Primary social emotions toward a non-player character: an examination of moral decision-making in video games
Article information
Abstract
This exploratory study examines the emotional responses toward non-player characters (NPCs) when challenged with an in-game moral decision (i.e. justified versus unjustified killing) and when exposed to game statistics on whether their moral choice aligned with the majority public. Using Buck et al.’s (2014) social and moral emotions (SAME) scale, participants (n = 301) completed an online experiment measuring their primary social emotions after being presented with a written scenario and screenshots from a popular video game that required participants to make a moral decision. Findings provide overall support for the interdependent connections in primary social emotions and revealed that embarrassment and shame were higher towards the NPC in the unjustified choice. Study results provide practical health implications for video game developers to develop and design more elaborate narrative paths that delve into the ethical tendencies of video game players.
Introduction
Video games are one of the most widely popular and advanced forms of interactive entertainment media (Vorderer, 2000). According to the Entertainment Software Association (2020), there are approximately 214.4 million U.S. Americans who play video games. Video games are a unique form of interactive entertainment that can be conceptualized by Jansz (2005) with four elements: interactivity between video game player and media, rule-based nature in gameplay, variable and quantifiable outcome, and required video game player effort (i.e. player must apply some effort in completing the game). Past research has explored a variety of effects from video games such as real-life aggression (e.g., Anderson & Bushman, 2001; Barrington & Ferguson, 2022; Weber et al., 2006), emotions (e.g, Yao et al., 2019), and prosocial effects (e.g., Grizzard et al., 2014; Shoshani et al., 2021). Current research has explored moral decision-making during gameplay (e.g., Ferchaud & Beth Oliver, 2019; Funk et al., 2003; Hartmann et al., 2014; Holl et al., 2022). Additionally, prior studies have found that video games can significantly impact one’s health (Hasan, 2017; Roy & Ferguson, 2016). For example, research has shown that violent video games can negatively impact one’s health by increasing stress levels (Hasan, 2017), decreasing cardiac coherence (Hasan et al., 2013), and increasing stress-related hormones in the blood (Lynch, 1999). Additionally, chronic stress can lead to coronary heart disease or hypertension (Steptoe & Kivimaki, 2012). Thus, it is important to examine how video games may evoke stress through players emotions.
Recently, in-game tasks that explore moral judgments have emerged as a popular and common plot device in video games (Weaver & Lewis, 2012). Games such as Mass Effect (Bioware Corporation, 2007) and Fallout 3 (Bethesda, 2008) require players to choose between actions that can impact non-player characters (NPCs) and affect their in-game skills (Schulzke, 2009). For example, players are required to make a moral choice of deciding which NPC to sacrifice in battle. Adventure-style role-playing games (RPG) such as The Walking Dead (Telltale Games, 2012) explore moral issues (e.g. which NPC to save) by requiring the player to make an on-the-spot decision between two to four choices that will impact the storyline of the game. Unique to The Walking Dead (Telltale Game, 2012), the game introduces an element to moral decision-making by tracking all player decisions and revealing the decisions that other players have made at the end of the game. The tracking of these in-game decisions not only allows game developers to tailor future game episodes (Klepek, 2012) but also provides the players with an opportunity to actively think about and reflect upon their moral choices (Saucier, 2015).
While research has focused on the emotions that emerge during video game play (e.g., Yao et al., 2019), few studies have examined the emotions toward NPCs reported by individuals who had to make a moral in-game decision. By utilizing the in-game moral choices from The Walking Dead (Telltale Games, 2012-2019) and the primary social emotions theoretical framework, the primary purpose of the present study is to examine how moral decision-making will influence an individual’s emotions toward NPCs. Specifically, how will the emotions of guilt and shame be impacted by an in-game moral decision? Furthermore, there is a gap in the literature about moral decision-making in video gameplay when statistical comparisons of other video game players’ moral decisions are revealed to the individual. Therefore, the second goal of the present study is to understand how players are impacted by the moral decisions presented in video games when compared to other players.
Literature Review
Primary Social Emotions
Studies have demonstrated that exposure to playing video games can lead to emotional reactions such as guilt (Carnagey et al., 2007; Grizzard et al., 2014; Hartmann & Vorderer, 2010; Lin, 2010), shame (Hartmann & Vorderer, 2010; Shafer, 2012; Lin, 2010; Weaver & Lewis, 2012), or enjoyment (Jansz, 2005). Much of these research studies focus on how players feel after making a moral decision. However, there is a lack of studies examining emotions felt toward NPCs. Buck (2014) suggests that primary social emotions (i.e. pride, arrogance, guilt, shame, envy, jealousy, pity ,and scorn) exist in interpersonal relationships between oneself and in comparison to others. Two aspects of fundamental social motives consist of the need to be loved and the need to follow or exceed the expectations of others (Buck, 2014). Thus, primary social emotions can provide a theoretical framework to explain the emergence of some emotions over others in a given moral situation.
Buck (2014) explains that each primary social emotions consist of the following elements: each primary social emotion has a twin emotion associated as a result of the need to meet or exceed expectations and the need to be loved (i.e. guilt is the twin of shame) and each primary social emotion has an opposite emotion (i.e. pride is the opposite of guilt). The discussion of primary social emotions suggests that hypotheses can be derived from positive and negative correlations among the eight emotions (Buck, 2014). For example, a positive correlation exists between primary social emotions with both its twin and its reciprocal emotion (Buck, 2014). As it pertains to video game scenarios, the social and primary emotions scale could provide a unique framework to explore the emotions felt toward NPCs. Therefore, the following hypotheses are posited based on the interdependent connections of primary social emotions:
H1a: Feelings of guilt will be positively correlated with feelings of shame.
H1b: Feelings of envy will be positively correlated with feelings of jealousy.
In contrast, a negative correlation exists between primary social emotions with its opposite emotion (Buck, 2014). Thus, the postulates attached to primary social emotions can give insight into specific emotions towards the NPC that emerge from making moral choices. The following hypotheses are posited based on the primary social emotions having a negative correlation:
H1c: Feelings of guilt will be negatively correlated with feelings of pride.
H1d: Feelings of envy will be negatively correlated with feelings of arrogance.
Video Games and Moral Choices
The nature of video games often subjects players to perform acts of violence such as killing or destroying people, creatures, and buildings (Klimmt et al., 2006). Bandura (2002) suggests that acts of violence in the real world would provoke an adverse reaction due to one’s moral standards known as moral disengagement theory. However, acts of violence in video games are often met with enjoyment rather than disdain (Klimmt et al., 2006). For example, Klimmt et al (2006) conducted a qualitative study examining the moral concerns during video game violence. Results from ten interviews revealed that players were able to enjoy video game violence because their focus was on winning the game and attributed video game violence as a consequence of good performance. The authors suggest that enjoyment from video game violence is a result of “moral management strategies.” Thus, moral disengagement is evident in video games where moral conflict arises and can impact the emotions felt by players. One proposed explanation for the enjoyment of violent acts is the concept of moral management. According to moral disengagement theory, video game players utilize strategies that can minimize or suspend moral standards in order to experience enjoyment (Bandura, 2002; Bandura et al., 1996).
Bandura et al. (1996) identified the following eight different moral disengagement cues that can influence a video game player’s perception of the violent act: moral justification, euphemistic labeling, advantageous comparison, displacement of responsibility, diffusing responsibility, disregard or distortion of consequences, and attribution of blame. Moral justification suggests that a moral purpose can validate committing a violent act (e.g., murdering for the greater good of society). Euphemistic labeling refers to strategic labeling of a heinous act to appear less harsh (e.g., “friendly fire” used to describe the accidental killing of soldiers). Advantageous comparison implies comparing the severity of an act to the opponents (e.g., “sacrificing a few to save thousands”). Displacement of responsibility proposes that personal responsibility is placed on another individual. Disregard or distortion of consequences happens when damaging consequences are ignored or reduced (e.g., lack of blood following a stabbing). Dehumanization refers to the portrayal of victims as lacking human qualities. Attribution of blame suggests blaming the adversaries for an attack.
A content analysis of 17 of the top-rated first-person shooter games and a systematic literature review of moral disengagement theory by Hartmann et al. (2014) found support that moral disengagement cues were embedded in violent video game genres. The study also found that many plots of first-person shooters involve a black-and-white conflict where it is evident who the antagonists are (Hartmann et al., 2014). Thus, moral disengagement can be conceptualized as a strategy for video game enjoyment in which the violent acts are seen as less morally problematic due to concepts of justification and displacement (Klimmt et al., 2006).
Shafer (2012) examined moral disengagement in moral decision-making in video games with an experiment in which 83 participants were randomly assigned to one of three game situations with moral choices. Results demonstrated that moral disengagement and moral activation were directly related to moral choices in video games. The author suggests that players who were morally activated were significantly more likely to choose the less morally-charged decision whereas players who were morally disengaged were significantly more likely to choose the evil option (Shafer, 2012). In addition, Krcmar and Cingel (2016) found that players have almost an equal amount of both strategic and moral reasoning when playing games. These findings suggest that players in the virtual world of gameplay utilized a lot of the same reasoning processes as in the real world. In summary, previous studies have demonstrated that moral choices in video games are influenced by one’s moral engagement or disengagement of moral decision-making.
Guilt and Shame in Video Games
Moral disengagement theory has also been utilized in video game research examining guilt and shame. Lin (2010) found that players reported different emotions toward human opponents and monster opponents based on techniques to reduce one’s morality in decision-making. Likewise, a study by Hartmann and Vorderer (2010) examined the moral emotions experienced in virtual violence found in first-person shooter video games. Results revealed that players felt less guilt and negative emotions if the virtual violence was justified (i.e. fighting evil or saving the world). Thus, studies have demonstrated that moral decision-making in video games can be tied to feelings of guilt.
Recent video game research has explored the emotions of guilt and shame that is felt by the player when faced with a moral decision in gameplay (Bushman & Huesmann, 2001; Hartmann et al., 2010; Lin, 2010; Mahood & Hanus, 2015). The role of perceived guilt has been also studied in the context desensitization in game-play (Carnagey et al., 2007). Desensitization can be defined as the elimination of cognitive, emotional, and behavior responses to the stimulus (Rule & Ferguson, 1986) and one implication of heavy video game violence is desensitization (Bushman & Huesmann, 2001). Other scholars have associated desensitization to characteristics such as having less sympathy for a victim of violence, reduction in perceived guilt of a violent perpetrator, and increase in aggressive behavior (Carnagey et al., 2007). For example, someone who is desensitized to violence from video games may be less inclined to help someone in danger and may view violence as normalized in society, thus having an overall reduction in guilt.
In the context of moral decision-making, a study by Lin (2010) found that participants reported more feelings of guilt and shame when participants shot a monster video game opponent as opposed to a human video game opponent. Findings suggest that participants reported fewer feelings of shame and guilt when killing a monster opponent because dehumanization techniques were used in gameplay leading to fewer moral judgments. In addition, a study by Mahood and Hanus (2015) explored feelings of guilt and shame through role-playing games (RPG). RPG proposes a unique video game platform because it focuses more on the narrative structure of the game such as building a video game character’s persona (Schulzke, 2009). Thus, The Walking Dead (Telltale, 2012), an episodic interactive RPG that focuses on character and story development where on-the-spot moral decisions will influence the storyline of the game, provides an interesting platform to explore emotions in gameplay.
Therefore, based on the summary of research findings demonstrating a relationship between feelings of guilt and shame when playing a video game that challenges players to make a reprehensible action (Lin, 2010; Mahood & Hanus, 2015), it is reasonable to assume that when challenged between killing one of two NPCs, guilt and shame will be higher towards the one who is not saved as opposed to the one who is saved. Therefore, the following hypotheses are proposed:
H2a: Feelings of guilt will be higher towards the NPC who is not saved than the one saved.
H2b: Feelings of shame will be higher towards the NPC who is not saved than the one saved.
As previously mentioned, research revealed that guilt and shame are influenced by the justification of the moral decision (Hartmann et al., 2010; Hartmann & Vorderer, 2010; Mahood & Hanus, 2010). In an experimental study, Hartmann et al. (2010) found that participants reported feeling more guilt when the video game violence was considered “unjustified” (i.e., supporting the torture camp) versus “justified” (i.e., freeing innocent prisoners from a torture camp). An experimental study where 75 participants were recorded playing a portion of the video game Fallout 3 conducted by Weaver and Lewis (2012) found that the moral decisions of players mirrored the moral decisions made in real-life situations and acting in anti-social ways increased guilt in gameplay. The authors suggest that players are able to interact with NPCs as a result of the suspension of disbelief (Weaver & Lewis, 2012). However, Mahood and Hanus (2015) argue that these studies focusing on player guilt were conducted using the video game genre of first-person shooter (FPS) games in which the primary motive is to shoot and destroy the enemy and the narrative of the video game story remains secondary.
Exploring video games that focus more on narratives such as role-playing games, Mahood and Hanus (2015) found that players reported more feelings of guilt in decision-making when exposed to a narrative about an NPC who has conducted immoral acts (i.e., humiliating a homeless man, killing an NPC’s mother) versus an NPC who has conducted moral acts (i.e., showing kindness towards a homeless man, saving an NPC's mother). Therefore, the element of justification found in gameplay can impact feelings of guilt and shame among participants. Based on these past findings, it is predicted that the justification of the decision will impact the emotions of guilt and shame. The following hypotheses are proposed:
H3a: For the justified decision, feelings of guilt will be higher towards the one not saved.
H3b: For the justified decision, feelings of shame will be higher towards the one not saved.
Video Games and Emotions
Research examining the emotional implications of moral decision-making in video games have been mixed (e.g., Funk et al., 2003; Hartmann & Vorderer, 2010; Weaver & Lewis, 2012). For example, a study by Funk et al. (2003) examining exposure to violent video games found a reduction in empathy and attitudes toward violence among participants. After exposure to either a violent or nonviolent game, 66 children completed questionnaires and findings demonstrated that long-term exposure to violent video games lowered preexisting empathy as opposed to increasing empathy. Researchers suggest that lowered empathy is a result of desensitization from exposure to video game violence (Funk et al., 2003).
In contrast, a study by Jansz (2005) examined the appeal of violent video games and found that emotional experiences felt during gameplay attributed to positive emotions such as enjoyment rather than negative feelings such as frustration. The researcher proposes that enjoyment from violent video games is a culmination of experiencing both positive (i.e. joy, accomplishment) and negative emotions (i.e. fear, anger, empathy).
Research has also found prosocial effects from exposure to video games (Grizzard et al., 2014). For example, Grizzard et al. (2014) found that players who chose to play an immoral character elicited more feelings of guilt, but also lead to an increase in prosocial benefits like sensitivity. Therefore, video games may impact players to experience either positive emotions (i.e. enjoyment, triumph) or negative emotions (i.e. guilt, shame). Thus, the following research questions are proposed to explore the various primary social emotions that can emerge:
RQ1: How will primary social emotions differ toward the one saved and the one not saved?
RQ2: How will primary social emotions differ between the justified decision and the non-justified decision?
Unique to The Walking Dead, players are exposed to the percentage of other players who agreed or disagreed with their moral decision. To date, no research has examined how a player’s emotion will change based on exposure to knowledge of other players’ moral decisions. Therefore, the following research question is proposed to explore how social primary emotions may change:
RQ3: How will primary social emotions change based on statistical knowledge about agreement with one’s moral decision?
Method
Participants
Undergraduate students (N = 302) from a large northeastern university introductory communication course and an introductory nonverbal class participated in this study. Participants received one-course extra credit for their participation. A total of 162 female and 140 male students participated in the online experiment. Ethnicities are reported as follows: American Indian or Alaska Native (n = 4), African American (n = 16), Asian (n = 31), Native Hawaiian or Other Pacific Islander (n = 2), Caucasian (n = 224), Hispanic/Latino (n = 30), other (n = 6).
The online experiment was hosted on Qualtrics and presented participants with a gameplay scenario from The Walking Dead in which a moral decision is required. Specific emotions-related questions pertaining to their decision were asked immediately after the decisions made in gameplay. Participants were also exposed to a percentage of players who agreed with the participant’s decision made during gameplay. Two follow-up questions were then asked after the exposure of the percentage.
Materials
In this experiment, participants were exposed to a gameplay scenario from The Walking Dead, an episodic point-and-click adventure role-playing video game. The interactivity element in video games provides players an opportunity to assume a role of a character in a fictional setting and make decisions that impact the storyline (Schulzke, 2009). The current study used a scenario from The Walking Dead (Telltale Games, 2012). The Wall Street Journal (2013) reported that The Walking Dead sold approximately $40 million in revenue. Thus, The Walking Dead (Telltale Games, 2012) is a widely-popular game in society and is reflective of moral decision-making in gameplay.
Procedure
Participants were given a detailed text gameplay scenario that is comparable to the actual game. The following scenario was presented to the participant: “Imagine you are a survivor in a zombie apocalypse in Georgia. Finding shelter at a farm with other survivors, you come across a situation in gameplay where you are asked to make a tough decision on which video game character to save: Duck – a young boy who is loving – or Shawn – a kind and resourceful young man – are both attacked when trying to mend a fence.” Next, participants were asked to choose to save either a child’s life or a young adult male’s life. Participants were shown two screenshots side-by-side, each depicting the child and the young adult male in danger of being consumed by zombies. The participant is presented with the following question: “Which character do you choose to save? The character not selected will be killed by zombies.”
To further elicit negative emotions from the participant, participants were presented with a fictitious percentage of other players who did not agree with their choice. After answering a series of questions regarding their emotions toward the one saved and the one not saved, participants are shown the following text: “75% of players chose to save (the individual not chosen).” A high percentage of player disagreement was chosen in order to examine how participants report emotions when their moral choice is not among the majority. Afterwards, participants were asked the same series of questions regarding their emotions toward the one saved versus the one unsaved.
Measures
Social and Moral Emotions Scale
Buck et al. (2008) social and moral emotions (SAME) scale was used to assess participants’ emotions. After given one gameplay scenario, participants were asked the following question: “When thinking about the person I saved, I feel X toward myself.” Participants were asked to describe their emotions using a 7-point Likert-type scale, ranging from “not at all” to “very much” with various emotions such as “embarrassment”, “pride”, “guilt”, “shame”, “arrogance”, “triumph”, and “humiliation.” Participants were then asked the following item: “I would feel X toward the person I saved.” Participants were asked to describe their emotions using a 7-point Likert-type scale, ranging from “not at all” to “very much” with various emotions such as “pity”, “jealousy”, “scorn”, “envy”, “admiration”, “resentment”, “sympathy”, and “contempt.”
Participants were then asked the same questions regarding the person who was not saved. “When thinking about the person I did not save, I feel X toward myself.” Participants were asked to describe their emotions using a 7-point Likert-type scale, ranging from “not at all” to “very much” with various emotions such as “embarrassment”, “pride”, “guilt”, “shame”, “arrogance”, “triumph”, and “humiliation.” Participants were then asked the following item: “I would feel X toward the person I did not save.” Participants were asked to describe their emotions using a 7-point Likert-type scale, ranging from “not at all” to “very much” with various emotions such as “pity”, “jealousy”, “scorn”, “envy”, “admiration”, “resentment”, “sympathy”, and “contempt.” Participants were then asked the same questions regarding the person who was not saved.
Moral Dilemma
A moral dilemma can be conceptualized as the acceptability, or justification, one feels when asked to kill one person over another person in order to save others (Foot, 1967). For example, the “trolley problem” asks individuals if they would kill one person in order to save five people. Research has found that 90% of individuals would not choose to kill one person to save five (Hauser et al., 2007). Kawai et al. (2014) examined the trolley problem by manipulating the age of the sacrificial person (i.e., a 70-year old male stranger, a 20-year-old male stranger, a 5-year-old male stranger, or a 20-year-old male disabled stranger) and found that killing the child was the least acceptable option. Findings suggest that individuals may feel negative emotions toward sacrificing someone who has a shortened lifespan compared to others who have lived longer (Kawai et al., 2014). Thus, based on past research on moral dilemmas and several game reviews of The Walking Dead, saving the young boy is operationalized as the justified decision and saving the young adult male is operationalized as the unjustified decision.
Demographics
Participants were asked to report their sex, age, ethnicity, and the amount of gameplay per each video game genre.
Results
Primary Social Emotions
A Pearson correlation coefficient was calculated to test H1a-d, which posited a positive correlation between guilt and shame (see Table 1). A strong positive statistically significant correlation was found between guilt and shame, r(299) = .72, p < .001. This finding implies that participants who reported high levels of guilt also reported high levels of shame. Therefore, H1a is supported. H1b posited a positive correlation between envy and jealousy. A strong positive statistically significant correlation was found between envy and jealousy, r(298) = .77, p < .001. This finding implies that participants who reported high levels of envy also reported high levels of jealousy. Therefore, this hypothesis is supported. H1c posited a negative correlation between pride and guilt. A statistically significant weak negative correlation was found between pride and guilt, r(302) = -.16, p < .001. This finding implies that participants who reported high levels of envy also reported lower levels of jealousy. Therefore, H1c is supported. H1d posited a negative correlation between shame and arrogance. A statistically significant weak positive correlation was found between shame and arrogance, r(299) = .45, p < .001. This finding implies that participants who reported high levels of envy also reported high levels of jealousy. Thus, H1d is not supported.
Correlations
Primary Social Emotions toward Saved and Not Saved
H2a and H2b proposed that guilt and shame will be higher toward the one who was not saved. To examine H2a, H2b, and RQ1 – which questioned the differences between the primary social emotions (embarrassment, pride, guilt, shame, arrogance, and triumph) toward the one saved and the one not saved – six paired-sample t-tests were calculated (see Table 2). Four emotions were found to be statistically different towards the NPC who was saved and the NPC who was not saved. Embarrassment was reported to be statistically higher towards the one not saved, t(301) = -15.63, p < .001. Pride was statistically lower towards the one not saved, (t(303) = 18.37, p < .001. Shame was reported to be statistically higher towards the one not saved, (t (301) = -17.23, p < .05). Triumph was statistically lower towards the one not saved, (t (302) = 16.46, p < .001). Guilt and arrogance were not statistically different between the one saved and the one not saved. Thus, H2b is supported.
Paired T-Tests towards the One Saved and the One Not Saved
Primary Social Emotions towards the Justified versus Unjustified Decision
To examine H3a, H3b, and RQ2, independent-samples t-tests were conducted to compare the primary social emotions (embarrassment, pride, guilt, shame, arrogance, and triumph) and the individual saved (see table 3). There were significant differences in pride scores for the justified decision (i.e., saving the young boy), (M = 4.85, SD = 1.97) and unjustified decision (i.e., saving the young adult male, (M = 4.24, SD = 1.81), t(301) = 2.79, p = .006. Guilt scores were higher for the unjustified decision, t(302) = -2.395, p = .017. Triumph scores were higher for the justified decision, t(302) = 2.58, p = .010. Shame scores were marginally significant (p = .05) with scores higher for the unjustified decision (M = 3.12, SD = 1.85, n = 135) than the justified decision (M = 2.24, SD = 1.54, n = 165). Thus, H3b is supported.
Independent T-tests towards the Justified and Unjustified Decision
Primary Social Emotions among the Justified Decision
To examine primary social emotions for those who chose the young boy, the more justified choice, six paired t-tests were conducted to examine how individuals feel towards the saved and not saved (see table 4). Embarrassment, pride, shame, and triumph were statistically different between the one saved versus the one not saved.
Paired T-Tests towards the One Saved and Not Saved for the Justified Decision
Primary Social Emotions among the Unjustified Decision
To examine primary social emotions for those who chose the young man, the less justified choice because he is not a kid, six paired t-tests were conducted to examine how individuals feel towards the saved and not saved (see table 5). Embarrassment, pride, guilt, shame, and triumph were statistically different between the one saved versus the one not saved.
Paired T-Tests towards the One Saved and Not Saved for the Unjustified Decision
Primary Social Emotions among the Justified Decision after Exposure to Percentage
To examine the influence of exposure to the percentage of people in disagreement with the participant’s decision (i.e., saving a young boy or a young man) on primary social emotions (embarrassment, pride, guilt, shame, arrogance, triumph, and humiliation), six paired-sample t-test were calculated (see table 6). Four paired-sample t-tests regarding the one that was saved were found to be significant. Embarrassment, pride, shame, and triumph demonstrated a significant change, (t(300) = 16.51, p < .001. Guilt and arrogance did not significantly change.
Paired T-Tests for Primary Social Emotions among Justified Decision
Primary Social Emotions among the Unjustified Decision after Exposure to Percentage
To examine the influence of the percentage of people in disagreement with the participant’s decision (i.e., saving a young boy or a young man) on primary social emotions (embarrassment, pride, guilt, shame, arrogance, triumph, and humiliation) among those who chose the unjustified decision, six paired-sample t-test were calculated (see table 7). Four paired-sample t-tests regarding the one who was saved were found to be significant. Embarrassment, pride, shame (M = 2.66, SD = 1.65), and triumph all demonstrated significant changes. Guilt and arrogance did not significantly change.
Paired T-Tests for Primary Social Emotions toward the Unjustified Decision
Discussion
This exploratory study aims to 1) extend support for the interdependent relationships among primary social emotions and 2) understand the relationship between primary social emotions toward non-player characters (NPC) when forced to make in-game moral decisions. Buck (2014) posits that primary social emotions are interdependently connected with each having an opposite emotion (i.e., guilt is the opposite of pride) and a twin emotion (i.e., guilt is the twin of shame). Specifically focusing on the primary social emotions of guilt and shame, two emotions often explored in video game research focusing on moral decision-making (Grizzard et al., 2014; Lin 2010), this study found a strong positive correlation between guilt and shame and a mild negative correlation between guilt and pride, thus providing support for Buck’s (2013) primary social emotions theoretical framework. For instance, Buck (2013) suggests that someone who reports feeling pride is less likely to feel guilt or shame, and one who feels guilt is also likely to feel shame. Findings provide implications for emotional communication research by examining how individual emotions arise and are interdependently connected through social situations. Understanding the relationships between various emotions such as guilt and shame has practical implications for prosocial, relationship-enhancing effects such as maintaining healthy interpersonal relationships (Leith & Baumeister, 1998). Additionally, Kral et al. (2018) found that emotions evoked through video games can potentially enhance empathy among young children.
Utilizing the social and moral emotion (SAME) scale (Buck at al., 2008), the present study examined primary social emotions through the social situation of how participants feel towards NPCs. Specifically, the study examined how primary social emotions would differ toward the NPC who was saved versus the NPC who was not saved. For all participants, regardless of whom they chose to save, embarrassment and shame were statistically higher towards the NPC not saved. Findings align with existing video game research, which found the emotional impact that gameplay can have on players (Jansz, 2005; Lin, 2010). Health research has reported that negative emotions can have negative health implications (Brown, 2006; Dolezal & Lyons, 2017; Herter et al., 2021). For example, individuals who feel shame may feel personally flawed, which can threaten their identity and cause significant mental distress (Brown, 2006). Additionally, research has found that individual emotions can impact healthy behaviors (Agrawal et al., 2007; Herter et al., 2021). For example, Herter et al. (2021) noted that specific emotions (such as sadness) can positively impact intentions to become healthier.
When looking between participants who chose the justified decision (e.g., killing a young man) versus the unjustified moral decision (e.g., killing a young child), guilt was statistically greater for those who reported choosing the unjustified moral decision. This aligns with past findings from Allen and Anderson (2021), which found that participants felt more guilt after unjustified violence (i.e., killing of innocent humans) versus justified violence (i.e., killing creatures) and provide support for moral disengagement theory (MDT), which posits that individuals utilize various moral disengagement strategies to minimize guilt during gameplay.
Additionally, when comparing the two groups based on which NPC was saved, participants who chose the justified choice (i.e., saving the young boy) reported higher levels of pride and triumph, which are positive primary social emotions. Unsurprisingly, those who chose the unjustified choice (i.e., saving the young man) reported statistically higher guilt and shame. This supports past findings that feelings of guilt will be higher towards unjust actions in video games such as unnecessary violence and murder (Lin, 2010; Mahood & Hanus, 2015). Additionally, results contribute to moral dilemma research on the “trolley problem” by adding the component of in-game moral decision-making. Furthermore, findings align with past research, which found killing a child to be the least acceptable moral choice regardless if it saved others (Kawai et al., 2014)
As in-game moral dilemmas continue to become implemented into games, game developers have been keeping statistics on player in-game moral decisions and reported that approximately 60% of players are likely to choose the “good”, or justified, option (Shanley, 2019). A key contribution of this study is looking at how primary social emotions are impacted once participants are exposed to game statistics on whether their moral choice is aligned with the majority public. For participants who selected the morally justified choice (i.e., saving the young boy), embarrassment and shame increased after the knowledge that their choice was in the minority. Not surprisingly, triumph and pride, which are positive primary social emotions, decreased after exposure.
Surprisingly, for participants who selected the unjustified moral choice (i.e., saving the young man), embarrassment, guilt, and shame decreased when exposed to game statistics revealing that their moral choice did not align with the majority of players. A potential explanation could be that participants who selected the unjustified moral choice are desensitized to their moral choices and participants could potentially feel better about not saving the character if other players are saving the character. These results provide practical implications for video game developers to develop and design more elaborate narrative paths to delve into the ethical tendencies of video game players.
Limitations and Future Research
There are several limitations to this current study that may have impacted the results. Screenshots of the decision-making scenarios were used in lieu of actual gameplay, which may have taken away from the authenticity of the decisions. In actual gameplay, players are given five seconds to make a decision with fast-paced music cueing in at just the right moment in an attempt to depict the urgency of the situation. In the online experiment, participants had unlimited time to make their decision.
Another crucial difference between screenshots and the actual video game is the aftermath of the player’s decision. Previous research has explored the emotional implications of a lack of consequences of virtual violence (Hartmann & Vorderer, 2010). For example, the absence of consequences of virtual violence (i.e. bloodshed, screams, dying sequences) could result in the player feeling less guilt. Therefore, in real-time gameplay, once players decide which video game character to save, the video game shows the one who was not saved to be eaten by zombies. This aftermath scenario has implications for the social emotions felt by the player regarding their decisions.
Although this study is confined to screenshots of the game, real-time decision-making could impact the implications of primary social emotions. Future examinations in real-time decision-making could be examined. For example, how would social emotions change once the aftermath of the decision is shown to the participant? Would seeing the negative ramifications of a participant’s decision change their emotions regarding themselves and the video game characters? Follow-up studies could also record participant gameplay so a content analysis could be conducted in order to capture the nonverbal emotions from moral decision-making.