Games User Research Summit Poster Presentation

This past week, I was fortunate enough to attend the annual Game Developers Conference (GDC) as an Xbox Women in Games ‘Game Changer.’ I was able to connect with a lot of amazing folks in the video games industry, including the various researchers, analysts, and data scientists associated with the field of games user research. Since my own research in academia overlaps with the field, I attended the annual U.S. Games User Research Summit, held on the Tuesday of GDC, where I also presented a research poster showcasing the results of a gameplay experiment.

Me with my poster at the GUR Summit!

I’ve elaborated on the pilot study results in this blog post, which highlights the implications for how female game characters are perceived by female players when the same character is portrayed in sexualized clothing or non-sexualized clothing. In a nutshell, the same character portrayed in sexualized clothing was perceived as more incompetent than her non-sexualized portrayal. This finding was found in an action-horror game where female users played the same level of the same game for 15 minutes each. Using the same game character within the same video game (and game level) helps to minimize any confounding variables introduced by varied gameplay strategies and enemy encounters, which have not always been avoided in other similar studies (e.g. Behm-Morawitz & Mastro, 2009).

The second study outlined in the poster utilized the same female characters (but a different set of female participants) and asked questions about the player’s self-perception following gameplay. Participants completed a 2 x 2 (sexualization x game difficulty) experiment. Before participants arrived to complete the experiment, they were randomly assigned to the sexualized or non-sexualized character attire and a casual or normal game difficulty. In short, no differences were found between any of the four conditions with respect to player’s self-esteem, self-efficacy, and body image. You can view the full findings (in condensed poster format), as well as my thoughts on the study’s implications, in the below PDF version of my poster.

 

 

Admittedly, this poster is a condensed form of a research project where I measured additional variables (e.g., user’s self-discrepancy; identification with game characters), yet the additional variables did not change any of the outcomes. As such, I focused on the variables most likely of interest to an audience largely embedded in the video games industry. If you’d like to read the full paper, feel free to get in touch and I can send you a copy.

Since the study produced null findings, I will likely conduct a follow-up study with a different game and different characters to see if I can replicate the results (or not!). Either outcome would be interesting to discover.

If you’re curious as to why I limited the sample to self-identified female players (which I was asked about a couple times at the Summit), that is simply because I’ve been interested in women’s experiences with digital games as part of my broader research agenda as a PhD student.

Citations

Behm-Morawitz, E., & Mastro, D. (2009). The effects of the sexualization of female video game characters on gender stereotyping and female self-concept. Sex Roles, 61(11-12), 808-823. doi:10.1007/s11199-009-9683-8

Skillful or Incompetent? How a Video Game Character’s Sexualization Affects Their Perceived Skill

One of my over-arching research interests involves how the appearance of video game characters effects the player-experience, as well as how identification with game characters effects self-perception on social-psychological dimensions.

Last fall, I started a pilot study (think of it as a trial or test-run) for an experiment where college-age, self-identified women played a video game where the character was either dressed in a sexualized or in a non-sexualized outfit. Essentially, the pilot’s main purpose was to confirm that my experimental conditions were reliable, or to put it another way, perceived as consistent when multiple people did the study. Because I was interested in exploring the effect of a game character’s appearance on self-perception, I wanted to be sure that people would consistently rate the sexualized character as, well, sexualized, and that the same character dressed in casual attire was perceived as non-sexualized. Thankfully, my pilot study confirmed this, and I was able to launch the full experiment in January.

While I did eventually finish the full experiment in April, I’ve yet to actually sit down and examine the results (but soon, after I finish analyzing my character design interview data!). However, I did find a somewhat unexpected – but nonetheless interesting – outcome from the pilot that sheds some insight on how a sexualized appearance of a game character influences their perceived skill.

The Pilot Study

For my experiment, forty-three undergraduates played the game Resident Evil: Revelations 2. All participants self-identified as female and the majority identified as White/Caucasian (81%). The average age of players was twenty years old. Everyone in the study played the Story Mode portion of the game for fifteen minutes. Following game play, participants filled out a questionnaire about the video game and the playable game characters, Claire and Moira.

As mentioned above, the pilot tested whether my conditions were reliable. Each participant was randomly assigned to either play as Claire and Moira in their sexualized or non-sexualized attire. Resident Evil: Revelations 2 was perfect for testing these two conditions because the story can be played with either characters’ default or bonus costumes (the player can switch between Claire and Moira in the story mode, as they both appear onscreen at the same time). For the purposes of my study, their default costumes were used for the non-sexualized condition. Claire’s bonus Rodeo costume and Moira’s bonus Urban Ninja costume were used for the sexualized condition.

Left to right, top to bottom: Moira (default/non-sexualized attire), Moira (bonus/sexualized attire); Claire (default/non-sexualized attire), Claire (bonus/sexualized attire).

After game play, everyone was asked a series of questions about the characters. I wanted to ensure that the characterizations were consistent despite the changes in attire. The only difference I hoped to find in the study was for differences in sexualization between the default and bonus attire conditions (which I did find). In addition to several questions about the characters’ attire, both characters were assessed based on six adjectives using a 7-point semantic differential scale. The characters were rated on the following: attractive/unattractive; strong/weak; aggressive/submissive; violent/passive; skillful/incompetent; good/evil.

No significant differences between the two portrayals were found in terms of Claire’s and Moira’s attractiveness, strength, aggression, violence, and moral character. However, significant differences were found for Claire’s skill, in which participants rated sexualized Claire as more incompetent  (i.e. less skillful; Mean = 2.70, Standard Deviation = 1.75) than non-sexualized Claire (M = 1.63, SD = 1.01). This outcome was not the case for Moira.

This difference may have emerged because Claire is the main playable character who is capable of attacking enemies with a gun whereas Moira can only attack enemies with a crowbar. As such, most participants played as Claire for the majority of the time. The difference in Claire’s rated skilled by attire suggests that players may have deemed her sexualized attire as impractical for fighting zombies. This may have influenced Claire’s perceived competency as questionable given the game’s context in which a lack of clothing seems like poor judgement.

Costume and Context Matter

Although I was not anticipating any differences in Claire’s skill between the sexualized and non-sexualized attire conditions, differences did indeed emerge. What does this tell us? I think it stands to reason that the sexualization of female game characters as an expression of empowerment will not always hold for specific contexts. Given that Resident Evil: Revelations 2 is an action-horror game where the main characters are kidnapped and imprisoned against their will, seeing Claire in a state of relative undress likely enhances her vulnerability in the situation which could have an effect on her perceived skilled, or ability to handle the circumstance.

Claire’s sexualization had the effect of diminishing her perceived competency. Given that players were fighting zombie-like enemies in a run-down facility, her attire may have conveyed a lack of sensible judgement on her part. Given the outcome, I think it demonstrates the importance of considering how sexualization of a game character is interpreted within the context of gameplay. When a game presents a situation where a character is vulnerable and fighting for their survival, as with Resident Evil: Revelations 2, a sexualized appearance may influence whether that character is deemed competent enough to handle the situation. This is especially important for game designers to consider when creating a character they want to portray as skilled and competent, despite the overwhelming odds stacked against them.

Obviously, it should go without saying that a different game and a different character may produce different results. Furthermore, participants in this study went into the game ‘blind’ and were unaware that the sexualized Rodeo attire is a bonus costume not intended as the default attire worn by Claire in the game’s story mode. Her default attire in the story is non-sexualized, and it is a player’s choice which costumes they want Claire and Moira to wear in the game. Understanding this context, as well as enabling a choice of costumes, could also lead to different results. Expect more nuanced findings once I do the data analysis for the full experiment, in a few months!

UX Playtesting Analysis: Tori

Several weeks ago – not long after I returned from my first Game Developer’s Conference – I realized something important: I’d like to apply my social scientific research skills and knowledge of media theories towards games user research. I’m fortunate enough to be in an interdisciplinary Media School where I’ve established a good rapport with many of the faculty members in the Game Design program as an Associate Instructor and the Lead Intern for Women in Game Design. After considering that I had certain skills that might be useful to the student game developers in the program, I pitched an idea to Professor of Practice Mike Sellers: his game design students in Workshop I, the first major game development course required for all students pursuing the B.S. in Game Design, could benefit from UX analysis and thorough playtesting feedback. So why not receive a little help from a Ph.D. student eager for more experience in conducting this kind of research?

Prof. Sellers agreed that I could offer useful insights and a collaboration was born!

The first student game project I’ve playtested is Tori, a meditative music exploration game in which players assume the role of a small bird collecting sounds from a stylized environment made up of several small islands. I recently sat down with Joseph Adams, a developer on Team Tori, for my first playtesting session with Tori where I offered some feedback on general usability, discussed below.

Gameplay and Feedback

Currently, the ‘bubbles’ which represent the aura of sounds for different objects in the environment are all colorized the same – a pale yellow color. These sounds can be collected by the player and dropped, in bubble form, to a new location. The collection and re-placement of sound-bubbles is an integral mechanic of the game’s progression system.

Collecting sound bubbles in Tori

I suggested that the different objects, and thus sounds, might be color-coded, so that when player’s drop these sound-bubbles in new locations, they might more readily recall which bubble represents the sound attributed to a rock, bell, lantern, or other unique objects in the world. A color-coding scheme might be integral further into development, as players will be tasked with dropping sounds in new locations in order to replicate a tune provided by a non-playable character, in order to unlock new islands to visit and explore. As is, the placement of bubbles that are all the same color, especially when the player is expected to imitate a tune using the collected sounds, could get a bit confusing. Currently, there is no feedback that identifies what object the sound-bubble represents once it has be removed from the object and placed elsewhere in the environment. Color-coding the sound-bubbles to be unique based on the object of origin could simplify the identification process.

Particle effect on wingtips to signify increased velocity

I also noted that additional audio-feedback might improve player experience. Currently, when players fly downward, the speed of flight increases. This interaction is paired with visual feedback of a particle effect trailing from the tip of the bird’s wings – a ‘speed trail,’ perhaps. It’s pretty and conveys the message, but might be paired with an accompanying sound effect, such as rushing wind, to match the visuals and reinforce the feedback.

This particle effect was misinterpreted by a player during a recent playtest. I took notes, along with student developer Joshua Smith, during a session with university students who played Tori for the first time. We invited the players to voice their thoughts as they played, and a participant misinterpreted the ‘speed trail’ as visual feedback to signify when the bird is carrying a sound bubble. Thus, playing an accompanying sound effect (e.g. rushing wind) might reinforce the visuals as the effect of speed during flight.

All in all, Tori is a simple and beautiful game that shows a lot of promise for its unique aesthetics and features. I look forward to conducting more playtests with new users and discovering new ways to enhance the project’s usability and player experience!

Follow Team Tori on Tumblr & Twitter for more updates. Many thanks to Prof. Mike Sellers and Team Tori for collaborating with this video game researcher.

Re-post:‘Through the Gates,’ ep. 31: Jess Tompkins on women in gaming

Indiana University’s Office of the Provost and Executive Vice President featured me on a segment of their weekly podcast, ‘Through the Gates.’ You can give it a listen on SoundCloud. Let me know what you think in the comments.

Tompkins describes her life-long enjoyment of video games and the gender politics that manifest in the game design industry and gamer culture. She explains that, in her early years, the sexualization of women in games “seemed normal,” saying, “I was being exposed to sexualized women in advertising, in films, long before I saw sexualized women in games.”

-From ‘Through the Gates’ official blog.