful”, running counter to many previous studies suggesting it has a negative impact. There is more and more research demonstrating the potential positives of engaging in video games (e.g. Eichenbaum, Bavelier, & Green, 2014) but that is another post for another time. But I hope I have demonstrated that the evidence for the negative impacts of video games (especially with regards to causing real-world violence) are overstated.

References:

2DJimmyN. (2008). Angry Video Game Nerd Artwork. Available at: http://2djimmyn.deviantart.com/art/Angry-Video-Game-Nerd-Artwork-74392515. [Accessed: 5th April 2016].

Anderson, C.A. & Dill, K.E. (2004). Video Games and Aggressive Thoughts, Feelings, and Behaviour in the Laboratory and in Life. Journal of Personality and Social Psychology, 78 (4), 772-790.

Anderson, C.A.; Shibuya, A.; Ihori, N.; Swing, E.L.; Bushman, B.J.; Sakamoto, A.; Rothstein, H.R.; & Saleem, M. (2010). Violent Video Game Effects on Aggression, Empathy, and Prosocial Behaviour in Eastern and Western Countries: A Meta-Analytic Review. Psychological Bulletin, 136 (2), 151-173.

Chambers, C. (2012). Changing the culture of scientific publishing from within [online] Available at: http://neurochambers.blogspot.co.uk/2012/10/changing-culture-of-scientific.html [Accessed on: 19/09/2016]

Drummond, A. & Sauer, J.D. (2014) Video-Games Do Not Negatively Impact Adolescent Academic Performance in Science, Mathematics or Reading. PLoS ONE 9(4): e87943. doi: 10.1371/journal.pone.0087943

Eichenbaum, A.; Bavelier, D.; & Green, C.S. (2014). Video Games: Play that Can Do Serious Good. American Journal of Play, 7, 50-72.

Elson, M. (2016). Flexibility in Methods and Measures of Social Science, [online] 3 April. Available at: http://crtt.flexiblemeasures.com/ [Accessed on: 05/04/2016].

Elson, M. (2014). Press CRTT to measure aggressive behavior: The unstandardized use of the competitive reaction time task in aggression research. Psychological Assessment, 26 (2), 419-432. 

Etchells, P.J.; Gage, S.H.; Rutherford, A.D.; & Munafo, M.R. (2016). Prospective Investigation of Video Game Use in Children and Subsequent Conduct Disorder and Depression Using Data from the Avon Longitudinal Study of Parents and Children. PLOS ONE. Available at: http://dx.doi.org/10.1371/journal.pone.0147732. [Accessed on 05/04/2016].

Ferguson, C.J. (2007a) Evidence for publication bias in video game violence effects literature: A meta-analytic review. Aggression and Violent Behaviour, 12 (4), 470-482

Ferguson, C.J. (2007b) The Good, the Bad, and the Ugly: A Meta-analytic Review of Positive and Negative Effects of Violent Video Games. Psychiatry, 78, 309-316.

Ferguson, C.J. & Kilburn, J. (2010). Much Ado About Nothing: The Misestimation and Overinterpretation of Violent Video Game Effects in Eastern and Western Countries: Comment on Anderson et al. (2010). Psychological Bulletin, 136 (2), 174-178.

Ferguson, C.J.; Miguel, S.M.; Garza, A.; & Jerabeck, J.M. (2012). A longitudinal test of video game violence influences on dating and aggression: A 3-year longitudinal study of adolescents. Journal of Psychiatric Research, 46 (2), 141-146.

Ferguson, C.J. (2014). Does Media Violence Predict Societal Violence? It Depends What You Look At and When. Journal of Communication, 65 (1), E1-E22.

Ferguson, C.J. (2015). Do Angry Birds Make for Angry Children? A Meta-Analysis of Video Game Influences on Children’s and Adolescents’ Aggression, Mental Health, Prosocial Behavior, and Academic Performance. Perspectives on Psychological Science, 10 (5), 646-666.

Furuya-Kanamori, L. & Doi, S. A. R. (2016). Angry Birds, Angry Children, and Angry Meta-Analysts. A Reanalysis. Perspectives on Psychological Science, 11 (3) 408-414.

Hasan, Y.; Begue, L.; Scharkow, M.; & Busman, B.J. (2013). The more you play, the more aggressive you become: A long-term experimental study of cumulative violent video games effects on hostile expectations and aggressive behaviour. Journal of Experimental Social Psychology, 49, 224-227.

Higgins, J.P.T. & Green, S. (2011). Trim and fill. Cochrane Handbook for Systematic Reviews of Interventions [online] Available at: http://handbook.cochrane.org/chapter_10/10_4_4_2_trim_and_fill.htm [Accessed on 05/04/2016].

Hilgard, J.; Engelhardt, C.R.; Bartholow, B.D.; & Rouder, J.N. (2016). How Much Evidence Is p > .05? Stimulus Pre-Testing and Null Primary Outcomes in Violent Video Games Research. [online] Available at: http://pcl.missouri.edu/sites/default/files/HilgardPPM.pdf

Hilgard, J.; Engelhardt, C.R.; & Rouder, J.N. (2016). Overestimated Effects of Violent Video Games on Aggressive Outcomes in Anderson et al. (2010). Early view article. Published 11/03/2016. Available at: https://github.com/Joe-Hilgard/Anderson-meta/blob/master/meta_ms.pdf. [Accessed: 5th April 2016].

Przybylski, A.K. (2014). Electronic Gaming and Psychosocial Adjustment. Pediatrics, 134 (3). Available through: PsychBrief’s DropBox <https://www.dropbox.com/home/Public?preview=Electronic+Gaming+and+Psychosocial+Adjustment.pdf> [Accessed 05/04/2016].

Przybylski, A.K. & Mishkin, A.F. (2015). How the Quantity and Quality of Electronic Gaming Relates to Adolescents’ Academic Engagement and Psychosocial Adjustment. Psychology of Popular Media Culture. Advance online publication http://dx.doi.org/10.1037/ppm0000070

Przybylski, A.K. & Weinstein, N. (2017). A Large-Scale Test of the Goldilocks Hypothesis:
Quantifying the Relations Between Digital-Screen Use and the Mental Well-Being of Adolescents. Psychological Science, 28 (2), 204-215.

Tedeschi, J.M. & Quigley, B.M. (1996). Limitations of laboratory aggression paradigms for studying aggression. Aggression and Violent Behaviour, 1 (2), 163-177. 

Tedeschi, J.M. & Quigley, B.M. (2000). A further comment on the construct validity of laboratory aggression paradigms: A response to giancola and chermack. Aggression and Violent Behaviour, 5 (2), 127-136.