I once contemplated on doing a content analysis of anime characters over the summer, but the problem rested on which anime to analyze and justify (to the marketplace of ideas) why I should do such content analysis. But then, I had to do a focus group just to get a bearing on my focus on crosscultural media use.
Edward Downs (University of Minnesota) and Stacy Smith (University of Southern California) published an article in Sex Roles’ June 2010 issue about a content analysis on videogame characters’ sexual appearance. The catch is that they conducted the content analysis in 2003, and they published this study 2010… What’s going on?
This study examined male and female sexuality in video game characters. The top 20, best selling console (Microsoft Xbox, Sony PlayStation2, and Nintendo GameCube) video games from the U.S. market for fiscal year 2003 were content analyzed. The 60 video games yielded a total of 489 separate characters with an identifiable sex for coding. Chi-square analyses indicated that female characters (n = 70) were underrepresented in comparison to their male counterparts (n = 419) as hypothesized. In comparison to male characters, females were significantly more likely to be shown partially nude, featured with an unrealistic body image, and depicted wearing sexually revealing clothing and inappropriate attire as also predicted. Implications for these findings are discussed using social cognitive theory as a theoretical anchor.
Their title was entitled “Keeping abreast of hypersexuality…”, which was funny.
The authors used social cognitive theory as a theoretical anchor. Well let’s do a little crash course. Social cognitive theory posited four processes in observational learning. The first process is attentional, what grabs your attention and is usually an attractive model, therefore sexually attractive videogame characters, theoretically, would attract your attention and you are more likely to play that game, heuristically. Of course, gamers would choose games based on more complicated criteria, like word of mouth. The second process is retention, did you remember what you saw? Two key factors are relevant to videogames, the first repetition of exposure, players are repetitively exposed to the character and over a longer period of time than television would. The second is based on gender schema theory which posited that exposure to stereotyped, sexual interactions on television would form a schema of gender stereotypes (i.e. women as submissive, demure, weak and highly sexual individuals). Gamers are more likely to remember and recount cool characters, like the Master Chief or Zoey (Left 4 Dead). However, their exposure to unmemorable characters would still leave a mark on their schema of videogame women (boobs). The third process is production, this is where attitudes and beliefs reside and influence how people would behave in situations or towards certain demographics in real life. The final and fourth process is motivational, will you act out what you were exposed from videogames and the attitudes and beliefs about women?
Videogame character demographics
Referring back to Dmitri Williams et al. (2009) census data, the demographic of videogame characters is dominantly characterized as overly white adult males with some grains of diversity, age, gender and race. Other census data showed some agreements, and 20% of female characters’ appearances are overly sexual.
Following that, they had two interesting research questions: does the ESRB rating system consistently filter out depictions of overt sexuality? Is there a particular brand of video game console that displays sexualized characters more often than others?
Like I mentioned earlier, they conducted a content analysis of videogames in 2003. They picked the top-20 videogames for each console (i.e. x-box, PS2, and Nintendo Gamecube) for a total of 60. The sample included top-selling videogames according to the NDP 2003 sales data. They picked these games reasoning that these games would represent what is most frequently exposed to the player base. Their sample consisted of 22 E-rated games, 23 T-rated games and 15 M-rated games. They included a list of videogames in their appendix.
Unit of analysis and results
Since this is qualitative research involving coding media material, they used simple categorical coding schemes (Yes, I’m calling them simple and categorical).They had three gamer undergrads to play each game for 20 minutes and these were recorded on VHS tapes…They got three other undergrads to coding the videogames over the summer of 2004 (back then I think I was at Dawson College) and coded human characters (N = 489) based on several variables:
Sex and race : As expected, over 85% of the sample are male while the under 15% are female. It’s exacerbated when it comes to playable characters versus NPCs. As for race, 50% (n = 246) are white, followed by African at 21% (n= 36), Asian at 7%, then Hispanic at 3%.
Sexually revealing clothing: “any garment …to enhance, call attention to, or accentuate the curves or angles of any part of the body.” Yes, no, can’t tell or not applicable. See! That’s why I call it simple, but easy to understand. 41% of females are confirmed versus 11% among males.
Nudity: amount of exposed skin. From none, partial (i.e. bare midriff, cleavage), to full (i.e. full on or flimsy clothing). 43% of females were partially or full on nude in contrast to 4% for males.
Body proportion: they coded them on this basis: “if the characters looked like an average male or female.” Coded as either realistic, unrealistic or can’t tell. Better read up the virtual census articles, see male and female. 25% for females versus 2% for males.
Sex talk: verbal reference or dialogue regarding sex or sexual issues. Doesn’t happen that much and is therefore can’t be used in their analysis.
Sexual behaviour: sexy dancing (sorry an inside joke). Actually, it’s as characters’ behaviours that imply a sense of likely sexual intimacy, like Catherine? If only, they repeated the study with Mass Effect… no they didn’t find any sexy dancing or sex scenes.
Appropriateness of attire: Are the clothes functional and suitable with respect to context? So, bikini armour is not appropriate in a dungeon raid. 16% among females versus 2% among males.
Breast size: Pettanko, average, or boobs of steel. 25% of females were found to have huge breasts. Nothing is said about the flat chested.
Waist size: disproportionally small, average or disproportionally large which were recoded in the analysis as either small waist versus not small waist. 40% of females versus 1% males do have small waists.
To answer the research question: “does the ESRB rating system consistently filter out depictions of overt sexuality?” The authors said that they did a fairly good job, but the depictions do show up in small numbers in E-rated games. But looking at absolute numbers, they are found more often in the teen or mature rating. Nudity was the only significant variable where it was more likely to be found in T or M-rated games than E-rated games. Looking at sex differences, the numbers were so low (n < 5) that it’s probably worth looking at the body census studies.
Research question 2 asked if a console showed more sex than the others. The answer is kind of convoluted. Nintendo’s Gamecube have more characters (31%) that wore sexually revealing clothing than the Xbox (12%) or the PS2 (10%). I expected the Xbox to be the culprit since I viewed it as a “frat boys” console. However, I’d like to point out the numbers are driven by the sample and perhaps the coding schemes. Take a look at the table below.
Clearly the Xbox boast the most sexualized female characters than the other consoles. But, Dead or Alive: Extreme Beach Volleyball is among the Xbox games sampled. So, the game probably accounted half of the variance in women’s appearance in the Xbox category. As for the PS2, I’m thinking of Xenosaga: Episode 2 and Dynasty Warriors 4. What’s surprising is that the sexually revealing clothing for the Gamecube was driven in the men’s section at 33 characters wearing something sexual. I looked over the list and I can’t think of which game that drives such numbers? Is it Def Jam Vendetta? The Sims?
What was the take home message again? Ah right, videogames aren’t representing female characters in a sensible manner, they are overly sexualized. Do something, you’re raising men’s expectations of the female physique and you’re probably hurting female gamers’ body image, like television does.
The authors noted limitations. For starters, it’s a content analysis and the inferences are driven more by theory and previous similar studies. Second, they wanted to check out the cutscenes, especially the ending scenes since they reward the player with something, and that something might be sexual. Additionally, they suggested looking at the range of portrayals, acts and behaviours in games.
Here’s an interesting quote about the PS3 and the Xbox 360’s high price:
These price points suggest that the Xbox 360 and PS3 are catering towards an older audience who can afford a more powerful game console. If this is the case, it would not be surprising if more sexual content were included in games for these consoles or if more companies produced genres of games for these consoles that included more sexualized game characters.
First, I’d like to point Nintendo’s history towards family-friendliness and the directions it took with the Wii towards casual gaming. The price has nothing to do with age, really. It’s more about loyalties towards the consoles and its affiliating game series. For example, the Final Fantasy series since FF7 has been on the Playstation. Halo is Microsoft property and is unlikely to cross over to the Wii or the PS3. That said, age has nothing to do with it. Will there be more sexualized characters in those consoles? That depends on the big league developers. At present, a trend towards sensible character development seemed to be on the way. Again, I tend to believe it is the player base that decides what content they want in their games and they are probably very vocal of the games they play.
Here are some of thoughts that nearly ran away from my head while reading the article. The coding of whether a sexual image is thereor not needs further study with more nuances. In particular, filming techniques should be examined in how games depict characters. Think of the camera angles that one would focus on, in some good games they might show a highly sexualized women, but her physique may not be the focus of the camera whereas the breast or panties has more screen time than the owner’s face (I’m looking at you, Tomonobu Itagaki). Maybe even a cut could interact where a scene changes from something innocuous to breasts or perhaps a nudity scene.
The authors argue that the active interactivity should intensify the learning processes as posited from social cognitive theory. Well, I’m a bit sceptical about that. I argue is the procedural elements (i.e. interactivity) that has factors in it, does playable female characters have a greater effect than NPCs or even sidekick roles, does camera controls allow players to ogle on the physique, does the game allow some kind of sexually implied activity. For example, a spell that basically renders Bayonneta nude for a couple of seconds. Furthermore, I argue it’s not fair that because female characters are expressing such overt sexual depictions that it’s potentially harmful towards sexual attitudes. How the player, the game, the creator and the character deal with this issue sensibly or not is more powerful in forming sexual attitudes. Fortunately, the issue of sexualized appearance is gaining momentum as I noticed in some developer blogs (psychochild blog), comics (nerf now, extra credits, Sexy videogameland). Perhaps, gamers are tiring of seeing such boyish characterization of women in videogames and want to see more mature/sophistication or that constant warnings from sex role media experts have started to rub off on everyone (unlikely).
Downs, E., & Smith, S. (2010). Keeping abreast of hypersexuality: A video game character content analysis. Sex Roles , 62 (11), 721-733-733. doi: 10.1007/s11199-009-9637-1