Methods in Human Computer Interaction/Project Proposal Turing

Introduction

For as long as there have been video games, there have been concerns about unintended psychological consequences of video games for players. Even during the mid-1980s, when video games were of a complexity and sophistication barely on the same scale as modern gaming (the original Super Mario Bros. was released in 1985), concerns were already being raised about the impact that such games might have on the development and education of young players. As video games and gaming culture have evolved, so too have the accompanying social and psychological concerns. A large body of research has arisen to address concerns about the real-world consequences of graphic violence and female objectification in modern gaming.

Of interest to the present study is how visual information from gameplay in popular video games might be processed by the user in ways that promote negative stereotypes of women in general, and promote a narrative of women as sexual objects more specifically. For instance, Fox, Bailenson, & Tricase (2013) conducted a study in which they recruited female participants to don a head-mounted display (HMD) and view themselves as avatars in a virtual world. The researchers manipulated the attire of the avatars so that outfits ranged from a fairly conservative style of dress to a significantly more suggestive style of dress, and then measured how the avatar’s appearance affected participants’ body-related thoughts and acceptance of rape myths. The results of the study indicated that participants whose avatars had a more sexualized style of dress reported significantly more body-related thoughts than their counterparts with more conservatively dressed avatars. Further, participants in the sexualized condition whose avatars used the participants’ own face reported greater rape myth acceptance than when the avatar had a computer-generated face. The authors concluded that participants’ reactions to their avatars conformed to the so-called Proteus effect, which is the idea that people will modify their attitudes about themselves based on observations of their own actions. The implication is that female players will tend to internalize the appearance and behaviors of their virtual avatars in video games, and this can be detrimental if avatars depict femininity in a demeaning light.

Subsequent research has emphasized that both men and women share attitudes and beliefs that predispose them toward video game sexism. Fox and Tang (2014) recruited a large sample of both male and female participants and measured how their scores on scales of lifetime video game play, conformity to masculine norms, social dominance orientation, and empathy, correlated to the scores on the Video Game Sexism (VGSS) scale. Across all ages and sexes, high scores on the masculinity norm of power of women, the masculinity norm of heterosexual presentation, and social dominance orientation were found to be significant predictors of sexism in video games. Thus, the risk is not simply that female players will tend to internalize negative depictions of femininity in video games, but also that men will have negative attitudes toward women reinforced by the same depictions.

It is considered gospel among many researchers of compliance and persuasion that it is easier to shape the behavior of individuals using superficial cues (i.e., the peripheral route to persuasion) than through direct arguments (i.e., the central route to persuasion) . In other words, the more individuals are motivated to attend to or elaborate on information they are presented with, the less vulnerable they become to persuasion. Thus, the extent to which players’ attitudes toward women in the real world are affected by the depiction of women in video games may depend on the amount of processing that goes into evaluating these depictions.

Research Question

One interesting question, then, is the psychological depth at which images of scantily clad female avatars are processed by players. This point may be critical to understanding the depth to which resulting negative attitudes toward women are held, and useful when attempting to design interventions. One possibility is that players process the sexualized depictions of female avatars at a very superficial level. Specifically, players might make more positive or negative evaluations of their game experience based on the how revealing they judge their avatar’s body armor to be, but fail to interpret relatively less armor as a strategic liability. On the other hand, players may report finding a game more or less pleasant based on the sexualized depiction of a female avatar wearing less armor, but they may also interpret less armor as making their avatar more vulnerable during a confrontation, and revise their expectations of performances in such confrontations accordingly. This deeper level of processing would be indicated if, given similar combat scenarios, players reported a disproportionate level of satisfaction for victories in which their avatars wore less armor (indicating that the character had won despite greater vulnerability) than for victories in which their avatars wore more armor, and this should be true irrespective of the gender of either the player or their avatar. We thus hypothesized that wins involving game play with a character wearing less armor would result in higher ratings of self-reported performance than wins involving a character wearing more armor, irrespective of the gender of the character. Correspondingly, Losses involving game play with a character wearing less armor will result in higher ratings of self-reported performance than losses involving a character wearing more armor, irrespective of the gender of the game character.

Our participants will be recruited from an undergraduate research pool with minimal restrictions. This means that we will likely be getting players from both sexes who span a wide spectrum of gaming skill levels, and who come from a variety of cultural backgrounds that may embrace different attitudes toward sexuality and gender roles. This will further allow us to see how the results of the above comparisons might be affected by player gender, player skill level, and player score on a gender attitude scale. We will therefore collect data concerning game performance as an indication of skill, as well as collecting demographic data and behavioral/attitude measures to be included in our analyses.

Methods

Participants

Students will be recruited from introductory psychology courses at a large Midwestern university. They will receive course credit for their of participation. They will have other options to receive course credit outside of participating in research. Students are representative of a young adult population in Midwest United States.

Experimental Measures

Implicit Attitude Test

The gender IAT will consist of seven tasks. All task consist of participants reacting to stimuli presented in the middle of a computer screen. The participant will either press the “e” key to indicate the stimuli belongs to the group on the left or the “i” key to indicate the stimuli belongs to the group on the right. The participant is asked to use their left hand to press the “e”’ and their right hand for the “i” key. In the first task, the participant will sort the faces of people in “Man” or “Woman.” The sides in which “Man” or “Woman” are sorted is counterbalanced across participants. In the second task, participants will categorize a group of words into positive or negative categories. Positive words; good, joy, love, peace, wonderful, pleasure, glorious, laughter, and happy. Negative words; bad, agony, terrible, horrible, nasty, evil, awful, failure, and hurt. The sides in which positive and negative words are sorted is counterbalanced across participants. On the third task, participants complete a combined tasked where one side can be for Man/Negative and the other Woman/Positive. In this task, participants should respond to woman and positive words with the same key. While responding to men and negative words with the other key. The fourth task is a repetition of the third, but the participant will see a wider variety of words and faces. The fifth, sixth, and seventh tasks are repeat of the first, third, and fourth tasks respectively, however, the sides “Man” and “Woman” are sort to is swapped. Hence, if men were sorted to the right in the first, third, and fourth task they will now be sorted on the left side.

In the IAT, implicit attitudes are measured by difference in reaction times, and correct response in sorting. A participant with negative implicit attitudes towards women will show slower reaction time and more errors when sorting positive words and women into the same than when sorting negative words and women together. An implicit attitudes test is required, because participants are often unaware of their own attitudes, or unwilling to disclose their bias. It is important for us to measure the implicit attitudes of our participants to control for possible pre-existing negative attitudes towards women. This control will allow us to isolate the effect on players’ performance to be from our manipulation as opposed to previous attitudes.

Eye Tracking

Eye Tracking software and hardware will be used to measure the location of the player’s gaze. The eye tracker will let us record how much time the player spend looking at their environment as opposed to their avatar. These times can be used to determine if one avatar has receive more perceptual attention than any other. Knowing how long players focus on their avatar can also provide context to differing levels of in-game performance.

Video Game

We will use Elder Scrolls 5: Skyrim as our game manipulation. This game was chosen as it provides a variety of tools and modification for researchers to control and record games variables. The game is also an exemplar of popular role-playing games. It has sold over 20 million copies in the first year of its release. The game allows user to personalize their avatar by giving players control over the clothes worn, the appearance of the avatar, the race of the avatar, the gender of the avatar, and the decisions of the avatar. For our study, the player will not have control over the appearance of the avatar. Additionally, the player will be locked into third person camera view in which the player can see both the avatar and their immediate surroundings. Regardless of condition, all players will only have access to a standard shield and sword equipment as weapons for their play session. Player equipment, regardless of condition and appearance will offer the same statistics to the player. Their play session will be limited to one of the game’s cave level. In this cave, the player will face a variety of game enemies of increasing difficulty. The player will fight against 2 low level enemies in a small room, fight 4 low level enemies consecutively in a winding corridor, 4 low level enemies and 3 medium level games in a large room, and finally 4 low level enemies, 3 medium level enemies, and 1 high level enemy in a multi-level fortified position.

Recorded in-game variables will be used to measure the performance of the players. Recorded variables will include damage dealt by the player, damage dealt to the player, amount of hits blocked by the player, and time taken to complete the cave. Time to completion can be a measure of mastery and how easy the player finds the game. However, by itself it can be incomplete as player can have different styles of playing, or they can just like looking around the cave and exploring the space. Hence, damage dealt by the player, to the player, and weapon mastery via blocks will also provide another dimension or game performance. All these variables are automatically recorded by the game and accessible in an in-game menu.

Avatar Design

Overall Appearance

In game, avatars are meant to be some sort of creation by the player. In order to control for differences in perception of the avatar’s basic phenotype we decided to standardize the appearance of the avatar. Within the psychology program at Iowa State University (the population from which many participants were selected) The demographics reported by the university aided in creating an avatar that could be perceived as being the most representative. In 2014, out of the 672 total majors, only 119 were classified as belonging to any minority. As such, it was determined that having a paler avatar would be more in line with the sample that would be pulled from psychology classes. Hair color design was set to a medium blonde/brown. While this trait is polygenetic (i.e. multiple genes code for the trait of hair color), the dominant allele codes for brown; an ambiguous medium hair was chosen to best capture the variation within the potential sample. Facial design was of least concern due to the fact that players would interact with the game from a third person point of view.

Armor Modification

Elder Scrolls 5: Skyrim is the base upon which avatar and game design is built upon. That being said, there are no armor sets for warrior classes that are revealing enough to be classified as a low armor condition. Bethesda Game Studios (the developer of the Skyrim games) has made the code available for players; this permits players/independent developers to modify the in game world. It is with such a modification (mod) that we will create more extreme high and low armor conditions. In the appendix, examples of high armor on the aforementioned male avatar as well as low armor on the female avatar.

Semi-Structured Interview

A semi-structured interview was chosen as the best way to assess post intervention satisfaction and perceived success. Numerous HCI studies have found this methodology to be successful in communicating with participants . Structured interview would not have allowed for researchers to gain a rich body of data from the participants with regard to their feelings toward the game and avatar. Instead, their responses would have been limited to the already created set of questions. On the other hand, a completely unstructured interview could potentially lead to more noisy data as well as inconsistencies in data between respective participants (i.e. participant 1 may not mention performance in detail and participants 2 and 3 may). Considering the possible wide range of gaming experience in the participants, a semi-structured interview allowed for the researchers to have a consistent set of questions for all of the participants, but also allowed for more in-depth and individualized answers.

Interview Questions

• Did you enjoy yourself?
• How do you feel that the boss battle went?
• Are you satisfied with your performance?
• Did you feel like there were any aspects of the game that were affecting your performance negatively?

    a.       What?
    b.       Why?

• Did you feel like there were any aspects of the game affected your performance positively?

    a.       What?
    b.       Why?

• Did you feel a personal connection to your avatar?

    a.       Why?

• What would you have changed about the character if you had made it yourself?

Interview Code Book

[1] Negative armor

           	Comments having to do with negative feelings about armor as protection

. [2] Positive armor

           	Comments having to do with positive feelings about armor as protection.

[3] Negative gender

           	Comments having to do with negative or sexist interpretations about the gender of the avatar.

[4] Positive gender

           	Comments having to do with positive or empowering feelings about the gender of the avatar.

[5] Sexualization

           	Comments referring to the sexual appeal of the avatar.

[6] Boss design

           	Comments having to do with the programming, specifically in reference to the boss. (i.e. movement or attacks)

[7] Game design

           	Comments having to do with all other aspects of UI (i.e. avatar movement, attacks, color balance, reaction times)

[8] Other

           	Any other comment not otherwise classified in the codebook.

Demographics

Are you:

· Male

· Female

What is your current age? (in years)

How would you classify yourself? · African American

· Asian/Pacific Islander

· Latino/Hispanic

· Multi-Racial

· Native American

· White

· Other(Specify)

What was your GPA (on a four point scale – i.e., 0.0 – 4.0) in the previous semester?

What is the highest level of education your mother (or stepmother) finished?

· Some high school

· High school

· Some college

· College

· Graduate or professional school

· Don't know

What is the highest level of education your father (or stepfather) finished?

· Some high school

· High school

· Some college

· College

· Graduate or professional school

· Don't know

What is your parent's current combined annual income?

· $19,999 or less

· $20,000-$39,999

· $40,000-$59,999

· $60,000-$99,999

· $100,000-149,999

· $150,000-$199,999

· $200,000 or more

· Don’t know

Is English your native language?

· Yes

· No

Video Game Use Questionnaire

On a typical school day (Monday through Friday), for how many hours do you play video games during each of the following times?

6 am - Noon _____ hours/day

Noon - 6 pm _____ hours/day

6 pm - Midnight _____ hours/day

Midnight - 6 am _____ hours/day

On a typical weekend day (Saturday or Sunday), for how many hours do you play video games during each of the following times?

6 am - Noon _____ hours/day

Noon - 6 pm _____ hours/day

6 pm - Midnight _____ hours/day

Midnight - 6 am _____ hours/day

Instructions: Please think of the five video games that you have played for the greatest amount of time from when you were in 7th grade until the present. Include computer, mobile, console/TV, and arcade games. Please write down the titles of these games on the blank lines below.

If you have never played a video game in your life, please check here and go on to the next questionnaire

What is the title of your most played game?

In recent months, how often have you played this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 11th & 12th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 9th & 10th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 7th & 8th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

How creative is this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

How creative do you feel when you play this this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

Which of the following categories best describes this game? Check all that apply. · Logic/puzzle

· Driving

· Sport

· Role-playing

· First person shooter

· Third person shooter

· Real-time strategy

· Turn-based strategy

· Other

What is the title of your 2nd most played game? In recent months, how often have you played this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 11th & 12th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 9th & 10th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 7th & 8th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

How creative is this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

How creative do you feel when you play this this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

Which of the following categories best describes this game? Check all that apply.

· Logic/puzzle

· Driving

· Sport

· Role-playing

· First person shooter

· Third person shooter

· Real-time strategy

· Turn-based strategy

· Other

What is the title of your 3rd most played game? In recent months, how often have you played this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 11th & 12th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 9th & 10th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 7th & 8th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

How creative is this game? Rarely Occasionally Often 1 2 3 4 5 6 7

How creative do you feel when you play this this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

Which of the following categories best describes this game? Check all that apply.

· Logic/puzzle

· Driving

· Sport

· Role-playing

· First person shooter

· Third person shooter

· Real-time strategy

· Turn-based strategy

· Other

What is the title of your 4th most played game? In recent months, how often have you played this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 11th & 12th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 9th & 10th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 7th & 8th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

How creative is this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

How creative do you feel when you play this this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

Which of the following categories best describes this game? Check all that apply.

· Logic/puzzle

· Driving

· Sport

· Role-playing

· First person shooter

· Third person shooter

· Real-time strategy

· Turn-based strategy

· Other

What is the title of your 5th most played game? In recent months, how often have you played this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 11th & 12th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

During 9th & 10th grades, how often did you play this game? Rarely Occasionally Often 1 2 3 4 5 6 7

During 7th & 8th grades, how often did you play this game?

Rarely Occasionally Often

1 2 3 4 5 6 7

How creative is this game? Rarely Occasionally Often 1 2 3 4 5 6 7

How creative do you feel when you play this this game? Rarely Occasionally Often 1 2 3 4 5 6 7

Which of the following categories best describes this game? Check all that apply.

· Logic/puzzle

· Driving

· Sport

· Role-playing

· First person shooter

· Third person shooter

· Real-time strategy

· Turn-based strategy

· Other

Player Experience Needs Satisfaction (PENS) Questionnaire

1- Strongly Disagree

2

3

4- Neither Disagree or Agree

5

6

7- Strongly Agree

I feel competent at the game.

I feel very capable and effective when playing.

My ability to play the game is well matched with the game's challenges.

The game provides me with interesting options and choices.

The game lets you do interesting things.

I experienced a lot of freedom in the game .

I find the relationships I form in this game fulfilling.

I find the relationships I form in this game important.

I don’t feel close to other players.

I care about the game characters and enjoy being involved with them.

I care about other players in this game and enjoy playing with them.

When playing the game, I feel transported to another time and place.

Exploring the game world feels like taking an actual trip to a new place.

When moving through the game world I feel as if I am actually there.

I am not impacted emotionally by events in the game.

The game was emotionally engaging.

I experience feelings as deeply in the game as I have in real life.

When playing the game I feel as if I was part of the story.

When I accomplished something in the game I experienced genuine pride.

I had reactions to events and characters in the game as if they were real.

Learning the game controls was easy.

The game controls are intuitive.

When I wanted to do something in the game, it was easy to remember the corresponding control.

Analysis

Since this is just a proposal, we don't have actual results to analyze. Below is an example of the kind of results we could expect and the corresponding analysis of those results.

For the analysis of each subject, we focused on a few variables. The most important of which is eye tracking data, which will be separated into groups by gender, then further broken down by the gender of the avatar, and then again by type of armor. We also collected data to measure performance, which will then be compared against the subjects’ answers of the questionnaire. Performance is measured based on damage taken and damage dealt.

Results of Eye Tracking

Analysis of Eye Tracking data:

As we can see, female avatars with revealing armor, cause more time to be spent looking at the female avatar by both genders, but significantly more by male players. Possibly more surprising, male avatars in revealing armor also garner more looks, and also by both genders. Most likely due simply to players getting acquainted with the avatar they will be playing, all subjects spent some amount of time looking at their avatar at the beginning of play, but in the cases of the revealing armor, more time was spent looking at the avatar at the beginning of and during gameplay.

Results from gameplay:

Analysis of gameplay results correlated with Eye Tracking data:

Players who reported and displayed competence in playing games spent less time looking at their avatars. Those “gamers” displayed higher skill during gameplay by taking less damage from and dealing more damage to enemies. Likely due to having seen avatars like these before, they spent less time looking at their avatar overall, these people also spent less time looking at both their health bar and any other part of the Heads Up Display (HUD). People who fall somewhere in the middle of gaming experience spent more time than the experienced “gamers” looking at the avatars and more time looking at the HUD, presumably before it took more time to understand what everything was. Having less experience would mean that displays like our game’s HUD has are less familiar. These “middle experience” players can be identified by having taken more damage and dealt less damage than the experienced players, but not enough of a difference to be the inexperienced players. The inexperienced players, what some “gamers” might call “noobs,” can be identified by having taken the most damage and dealt the least damage. They, on average, spent more time looking at their avatar, more time trying to figure out what was on the HUD, and also had less focused eye movement. What is meant by this last one is that the more experienced players typically kept their eyes on the task. Either looking at their health bar when they took damage, looking at the enemy, or at anything that would be important at any particular time during gameplay. Inexperienced players logged more time with the eye tracking data that we interpret to mean they were looking at the keyboard.

Analysis of gameplay results and survey

We wanted to see if players with revealing armor would report having higher skill levels than those with full armor even though they performed just as well. What we found was that, on average, people gave themselves higher scores for performance, if their avatar had the revealing armor.