Unless I get into graduate school, within 2 years I won’t be able to keep tabs on the latest psychological research in video games as I will sever my academic ties and go into corporate world. This will affect the future of this blog and its tiny importance in the video game world… I guess I could get away with little anxiety.
Another work from Christopher Barlett of Iowa State University, still a PhD student at the moment, has published a study that looks into the question whether realism bears influence in the relationship between violent content in video games and aggression.
Two studies were conducted that tested the moderating role of video game graphics quality in the relationship between video game content and aggression-related variables. In both studies, participants played either a violent or nonviolent video game on one of three video game systems with differing technological computing power (which contributes to the realism depicted in these video games). In Study 2, the moderating and mediating role of immersion was also tested. Results showed that video game violence exposure was related to aggressive cognitions and state of hostility. Video game technology did not moderate this relationship. Finally, immersion, as an individual difference variable, did not moderate or mediate this relationship. This suggests that aggressive cognitions and feelings occur independent of how technologically advanced the graphics are and the extent to which one feels as though they are immersed in a violent video game. Implications and future research are discussed.
There are few studies that examined the effects of realism using commercially available video games. Among those (that I am aware of) who studied this question are James D. Ivory, Marina Krcmar, Kirstie Farrar and Kristine Nowak. Several researchers have briefly discussed that issue, for example Olson and Kutner’s book Grand Theft Childhood and Shapiro’s chapter.
There are a few key concepts to introduce before going into the study. Video game realism is somewhat tricky to define as some go by sensory information (i.e. graphics, sound quality, etc.), or something psychological realism as referred to absolute perceived realism, where other elements of the environment (settings, characters or actions) can be related to the real world. To sum it up, I’ll use games based from the Cry engine (Far Cry series and Crysis): sensory realism: what you see looks real, but it is not real, so Crysis looks science-fictiony while Far Cry looks so real. Absolute perceived realism, it is not necessary to make it look real as long as people are convinced that their environment is real (I recall a similar line from Stephane Bouchard du l’Universite du Quebec en Outaouis), so Far Cry is realistic up to a point until the player picks the unrealistic stuff (see Yatzhee).
Going into rant mode, most politicians, pundits and average people discussed realism based on sensory quality. They argued that as games become more realistic the aversive behavioural effects it has on individuals worsens. It’s a fine logical assumption and makes common sense, but common sense can be overthrown which could mean that graphical realism may be the tip of the iceberg and psychological realism lies submerged in the water.
Immersion is another key concept in realism to which IMO represents the whole iceberg. It relates pretty much to telepresence, where individuals feel as though they are in the virtual space. There are several studies that tested the relationship with telepresence and video game violence . Currently, there are mixed results about the role of immersion in the media violence relationship effects, some of it are attributed to methodological differences, some on operational definition, and others may be due to chance. So let’s get to it.
Participants: 245 undergraduates (128 males and 117 females), average age is 19.07 (SD=2.53), 67.8% are first-years, 89.4% are Caucasians, average play time per week is 6.88 hours (SD=7.89). Looks about average for the university population. However, some participants were excluded from the analyses due to their being figuring the purpose of the study (see demographics and suspiciousness questionnaire)
Buss-Perry aggression questionnaire: measures for trait aggression. It has 29 items that are answered by a 5-point scale.
Story stems: three ambiguous stories where participants are asked to answer in their own words what the main character in these stories would do, say and think in these ambiguous situations. Two lab members (aka coders) read these answers and coded them according to content and frequency. For example, I’ll punch him and yell at him for being so stupid. I code that as punch him (physical aggression), yell at him (verbal aggression), for being so stupid (negative comment or psychological aggression or maybe verbal aggression). I just realized that the psych lab I am in has also this kind of measure, in fact, I did (still doing it) something similar for my honours thesis.
One general limitation from this measure is that participants would answer them hypothetically so their answers could be the most ideal behaviours, feelings and thoughts for them. But in my experience as a hypothetical story coder, although there are typical answers for each situation, there is still variability in the answers they give and some can be brutally honest in their answers. Since they’re answering after playing the video game and we find statistically significant differences. Then at least, there are effects.
Demographics and suspiciousness questionnaire: age, sex, blah blah blah. The suspiciousness questionnaire is to find out if participants figure out the true purpose of the study because some of the measures like the aggression measures are self-reported and be subjected to response bias and therefore we participants that can ruin the results.
Physiological arousal: heart rate device where participants are asked to press their right index finger on a sensor three times on several occasions. Not good enough because participants have different ways of pressing their finger on the sensor, either pressing
too hard or too light or they have small veins or too many variables to talk about. I say “meh”.
Video games used: Since they’re looking at differences in video game graphics, it can be pretty arduous on defining graphics and how we differentiate between low realism versus high realism. Try to have a 5 minutes discussion with someone and see how it goes. So, they’ve decided to use video game consoles as a (IMO) rough measure of graphical realism. Barlett and company acknowledged that it is simplistic and there should be more to it. But I ask those in the video game industry to help VGresearchers in defining and identifying components of video game realism.
It is pretty consistent and a clear difference in graphical realism between the games. Although, I’m a bit worried about the differences between the N64 and PS2 versions of Mortal Kombat, but I haven’t played
Mortal Kombat series, so I don’t know if
these differences are clear.
Participants get their baseline heart rate measured three times, complete the Buss-Perry aggression questionnaire. They are randomly assigned to one of the six gaming conditions, either violent or non-violent and either in SNES, N64 or PS2. They are given a tutorial, then 15 minutes of play. After that, heart rate measured three times, then complete story stems and the rest.
Number of participants assigned to gaming condition: Violent-SNES = 35, violent-N64 = 34, violent-PS2 = 33. Non-violent-SNES = 35, non-violent-N64 = 34, non-violent-PS2 = 27. So pretty much equal.
Physiological arousal: A violent content by game console ANCOVA analysis has found no significant results.
Aggressive scripts: using data from story stems as the dependent variable. Violent content by game consoles ANCOVA analysis. Main effects were found for violent content in that violent games in general increased aggression codes in story stems than non-violent games. Trait aggression was found significant and will be treated as a covariate or controlled for.
Interactions effects were found in that those in the violent-N64 were found to have significantly higher aggression scores than those in violent-SNES or violent-PS2. There were no differences between the violent conditions. The lowest score is the nonviolent-N64 group, however it doesn’t look like it’s statistically different from the other nonviolent groups.
This is quite weird and is quite counter to their hypotheses. Barlett et al. offered an explanation that it may have been frustration using outdated game controllers or that N64 games are harder to play than the rest. Or IMO, it may be possible that certain content in the N64 may have been more graphic than the other consoles. In any case, it’s just weird, let’s just not fuss over it and have it replicated in another study which leads to study 2. A side note, the data was collected in 2005-2006, but again it’s not far back in the past, so no complaining.
It’s essentially the same study with further improvements like more recent game consoles, more aggression measures and an immersion measure added to the procedure.
Participants: 143 undergraduates (94 males and 49 females), average age is 19.2 years, mostly first-years, Caucasians and average play time is 6.2 hours (SD = 7.40)
State hostility scale: measures participants’ current or state mood, specifically for aggressive feelings. 35-items answered on a 5-point scale.
Word completion task: 96 incomplete words that participants fill in the blanks, depending on how they fill in the words, they can be worded something aggressive, non-aggressive or a non-word. Measures for aggressive cognition.
Presence questionnaire: measures for immersion or telepresence. 27 items answered on a 7-point scale.
Video game rating sheet: 10-items answered on a 5-point scale. Participants rate the games on a several criteria, such as how violent, difficult or exciting it is.
Buss-Perry aggression questionnaire: measures for trait aggression. It has 29 items that are answered by a 5-point scale.
Physiological arousal: heart rate device where participants are asked to press their right index finger on a sensor three times on several occasions.
Demographics and suspiciousness questionnaire: the same as in study 1.
Video games used: They omitted the SNES from this study, no reasons given. IMO, the results would’ve been the same; there are logistical concerns (i.e. the need to get more participants for analysis equals more resources, not everyone gets the funding they want); statistical concerns (imagine the magnitude of the mathematics); it’s better to look at those that are more relevant.
Nintendo 64 (N64): Mortal Kombat 4 and Ken Griffey Jr. Baseball.
Playstation 2: Mortal Kombat: Deadly Alliance and MLB 2004.
The same as in study 1, get your heart rate measured, fill your questionnaires, play for 15 minutes, fill in more questionnaires and thank you and come again.
Number of participants assigned to gaming condition: Violent-N64 = 18, violent-PS2 = 20, violent-Wii = 15, nonviolent-N64 = 17, nonviolent-PS2 = 21, nonviolent-Wii = 17. Several participants were excluded from the analysis due to their being suspicious or having a clue to the purpose of the study. Refer to what I wrote earlier.
Video game rating results: As mentioned earlier, the authors discussed that something from the N64 made it different from the other game consoles. It turns out that participants rate the N64 games as harder to play than those who rate the PS2 or Wii games. So, a possible confound of frustration from the N64 games might’ve explained the weird results from study 1.
Analyses were again performed through ANCOVAs.
Aggressive feelings: data from the state hostility scale. Main effects were found for only game content, so only there are significant differences from the violent content of the games. No other factors were found to have any significant bearing, so technological advancement was found to be insignificant.
Aggressive thoughts: data from the word completion task. Main effects for violent content, no other factors were found to be significant in the analysis.
Physiological arousal: No significant results were found. So, there are no differences in the factors of interest. So whatever you’re playing a violent PS2 game or a non-violent N64 game, statistically there are no differences. But, I wouldn’t celebrate since I demand a more robust and reliable way of measuring physiological arousal (i.e. actually putting heart rate monitors on participants’ chests).
Immersion: data from the presence questionnaire. For the most part, no significant results were found, except for one subscale (the natural subscale which measures for how the virtual environment relates to reality) where participants in the nonviolent game rated higher scores than the violent games, regardless of game console. IMO, it’s probably the sport content that accounts for the difference since sports is something that relates highly reality in contrast to fighting opponents with magical powers.
The authors made further analyses using hierarchical linear regressions and found no significant results. So immersion in this study is not found to have a mediating nor a moderating role in the relationship between violent video game and aggression. But, I believe that current video game consoles haven’t achieved a level of immersion that might make such relationship significant and observable. So, when holographic gaming technology becomes commercially available, I’ll probably be an old geezer condemning it as it promotes violence using the same spiel of today (see Blascovich and Persky, 2007).
As it stands today, realistic graphics in violent video games has no significant relationship in individuals’ aggression. So it doesn’t matter if the graphics is crappy or beautiful in a game, what matters essentially is whether it has violence… if it has blood (Barlett et al. 2008), promotes and rewards violence, justifies violence, and restricts players to violent actions that has any significant influence in a game’s outcome (forgot source). Yes, yes there are some games that put you either in light or dark side of the force, but between that there’s violence. So it doesn’t really matter, because you’re just using different tools in combat.
Ranting off… please excuse me. Unless that violence is committed by the antagonists and the protagonists gets victimized for lots of it, much like in how Japanese visual literature works like. Of course, we’re going have to talk about identifying with the characters (Konijn et al., 2007). Hmm… I wonder if there are differences between reactive gaming aggression vs. proactive gaming aggression. Defence vs. offence? I guess it doesn’t matter. What about narrative context? What about prosocial acts or promoting camaraderie? Must play Valkyria Chronicles. Rant mode off.
The authors noted that the second meaning of video game realism (i.e. how things might actually relate to reality) has a significant bearing and this will be told in another study. For now using violent video game graphics argument in political rhetoric is not supported or inconclusive at best. The authors discussed about their immersion results and how some could argue participants’ not being immersed in video games were not paying attention and therefore affect the aggression results. On the flip side, the authors noted people who say they are not immersed in violent video games would not find it comforting it doesn’t matter because it’s the violent content that matters most.
Some limitations to write down: aggressive behaviour was not measured. Just aggressive feelings and thoughts. Results cannot be generalized to every gaming genre, IMO different gaming genres have different emphasis on graphics and how it is used.
With FPS games being the forefront at graphical realism. The fighting game genre hasn’t really much evolved as the way FPS games do since they included a lot of variety (driving tanks, flying, some stealth, using graphics to make cool cutscenes, or to create a more moody environment.) IMO, video game graphics would have a kind of interaction effect, in that video game graphics would have a significant effect from a violent FPS game than the rest of the gaming genre. Krcmar and Farrar had presented a paper that found differences in aggression and immersion between Doom 1 and Doom 3 in a conference some time ago, I’m not sure about its publication status.
Some more limitations: video game performance was not measured, they did not used the same immersion measure that was used in another study so comparisons with Ivory and Kalyaraman (2007) would be difficult. IMO, my “immersive feeling” with new video games last for several days of playing until I spot the shopped quality and notice the pixels or some of the cost-saving techniques of video game graphics (like using the same monster model with different colour patterns or attaching several lego pieces to make characters in an attempt for variety or most obvious using the same freaking voice actor voicing the enemy throughout the game). Oops another rant.
Barlett, C. P., Rodeheffer, C. D., Baldassaro, R., Hinkin, M. P., & Harris, R. J. (2008) The effect of advances in video game technology and content on aggressive cognitions, hostility, and heart rate. Media Psychology, 11, 540-560.