Competence-impeding electronic games and players’ aggression: the short-version (Przybylski et al., 2014)

Andrew Przybylski (University of Oxford) tweeted a press release foreseen to create a lot of twitter activity and press media.

What causes gamers to feel agressive when playing video games? @ShuhBillSkee et al point to new findings. http://t.co/k5apRCeGqs

— Oxford Internet Institute (@oiioxford) April 7, 2014

The Journal of Personality and Social Psychology ranks very highly in social psychology and its articles are very long, detailed and comprised of multiple studies. There are only two other video game publications in the Journal of Personality and Social Psychology. The first is Craig Anderson and Karen Dill’s 2000 seminal aggression videogame article, the second is Tobias Greitemeyer and Silvia Osswald’s 2010 prosocial videogame article.

Abstract

Recent studies have examined whether electronic games foster aggression. At present, the extent to which games contribute to aggression and the mechanisms through which such links may exist are hotly debated points. In current research we tested a motivational hypothesis derived from self-determination theory—that gaming would be associated with indicators of human aggression to the degree that the interactive elements of games serve to impede players’ fundamental psychological need for competence. Seven studies, using multiple methods to manipulate player competence and a range of approaches for evaluating aggression, indicated that competence-impeding play led to higher levels of aggressive feelings, easier access to aggressive thoughts, and a greater likelihood of enacting aggressive behavior. Results indicated that player perceived competence was positively related to gaming motivation, a factor that was, in turn, negatively associated with player aggression. Overall, this pattern of effects was found to be independent of the presence or absence of violent game contents. We discuss the results in respect to research focused on psychological need frustration and satisfaction and as they regard gaming-related aggression literature.

The reviewing process took very long from initial submission on November 12, 2012 to final acceptance on September 11, 2013.

The self-determination theory (Wikipedia), in short, posits that humans are motivated by three psychological needs: competence, autonomy and relatedness. These needs are psychologically innate, intrinsic and universal. That is these needs are not something imposed by external influences, they are natural human drives. The authors sought to apply self-determination theory within the violent videogame literature due to methodological flaws in past studies, studies that did not support predictions made by the General Aggression Model, the primary theory in explaining violent videogames effects on aggression.

Self-determination theory’s connection to human aggression is argued to be from the deprivation or thwarting of the three psychological needs. Studies on thwarting relatedness and autonomy found such associations to aggression. However, competence deprivation’s association to aggression needed more attention. Given that videogames heavily emphasize competence, this allows the authors to study how competence-deprivation might affect aggression, and they hypothesized similar effects as found with other psychological needs.

The authors wrote several paragraphs relating the current article to other gaming behaviours and effects. One relation is mastery-of-controls where the learned ability to effortlessly use a game’s control scheme is an essential step to have a good feeling of competence in playing a videogame. However, this does not mean that videogame controls and rules should be simplified, complex videogames can be especially satisfying, immersive and stress relieving. They argued for a balance between complexity and competence support, give players sustained support through their trials and errors in learning the game. When the game is not balance, people would complain as having a “steep learning curve”.

Another relation is rage-quitting where players violent quit the game because of being overwhelmed by superior players. The authors argued competence deprivation can result from repeated failures because of an imbalanced scaling of challenge to the player’s abilities. For example, a level that suddenly becomes impossibly hard in comparison to the previous level or the enemies’ level are scaled up too high and too quickly. Another example that is nicely tackled by game companies is matching players’ skills in multiplayer matches, so as not to upset players’ enjoyment, however (IMO) players’ sportsmanship still needs work.

Another relation is the categorization of players through marketing methods (i.e. demographics and marketing segmentation) resulting in market segments of casual and hardcore gamers demographics. This market segmentation of casual vs. hardcore players influence game developers’ design choices for master-of-control, intuitive controls for causal gaming and more complex controls for the hardcore.

In short, the authors seek to examine the structural and motivational as opposed to violent content, predictors for player aggression. Given the Journal of Personality and Social Psychology’s quality requirements, the authors examined their hypotheses in seven studies and I will give the short version for each.

Study 1

An experimental paradigm that is widely used to demonstrate the differences between non-violent and violent videogames’ effect on aggression. They even replicated one of the old studies (Anderson et al., 2004; study 3) by using the same old games. The experiment tested several hypotheses, one hypothesis was reported earlier in 2009, which examined enjoyment, preference and immersion (Przybylski et al., 2009).

They recruited 99 undergraduate students (41 male) who came in the lab individually. They filled out some questionnaire, notably the 35-item state hostility scale which assess for aggressive feelings. Then, they are randomly assigned to play either Glider Pro 4 (non-violent) or Marathon 2 (violent) for 20 minutes. After play, they completed state hostility scale again and completed the 3-item mastery-of-control which assess for competence needs.

Results

The participants filled out the state hostility scale before and after gameplay, so the researchers calculated the difference between them and use the standardized residual scores for analysis.

Regression analysis revealed that competence needs was a predictor for aggressive feelings in that as one feels less competent, aggressive feelings rises. They analyzed with the inclusion of game type as an independent predictor (non-significant) and as a moderator (non-significant).

They did found that game type predicted mastery-of-controls in that violent videogame elicit lower feelings of competence as compared to the non-violent game. So, conducting a mediation analysis they found that game type was not a significant direct predictor for aggressive feelings, but it has an indirect effect in that violent game type significant led to lower competence which in turn led to increases of aggressive feelings. In a way, violent videogames (through harder controls) did increase aggressive feelings.

Study 2

This experiment was also reported earlier for a different hypothesis (Przybylski et al., 2009). This experiment replicates study 1 by modding one videogame to control for all game structures, except for story and graphics.

They recruited 101 undergraduate students (36 male) who came in the lab individually. They filled out questionnaires, including the 35-item state hostility scale. Then, they are given a 20 minute tutorial of playing Half-Life 2. Then, they are randomly assigned to play either a non-violent version of Half-Life 2 where they were instructed to tag competitors teleporting them away or a violent version where they shotgun all the competitors. After play, they completed the state hostility scale again and completed the 3-item mastery-of-control scale.

Results

The participants filled out the state hostility scale before and after gameplay, so the researchers calculated the difference between them and use the standardized residual scores for analysis.

Much like in study 1, they found that competence needs was a predictor for aggressive feelings in that as one feels less competent, aggressive feelings rises. They analyzed with the inclusion of game type as an independent predictor (non-significant).

Because they used the same game for their violent and non-violent versions, they found no significance for violent content nor were there any interactions with content and competence with aggressive feelings.

Study 3

The authors move away from content, for now, and focused on how tweaking people’s competence needs affect aggressive thoughts. Study 3 is of great interest for game designers as it examined control scheme and mastery-of-control.

They recruited 104 undergraduate students (30 males) who came in the lab individually. They were randomly assigned either to play standard Tetris or screwed-up control scheme Tetris for 10 minutes. After play, they completed the lexical decision task (download example) where participants must quickly identify words as legitimately English words or wait if they identified non-legitimate English words. They were given the standard practice procedure beforehand and underwent 100 trials of which contained 60 nonwords, 20 neutrals and 20 aggressives. After that, participants wrote down the first five thoughts when thinking about the game and gave evaluative ratings to these thoughts, from positive, neutral to negative.

Results

Three regression analyses revealed that the screwed-up control condition led to less mastery-of-controls; faster reaction and more automatic access to aggressive thoughts; and more negative evaluations.

Two more regression analyses revealed that mastery-of-control led to slower reaction and less automatic access to aggressive thoughts and greater positive evaluation of the game.

Study 4

Same thing as study 3, but with more measures. This time bringing the additional relationships between control scheme complexity, mastery-of-controls, competence needs, feelings and enjoyment.

They recruited 141 undergraduate students (52 males) who came in the lab individually. Participants filled out 6 aggressive affect items (e.g. angry, hostile) from the PANAS. Then, they were randomly assigned to play either standard Tetris or screwed-up control scheme Tetris for 10 minutes. After play, they completed the 6 aggressive items again, 3-item mastery-of-control, 3-item competence needs and 4-item game enjoyment, the latter two were adapted from the Intrinsic Motivation Inventory.

Results

The participants filled out the 6 aggressive items before and after gameplay, so the researchers calculated the difference between them and use the standardized residual scores for analysis.

In a series of regression analyses, they found that playing the screwed-up control scheme led to lower feelings of mastery-of-controls and lower competence needs. They found that greater competence needs affect greater game enjoyment and lower aggressive feelings.

Moving on to more complex analyses, they found that the relationship between control scheme and competence needs is mediated by mastery-of-controls, thus there is an indirect effect. So, if you play a complex control scheme, you would feel that you have less mastery on the game which in turn led to lower competence needs.

Examining the relationship with aggressive feelings and game enjoyment, they found that control scheme was mediated by competence needs and in another analysis, they found that mastery-of-controls was also mediated by competence needs. Thus, there are two indirect paths leading to aggressive feelings and game enjoyment.

How about the relationship between aggressive feelings and enjoyment, they found that greater feelings is associated with less enjoyment.

In an interesting tangent, the Playstation controller’s symbols had intuitive meanings, but I heard that these meanings were lost outside of Japan. For me, the circle or rather its position on the controller represented “no” or “backspace” whereas the cross is “yes” or the button-mashing spot. For a basic primer, I recommend The Geography of Thought: How Asians and Westerners Think Differently… and Why. In another tangent, My old brother tried Armored Core 3 (2?) and found it difficult, mainly because its control scheme were complex and yet I heard that fans like it that way as it more satisfying and immersive once mastered.

Study 5

Given the previous studies we have seen so far, the authors sought to examine how practice or gaining experience in a new videogame would affect the relationships with behavioural motivations and aggressive feelings.

They recruited 112 undergraduate students (33 males) who came in the lab individually. They completed the 6 aggressive items. Then, they are randomly assigned to either practice Half-Life 2 for a 10-minute practice or 10-minute playing tetris, afterwards they are randomly assigned to either play the non-violent or violent version of Half-Life 2, the same one used in study 2 for 10 minutes. After play, they completed the same 6 aggressive items, 3-item mastery-of-controls, 3-item competence needs and game enjoyment. The participants were then offered to chose to spend more time with the game or browse the internet for the next 10 minutes. So, the amount of time spent on the game is used to assess player motivation.

Results

The participants filled out the 6 aggressive items before and after gameplay, so the researchers calculated the difference between them and use the standardized residual scores for analysis.

Participants who had practice with Half-Life 2 reported greater levels of mastery-of-controls, competence needs, lower feelings of aggression and greater game motivation.

Further analyses looking at mediating relationships, they found that mastery-of-controls mediated the relationship between experience and competence needs, so if you have experience, you would feel that mastered the game which in turn led to satisfying competence needs. They found that player experience was associated with lower levels of aggressive feelings and higher levels of player motivation, violent content had no moderating significance.

The authors examined how competence needs act as a mediator between master-of-controls and player experience with aggressive feelings and player motivation. They found that aggressive feelings and player motivation is indirectly affected by mastery-of-control and player experience, separately, and as mediated by competence. So, lower mastery or experience led to lower competence which in turn led to greater feelings of aggression. As for player motivation, more time is spent if they had more experience or felt greater mastery which is mediated by competence needs.

Finally, they found that player motivation led to lower aggressive feelings. Content was not a significant predictor.

In a tangent, game designers should consider effective ways of giving much practice and support for new players. Some games successfully fit tutorials into the story or the world of the game. In another tangent, when it comes to more complex control scheme, players’ personality characteristics, for example need for cognition (Espejo et al., 2005; Day et al. 2007), would be useful to consider.

Study 6

So moving from the user side to the game side… what happens if the game was being unfair or unbalanced.

They recruited 47 undergraduates students (21 males) who came in the lab individually. First, they were instructed to dip their hand in icy cold water (4 C^o) for 25 seconds ostensibly determined by a participant before them (not really, all participants dip their hands for 25 seconds). Participants were told this is to put them through a physiological stressor, but it was meant for the participant to experience how painful this task is as they will determine how long the next participant will dip their hands in. After, they completed the 6 aggressive items and then they were randomly assigned to play two versions of Tetris: standard Tetris or what looks like Hatetris (get a taste of it) where it calculated the blocks’ utility and gives you the worst one 75% of the time. Play time is 10 minutes. After play time, they completed the same 6 aggressive items and were ask how long the next participant should dip their hand in the water. On average, participants assigned 25.98 seconds to the next participants. So, any deviation from the average is indicative of their willingness to inflict more or less pain, thus aggressive behavior.

FYI, the hand-dipping in icy water was used in previous studies and a graduate colleague is using it in her experiment.

I saw a youtube video of the Impossible Game and made wonder how some people would persist such punishingly hard game. Masochism? Joking, but really I would just rage-quit after a few tries. But then again, Flappy Bird is something like that too, why is it so engaging?

Results

The participants filled out the 6 aggressive items before and after gameplay, so the researchers calculated the difference between them and use the standardized residual scores for analysis.

Their first regression analysis showed that those in the hatetris condition were feeling more aggressive than those in the standard Tetris.

Examining the relationship between playing hatetris vs. Tetris and participants’ aggressive behavior, their regression analysis revealed that those who played Tetris on average gave 22 seconds (3 seconds less than the global average) as opposed to those in the Hatetris condition who gave 29 seconds (7 seconds more than the global average).

Study 7

All experiments and no external validity, let’s switch to a survey study.

The authors surveyed 380 gamers (258 males) who participated in exchange for a raffle for a game of their choice. Respondents reported the three most played videogames in the past month. They also reported for each game their competence needs satisfaction, post-play aggressive feelings, and enjoyment.

The videogames were coded according to the ESRB rating scale and a violence content coding scheme of their own creation, they had 4 research assistants to sift through the game titles.

Results

Due to the nature of the dataset, hierarchical linear modeling was used. TL;DR version: competence satisfaction was negative associated with post-play aggressive feelings and positively associated with player enjoyment. Factoring in violent content, their results revealed that ESRB rating nor the violence coding was linked to aggressive feelings. It did not affect the association between competence satisfaction and post-play aggressive feelings. Finally, post-play aggressive feelings was negatively associated with player enjoyment.

Discussion

The take home message is that some gaming structures, such as difficulty, control scheme, are important factors in one’s sense of mastery and competence satisfaction which they have broad implications for the player, for this long-winded article, aggression. They also found that violent content have non-significant bearing on aggressive feelings, thoughts and behaviours. If you want to know more what self-determination theory can explain other things about video games, go buy Rigby and Ryan’s book Glued to Games. We’re nearing 3000 words and I’m guessing you are getting tired at this point.

Here are some points that caught my attention.

The authors discussed their current works in relation to other theories. The frustration-aggression hpothesis, for example. When you are blocked from your goal, you become frustrated opening a line towards aggression. Self-determination theory can explain the phenomenon as well as that your needs are thwarted. They also noted that frustration may elicit other emotional or motivational reactions and this shown in the current works. When competence was deprived, enjoyment and player motivation went down, so no doubt an array of reactions are likely to show up. So far, we have seen what is most relevant. Second, a bunch of other factors can moderate the link between needs thwarting and aggression. For example, a person’s autonomous self-regulation or self-control can impact their thoughts and feelings to an event.

As you gathered that the results seem to fly in the face of previous aggression videogame studies that found that violent content prime aggression. It does not necessarily contradict the previous studies, but pointed out a confound in those studies which was the differences in competence satisfaction that was inherent between a violent and non-violent videogame. The authors argued that the picture is thus more nuanced when looking at the nature of videogames and aggression. As many people used to say about the difference between television and videogames, it’s the interactivity!

So, if you are piqued by their findings, what can you do to advance this theoretical line? Here are what the authors proposed for you. First, you can help looking at how thwarting autonomy or relatedness needs would affect aggression. A talk with Mike Hanus, who does self-determination stuff, agreed that social ostracism research would be a good bet to check out for relatedness thwarting and he said deadlines, requirements or some deprivation of choice would be some good bet for autonomy thwarting. Second, you can help fund and design longitudinal study to examine long-term effects, so how does being chronically deprived of your psychological needs feels like? Third, more people and a more diverse sample, so maybe some vulnerable population or those who don’t typically participate in studies. Finally, they’ve only measured aggressive behaviours through one method, the hand-dipping in ice cold method, so maybe we should try the noise blast or the chili sauce or the tangram puzzle method.

There is one more limitation not mentioned in the article. This is noticeable for those in social sciences or are experts in statistics. The statistical method used throughout the paper raised eyebrows and rants from colleagues of mine who know a lot about statistics. The gist I got was that there are problems about the approach made by Baron & Kenny (1986) and should approach the statistical analyses proposed by Preacher & Hayes (2008). My colleagues also said that statistical techniques were done like this probably because of reviewers and should have analysed them in a more elegant way like causal-chain test mediation linking the studies. They’ve also recommended “that researchers and reviewers involved in studying psychological dynamics of electronic games adopt the recommendations advanced by Geoff Cumming in Psychological Science (2014).

There are you have it, there are several communication scholars already researching videogames through the self-determination perspective and hopefully this long-winded study spark some ideas in others.

Przybylski, A. K., Deci, E. L., Rigby, C. S., & Ryan, R. M. (2014). Competence-impeding electronic games and players’ aggressive feelings, thoughts, and behaviors. Journal of Personality and Social Psychology, 106 (3), 441-457. DOI:10.1037/a0034820