Blood and point of view in violent video games and its effect on reactive aggression (Krcmar & Farrar, 2009)

This post is a part of series of study reviews written by professors whom I have an interest in joining as their grad student.

Marina Krcmar, Kirstie Farrar and Kristine Nowak are being considered to be good advisors when I’m applying for grad school for this fall. Their names have shown up a lot in my research library, their research leans pretty much on the technical aspects of video games and this study reflects such leanings.

In this study, they’re looking into two aspects: point of view and the presence of blood in violent video games.


In this study, an experimental design was utilized to test, first, the effect of a violent game versus a no game control on physical and verbal aggression and retaliatory aggression against a confederate. In addition, the effects of two internal video game manipulations were explored. Overall, those in the violent game condition were more verbally and physically aggressive than those in the no game condition. In terms of internal game features, third-person play with the blood on, especially when combined with aggressive cognitions and to some extent, hostile affect, encouraged more aggressive outcomes.

The Association for Education in Journalism and Mass Communication has post on it and a link to the full-text study.

Going through their introduction of the study, it’s mostly the same thing you’d read in any study regarding violent video games effects on aggression: the General Aggression Model, previous studies on violent video games, the theoretical mechanism, such as arousal, cognitive and affective. Stuff that I shouldn’t skip, but I do.

Well, I’ll just write some highlights they wrote as to why violent video games increase aggression:

  • Players have an active role in a storyline in contrast a passive role in television.
  • It can be viewed as a learning environment since you are reinforced to which in-game actions helps you advance in the game and which doesn’t. One of the reinforcers which I find outdated in today’s mature video games is the usage of points. I don’t recall any recent critically acclaimed video games that possess a scoring scheme. I am pretty sure any junk games do have a scoring scheme, but then again such games appeal to the lowest common denominator of the population.
  • Character identification, mainly because video games have players assume the role of the hero. In particular and perhaps with greater effect, when playing first-person shooters where you literally see through the eyes of the hero.

They selected two video game features in this study: Point of view and the presence of blood.

Why point of view? Referring to the third point, it is possible that looking through the eyes of a soldier shooting enemies might make one feel more involved and more likely to identify with an aggressive character. To paraphrase their words. But, their previous study in 2006 makes the theory sound uncertain.

Why blood? In television, blood serves as an aversive cue that violence is bad which lessens aggression. In video games, it acts as a rewarding cue which increases aggression. I have some reservations about blood being the sole cue for aggressive in-game behaviours.

They also wanted to test out the General Aggression Model and see how it relates to video game features. Specifically, they wanted to see if the three aggression routes (affective, cognitive and arousal) might have a moderator or mediator role in the relationship between video game features and aggression. To explain, if you are playing violent video games and you are in a angry mood, or thinking of kicking your boss for being an ass or he’d just made you pent-up full of emotions, then you are more likely with the violent video game to be more aggressive (in all aspects of it, verbal, physical, etc.)

Their argument is that if a video game feature does increase aggression, then from theoretical point of view, it acts as a mediator since in-game violent acts must go through said feature which influences the relationship (increase or decrease) and that the three routes would therefore act as moderators. Why? because violence is present in both video game features. I’m a having a little trouble understanding this, so if there’s some kind of visual aid. I would appreciate it.


Participants: 186 undergrads (97 females, 89 males) from an intro to communications study class. Average age: 19.67, mostly white. The authors already addressed the concern that these participants might’ve been exposed to the violent video game debate or had any lectures about media effects, but thankfully the data was collected at the beginning of the semester, most of them aren’t even majoring in communications, so they’re good.


Video game ratings: participants rate the game they played in the experiment based on 9 criteria on a 7-point scale (i’m guessing on the scale). Easiness, enjoyable, frustrating, violent, how much blood and gore, pace, engaging, how much it maintains their attention.

Demographics: nuff’ said.

Video game exposure: Asked whether they had the game before (17 % do and it’s been taken care of). Asked how often they played different video game genres (i.e. first-person shooters, RPGs, etc.) on a 7-point scale.

Modified Buss-Perry: measures aggressive behavioural intentions. It’s modified from the original in that participants read an ambiguous story that might have been interpreted as an aggressive provocation. Using the reworded 29-items from the Buss-Perry questionnaire, participants were asked (on a scale of 1 to 6) on how characteristics of themselves they would react to the situation (e.g. I would hit this person.). The items are categorized on four subscales: verbal aggression, physical aggression, temper and resentment.

Confederate provocation paradigm: measure behavioural aggression. Participants first complete a filler questionnaire about “current events”, the questionnaire was designed to be very hard to answer. Then a female confederate comes in to tell the participant there’s one more thing to do and they must go to another room. Both participant and confederate walk to the other room, meanwhile the confederate looks at the filler questionnaire and insults the participant by saying their performance was terrible in comparison to other participants. When they are in the other room, the confederate tells the participant to evaluate her and that their answers would be confidential and would help determine if she would get funding. Of course, the participant must not suspect that this is all an act. This paradigm has been used before.

The participants’ evaluations of the confederate consist of 7 items answered on a 10-point scale. The important part of the evaluation is the funding items (2 items) which is used to measure behavioural aggression.

These aggression instruments are used to measure retaliatory aggression or reactive aggression. It’s a type of aggression in response to a harmful act (perceived or real, e.g. being insulted or humiliated, being yelled at by your boss or being punched by an asshole.), so this study (IMO) does not generalize to other types of aggression, say proactive aggression where a person initiates an aggressive behaviour.

Speed of association test: measures aggressive cognitions. Participants are told to read a word and write down the first word that comes to mind. There are 50 words in the test.

State Hostility scale: measure aggressive affect. 15-items answered on a 4-point scale.

Video game used: Hitman 2 on the Playstation 2. The game was selected because of the available options for changing the point of view and the presence of blood. Something tells about using the Playstation 2 for FPS gaming. I’ll discuss this issue in later.


Participants are randomly assigned to either a no-game condition or a game condition. Those in the game condition (n = 148) are further randomly assigned to play one of the four conditions. Either they play in first-person or third-person, with or without blood. Play time is 12 minutes. The participant play the game in god mode to control the differences in skill among participants, although this may raise some methodological concerns, but this is better than letting participants feels frustration from player death. After play, they undergo the confederate provocation, after that they complete the rest. They are then debriefed.


Their first hypothesis was that those in the violent game condition would be more aggressive than those in the no-game condition and that the relationship is mediated by aggressive cognition.

Using ANOVA to test this hypothesis, they found significant main effects in that those in the video game condition had scored statistically significantly higher on verbal aggression and physical aggression from the modified Buss-Perry measure. However, when it comes to behavioural aggression from the confederate provocation, it seems that those in the no-game condition gave significantly lower funding rating than the violent game condition. The authors contend that the no-game condition were likely to be irritated at not being able to play a video game and took it out on the confederate. No funding for you (on average)!

To see if cognition acted as a mediator, and this is where I got lost on how they did it, they found that aggressive cognition did not have any statistical significance as a mediator. So, violent video game play has a direct effect on aggression.

Six regression analyses were conducted to investigate point of view and presence of blood, separately, and aggressive cognition as a moderator using verbal aggression, physical aggression and funding rating as dependent variables. Gender is taken into account and the no-game condition was excluded from the analyses.

Verbal aggression and POV results: it became significant when there’s an interaction between POV and verbal aggression. So those in the third person condition with high scores on aggressive cognitions reported higher scores on verbal aggression. No other significant differences were found.

Physical aggression and POV results: Gender differences were found, obviously males were more aggressive, but let’s put that aside. Again, an interaction effect was found. Those in the third person condition with high scores on aggressive cognitions reported higher scores on physical aggression. But, looking at the tables in the paper, I don’t see the numbers indicating a difference. Little help here….

Funding rating and POV results: Gender differences were found, no details given. Significant interaction effect was found in that those in the third person condition with low scores on aggressive cognitions gave higher funding rating than the rest of the conditions. No other significant differences were found.

Verbal aggression and blood results: Statistically significant interaction effects were found when those in the blood condition with high scores on aggressive cognitions. No other differences were found.

Physical aggression and blood results: Just a gender difference.

Funding rating and blood results: Just a gender difference.

Next in their analyses is their examination on whether aggressive affect has a mediating role. Again using verbal aggression, physical aggression and funding rating as the dependent variable.

No significant results for verbal aggression and physical aggression. So, aggressive affect has no mediating role in the relationship. However, affect seems to have a statistically significant effect on funding rating, so a partial mediation effect.

Next is mixing in point of view and presence of blood with aggressive affect as a moderator using verbal aggression, physical aggression and funding ratings as dependent variables. Only the game conditions were included in the analyses. Analyses conducted with regressions.

For point of view: their results found no moderation effect or interactions from aggressive affect. Nada.

For presence of blood: No moderation effect or interactions from aggressive affect for physical aggression and funding rating. A significant interaction was found for verbal aggression: those in the blood condition with high scores on aggressive affect reported higher scores on verbal aggression than the rest. Those in the no-blood condition with low scores on aggressive affect reported lower scores on verbal aggression than the rest of the conditions.

Wait a minute, what happened to the analyses that mixes point of view and presence of blood together? I don’t see it in the paper or is combining the separate analyses into one in the discussion section a valid method?


After a coma-inducing read of the results section, I can finally rant!

A reminder is that reactive aggression is being studied, so behaviours after being “harmed”, in a hypothetical situation and in real life situation (staged, of course).

Overall, participants in the video game condition were more verbally and physically aggressive. Examining further with point of view and the presence of blood, with aggressive cognition and affect as a moderators in the relationship. It seems that playing in bloody third-person violent video games would increase our aggression, depending on what we’re thinking at the time. For example, you are thinking of what’s your mom cooking for dinner or you want to frag that player who insulted, the likelihood of resorting to violent outbursts or punching your monitor decreases or increases, respectively. It should be noted that aggressive cognitions does not necessarily lead to aggressive behaviours, aggressive cognitions is like adding weight on the behavioural aggression scale, but you need more factors until you reach a certain threshold.

On the other hand, how we feel would noticeably affect on how we would verbally aggress others depending if there’s blood or not. So you are calm or angry and if there’s no blood or blood, then the likelihood of aggressive verbal retaliation decreases or increases, respectively.

So why third-person games, and not first-person games, increase aggression? The authors argue that first-person play is more of no-person play since players are only playing with an arm that interacts with different sorts of weapons. Considering that Hitman 2 is used in the study, the degree of movement in either first-person or third-person is quite limited. I argue a more advanced game, say F.E.A.R. series which features a more vivid and fluid physicality of the protagonist might increase character identification and therefore (indirectly) aggression. For example, if we see the protagonist’s arms, legs and maybe chest. Mirror’s Edge might be a good template for increased identification since body movement needs higher player input and coordination, and hearing from initial reviews (of which I forgot to save some good links, but googled), such gameplay was rather unique, frustrating and yet induce motion sickness.

Another thought is perhaps that visual output from third person games is richer than first person. So player input in the game appeared more obvious and instantaneously. Therefore, players would feel more immersed since the virtual world appears contiguous to their perceptions. Attention might be related in that players in first-person games needed more focus on feedback of their actions and has less attention of other gaming features, say exploding bodies. It might be interesting to examine this idea using a first-person violent movie… The authors suggested that experience and familiarity with the medium might be a factor, well it does seem to go along with cognitive load since experienced players would have developed automatic scripts for various tasks, like switching weapons or emoting, and perhaps allowed them to make the gaming experience as natural as possible.

Observational learning is another possibility for the increased aggression. The authors argued that learning and imitation occurs if there’s character identification; where players can see their avatar’s act being performed so they can clearly see how one slits another one’s throat and therefore it’s much like the bobo doll experiment; and if there are reinforcements, such as blood as reward cue. Of course, I still doubt blood is the sole rewarding cue for violent in-game behaviours.

Another thing is the input/output routes, first-person shooters don’t work well on consoles. They started in computers and works well with keyboard and mouse. I heard that many FPS games had to include auto-targeting options because gamepads can’t even accommodate the complexity of FPS games (okay the last part is just big talk). IMO, having played Red Faction 2 on the PS2, looking around appeared clunky. However, it might’ve been unnoticeably clunky because the video game rating did not say there were differences between first- and third person games.

As for blood, well I do want to know why it only affected verbal aggression and not the other types of aggression. I got nothing in mind. Aside that it provides some support for previous studies and such.

All of the above theories need to be tested, since this kind of research is lacking in the literature. The authors suggested some future studies like comparing expert vs. novice players using the same experimental paradigm. Speaking of which, why not do a reverse experimental paradigm, like using the avatar machine?

The authors noted some limitations like whether “god mode” affects players’ cognitions by removing the competitive element of the game. Whether they should’ve used a male confederate instead of a female one. Something about the participant’s gender and their avatar’s gender affect the immersion factor (better reread Eastin’s work).

All in all, it’s an interesting study and I can see how these authors are trying to do in the video game research literature. Creating an aggression model that specific to video games.

Krcmar, M., & Farrar, K. (2009). Retaliatory aggression and the effects of point of view and blood in violent video games. Mass Communication and Society, 12, 112-138.


2 thoughts on “Blood and point of view in violent video games and its effect on reactive aggression (Krcmar & Farrar, 2009)

  1. As read in the handbook of computer games studies (p.238), the “subjective” perspective makes the first-person identification of character rather ambiguous because the player’s self-concept is “leaking” into the game and thus separates the character from the player. Identification fails because of the ambiguity.

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 )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s