Excerpt
Table of Contents
List of Figures
List of Tables
List of Abbreviations
Chapter 1: Introduction
Chapter 2: Literature Review
2.1 Literature Search
2.2 Self-Awareness
2.2.1 Objective Self-Awareness Theory
2.2.2 Self-Awareness and Self-Consciousness
2.2.3 Cognitive and Behavioral Reactions to Self-Awareness
2.2.4 Influencing Factors on Self-Awareness
2.3 Dishonesty
2.3.1 Motives for Dishonesty
2.3.2 Major Mechanisms of Dishonesty
2.3.3 Alternative Mechanisms of Dishonesty
2.3.4 Influencing Factors on Dishonesty
2.4 Virtual Reality
2.4.1 Technical Terminology
2.4.2 Experimental Advantages of VR
2.4.3 Psychological Effects in VR
2.5 Hypotheses
Chapter 3: Methodology
3.1 Objective
3.2 Experimental Design
3.2.1 Participants
3.2.2 Procedure
3.2.3 Treatments
3.2.4 Variables and Measurement
Chapter 4: Results and Discussion
4.1 Results
4.1.1 Descriptive Statistics
4.1.2 Primary Results
4.1.3 Secondary Results
4.2 Discussion
4.2.1 General Findings
4.2.2 Specific Findings of Self-Awareness
4.2.3 Specific Findings of Dishonesty
4.2.4 Implications for Managers and SMEs
Chapter 5: Limitations and Further Research 38 Chapter 6: Conclusion 40 Appendices 41 References
Abstract
Objective self-awareness theory (Duval & Wicklund, 1972) predicts that individuals behave morally compliant and honest under increased self-awareness. We1 argue that this proposition requires a critical reflection in the context of Virtual Reality (VR). The present thesis investigates self-awareness and dishonesty in VR and is among the first to provide such an exploratory reflection. In an experiment, we tested the effects of the two factors (1) mirror presence and (2) avatar choice on dishonesty in a highly-immersive VR setting. Participants played a mind game in VR that incentivized dishonest reporting without any fear of detection. We let participants choose between two human and one fictitious avatar. We also varied the presence of a mirror across participants. Our results show that participants took significantly more time to lie when facing a mirror or wearing a human avatar, indicating a higher state of self-awareness. However, this effect did not significantly influence the resulting level of dishonesty. We discuss a number of factors that could explain this finding and highlight the relevance of further research in the economics domain.
Keywo rds : Mirror presence, mask, dishonesty, self-image, Virtual Reality The editorial “we” is employed throughout the entire thesis
List of Figures
Figure 1: Screenshot of the avatar selection room
Figure 2: Screenshots of the two treatment rooms
List of Tables
Table 1: Demographic overview of the n = 45 experiment participants
Table 2: Overview of variables measured by the logging algorithm
Table 3: Descriptive statistics of the prediction task
Table 4: Analysis of balances in all treatment - avatar combinations in the experiment in a [3, 3] matrix
Table 5: Multiple linear regression model of different regressors and control variables on the dependent variable balance
Table 6: Analysis of response times in all treatment - avatar combinations in the experiment in a [3, 3] matrix
List of Abbreviations
Abbildung in dieser Leseprobe nicht enthalten
Chapter 1: Introduction
“Being entirely honest with oneself is a good exercise.” – Sigmund Freud
What is the nature of the self ? The search for an answer about the nature of human consciousness and the self historically roots in philosophy (e.g., James, 1890). During the 20th century, the domain of social psychology greatly advanced these concepts and generated more practical insights about interactions between the self (e.g., Erikson, 1968; Fenigstein, Scheier, & Buss, 1975; Freud, 1927; Gergen, 1971) and the environment (e.g., Ornstein, 1972; Ornstein, 1973; Sperry, 1969). One field of research which advanced in this period deals explicitly with the state of self-awareness and its implications on human cognition and behavior (e.g., Carver, 1979; Carver & Scheier, 1981; Duval & Wicklund, 1972; Hull & Levy, 1979). Interestingly, as Wicklund (1979) points out, “the person who becomes self-aware is more likely to act consistently, be faithful to social norms, and give accurate reports about himself” (p. 1). This line of thought has since inspired the potential of using self-awareness to promote ethical behavior (see Postmes & Spears, 1998).
Ethical behavior and moral standards are manifested through societal norms (Rosenbaum, Billinger, & Stieglitz, 2014). These societal norms reflect on the characteristics of the society and are of practical relevance in daily interaction (Abeler, Becker, & Falk, 2014). For this thesis, one moral standard is of particular interest: Honesty. As empirical research shows, dishonesty is clearly regarded as anti -social behavior in society (Goldsen, Rosenberg, Williams, & Suchman, 1960), but broadly prevalent in major parts of society (Mazar & Ariely, 2006). Examples range from insurance fraud (Ernst & Young, 2018) to income misreporting to reduce tax payments (Allingham & Sandmo, 1972; Herman, 2005, May 30), to lighter offenses, like keeping excess change in a restaurant (Azar, Yosef, & Bar-Eli, 2013). All these instances of dishonesty are not only antisocial, but bear huge costs on businesses and economies (Graham, Litan, & Sukhtankar, 2002; Weber, Kurke, & Pentico, 2003). It becomes clear that the field of dishonesty is increasingly relevant in the economics domain. Academic research aims to identify the mechanisms and factors that influence lying behavior (Jacobsen, Fosgaard, & Pascual-Ezama, 2018; Rosenbaum et al., 2014).
Our current understanding has already classified several such factors (Jacobsen et al., 2018; Rosenbaum et al., 2014), which are consistent with the theory of Duval and Wicklund (1972) about self-awareness. However, disruptive technological shifts change the external environment of these factors, and, therefore, call for an update of our theory (Innocenti, 2017). One disruptive technology that has greatly advanced in maturity and adoption in recent years is Virtual Reality (VR) (Gartner, 2017, 2018). VR provides the possibility to experience a seemingly real and physical simulation of a fictitious environment, brought to the individual through a ‘high tech’ helmet or headset (Oxford Dictionaries, 2019). The list of applications already covers psychotherapy (Dibbets & Schulte-Ostermann, 2015), engineering (Freeman, Salmon, & Coburn, 2016), architecture (Blauert, Lehnert, Sahrhage, & Strauss, 2000), and also business applications, such as marketing (Barnes, 2016) and research (Meißner, Pfeiffer, Pfeiffer, & Oppewal, in press). As businesses and managers are projected to spend more time in VR (Madary & Metzinger, 2016), an increasing volume of economic transactions likely occurs in VR in the future. However, just as in the ‘real’ world (RL), the risk of dishonesty in VR prevails as a costly threat for these transactions (Weber et al., 2003). We argue that this requires a critical re-assessment of self-awareness and its influence on dishonesty, in the context of VR. The purpose of this thesis is to assess the influence of two known factors of self-awareness on dishonesty in VR. First, looking into a mirror has been shown to induce self-awareness and decrease dishonest behavior (e.g., Diener & Wallbom, 1976; Falk, 2017; Vallacher & Solodky, 1979). Second, wearing a mask as a disguise has been shown to inhibit self-awareness and increase dishonest behavior (e.g., Diener, Fraser, Beaman, & Kelem, 1976; Miller & Rowold, 1979; Silke, 2003). We hypothesize that both effects can be transferred to VR with similar implications. Our objective is to test this claim in an experiment and discuss possible implications of dishonesty in VR. Therefore, our analysis adds empirical evidence to the research gap of self-awareness and dishonesty in VR. As we show, this evidence is of practical relevance. Our results allow for drawing practical implications to help managers and businesses in designing honesty-promoting virtual environments (VEs).
The development of this thesis is structured as follows: First, we will review the prevalent literature on self-awareness, dishonesty, and VR. This literature review identifies linking mechanisms between self-awareness and dishonesty, which could also exist in VR. We use these insights to develop two central hypotheses. Second, we outline the experimental methodology we employed to test these hypotheses. Third, data from the experiment is analyzed to formulate a model that describes the results. Fourth, we interpret and discuss our findings, and draw implications for managers and businesses. The thesis closes with an overview of possible limitations and an outlook for future research.
Chapter 2: Literature Review
2.1 Literature Search
The objective of our literature review is to provide a basic understanding of the necessary fields of research to practitioners and non-specialists in the economics domain. The three fields are (1) self-awareness, (2) dishonesty2 , and (3) VR. We specifically focus on linking factors and mechanisms between the fields, which can be used to develop the hypotheses. To keep this focus, we may not cover all sub-areas of each field in detail.
Our literature search strategy includes three approaches to find relevant papers in each field (see also Heugens & Lander, 2009). First, we researched papers using the four online databases EBSCO, JSTOR, PubMed, and APA PsycNET. The keywords searched for include “self”, “self-consciousness”, “self-awareness”, “mirror”, “mirror presence”, “deindividuation”, “deindividuation mask/masquerade/anonymity”, “honesty”, “dishonesty”, “dishonesty self/self-awareness/mirror/mask”, “lies”, “lying”, “virtual reality”, “body ownership”, “virtual reality consciousness/self-awareness/moral/truth-telling/lying/honesty/dishonesty”. Second, we applied the snowball technique to trace relevant references from the papers identified. In this step, we especially drew on a number of meta-studies and reviews from Gibbons (1990); Innocenti (2017); Jacobsen et al. (2018); Mol (2019); Postmes and Spears (1998); Rosenbaum et al. (2014); Silvia and Duval (2001). We searched EBSCO and Google Scholar to find referenced papers. Third, we also applied the cited-by technique to identify papers that cited papers of major theory, using Google Scholar. The papers used for this approach were from Carver and Scheier (1981); Diener (1977); Duval and Wicklund (1972); Gneezy (2005); Mazar, Amir, and Ariely (2008); Sanchez-Vives and Slater (2005); Slater and Sanchez-Vives (2016); Zimbardo (1969).
2.2 Self-Awareness
In the following part, we review the relevant theory of self-awareness. We start with a survey of the cognitive and behavioral reactions to self-awareness and conclude with identifying the influencing factors that drive self-awareness. We also draw implications for our hypothesis development.
2.2.1 Objective Self-Awareness Theory
Human consciousness, despite a vast body of research, cannot be entirely described by any finite definition so far (Seth, He, & Hohwy, 2015). A commonly employed approach differentiates between the two perceptions of the outer world, and the inner self (Edelman, 2003). This approach highlights the ability of a human to be aware of himself, thus to be self-aware (Carver & Scheier, 1981; Duval & Wicklund, 1972). Two definitions describe this phenomenon in more detail. First, G. H. Mead (1934) argues that the human self has the ability to think reflexively about itself as an own object, separate from the environment around it. He describes this ability as reflexivity (see also Leary & Tangney, 2012). Second, Duval and Wicklund (1972) argue that a human cannot solely be aware of the existence of the environment around him, but also of the existence of himself. They consider the self being able to actively think about itself as a separate object in the universe. Therefore, they define this state as objective self-awareness.
Following their definition, Duval and Wicklund (1972) developed the objective self-awareness theory, which discusses different implications of objective self-awareness on the cognition and behavior of an individual. According to their theory, the state of objective self-awareness can be initiated by any external factor that focuses the attention of an individual towards a salient inner self-dimension (see Gibbons, 1990). An increased self-awareness necessarily triggers a self-evaluative process (Wicklund, 1975), whereby the human individual compares himself to his ideal standards (Duval & Wicklund, 1972). According to Duval and Wicklund (1972), these standards are defined by a criterion of correctness, which comprises the mental model of correct attitudes and behavior for the individual (Silvia & Duval, 2001). More often than not, the individual realizes a negative discrepancy between his actual behavior and his desired attitudes and standards (Duval & Wicklund, 1972). This, in turn, exerts a motivational force to reduce this discrepancy (Duval & Wicklund, 1972; see also Wicklund, 1975). As a result, the individual may react in several ways, which will be reviewed later (see Section 2.2.3).
2.2.2 Self-Awareness and Self-Consciousness
How does objective self-awareness relate to human self-consciousness? According to Fenigstein et al. (1975), self-consciousness is a dispositional trait. They claim that some individuals consistently direct their attention rather inwards, whereas others consistently direct their attention rather outwards. In contrast, self-awareness is a situational state (Carver & Scheier, 1978; Duval & Wicklund, 1972). As Fenigstein et al. (1975) show, self-awareness can be induced in any situation by certain self-focusing factors, irrespective of disposition.
Though different in nature, Fenigstein et al. (1975) find that dispositional self-consciousness and situational self-awareness can have similar implications on human cognition and behavior (see D. M. Buss & Scheier, 1976; Scheier, 1976; Scheier & Carver, 1977). To name one, both high dispositional self-consciousness and high situational self-awareness increase aggression behavior in a comparable way, and even aggregate in effect (Scheier, 1976). Therefore, we can assume that most effects laid out in the following for different degrees of self-awareness are also valid for different degrees of self-consciousness.
2.2.3 Cognitive and Behavioral Reactions to Self-Awareness
Under salient self-awareness, an individual is likely to experience a negative discrepancy between himself and his standards (Duval & Wicklund, 1972; Wicklund, 1975). This discrepancy provokes several cognitive and behavioral reactions. In the following section, three such reactions are examined. They are ranked from the least to the highest mental effort (Heider, 1958), expressing mental favorability of the action (Silvia & Duval, 2001).
First, it has been assumed that an individual tries to actively avoid the self-awareness inducing stimulus, in response to the negative discrepancy (Duval & Wicklund, 1972). Wicklund (1975) suggests several avoidance actions. For instance, this includes the physical avoidance of mirrors (Burris & Lai, 2012; Duval & Wicklund, 1972) and of own voice tape-recordings (Gibbons & Wicklund, 1976). As we show later, both of these stimuli induce self-awareness (see Section 2.2.4). Furthermore, Wicklund (1975) hypothesizes that individuals even create distracting stimuli for avoidance, which is consistent with intentional distraction used in stuttering therapy (Dittmann & Llewellyn, 1969; Swift & Hedrick, 1917). We, therefore, expect to observe distracting behavior among individuals of increased self-awareness.
Second, if an individual cannot avoid the stimulus sufficiently, classical theory predicts he will try to reduce the discrepancy by changing his behavior to match his standards (Duval & Wicklund, 1972). According to the theory, self-awareness induces normative behavior (Wicklund, 1979). Numerous studies have demonstrated increased conformity of self-aware individuals to moral standards (e.g., Batson, Kobrynowicz, Dinnerstein, Kampf, & Wilson, 1997; Carver, 1974, 1975; Diener & Wallbom, 1976; Falk, 2017; Vallacher & Solodky, 1979). Employing this line of thought, we can hypothesize that individuals behave compliant with moral standards in situations of increased self-awareness.
Third, more recent research also suggests the possibility of an individual to adjust his desired standards to match his actual behavior, effectively leading to reduced discrepancy as well (Silvia & Duval, 2001). As Dana, Lalwani, and Duval (1997) show in two experiments, this occurs when an individual rather focusses on his standards than on his behavior. This relates to the phenomenon of moral hypocrisy (Batson et al., 1997; Baumeister & Newman, 1994), which describes a change of the ‘moral’ standards to ‘immoral’ standards (Silvia & Duval, 2001). More specifically, this means that an individual tries to appear morally compliant to himself and others while avoiding the actual costs of morality (Batson et al., 1997). In an extensive study, Batson, Thompson, Seuferling, Whitney, and Strongman (1999) observe that, when individuals are not aware of any standards ex-ante, they often construct standards ex-post that match to their immoral behavior. This relates to the phenomenon of ethical blindness (Chugh, Bazerman, & Banaji, 2005; Gino, Norton, & Ariely, 2010).
2.2.4 Influencing Factors on Self-Awareness
Several factors that direct attention to the self or to the environment, thus influencing the degree of self-attention and self-awareness, have been proposed in research (see Gibbons, 1990; Silvia & Duval, 2001 for a review). We proceed to categorize these into (1) factors that increase self-awareness and (2) factors that decrease self-awareness. This enables us to highlight one factor of each category. These factors constitute the basis of our hypotheses.
(1) First, we consider factors that increase self-awareness by guiding the attention towards the self (Wicklund, 1975). One of the most prominent of such is the presence of a mirror, allowing to observe one’s own presence (e.g., Brockner, Hjelle, & Plant, 1985; D. M. Buss & Scheier, 1976; Carver, 1974, 1975; Froming, Walker, & Lopyan, 1982). A similar effect has also been observed for the presence of an active recording device, such as a video camera or a tape recorder (e.g., Davis & Brock, 1975; Duval & Wicklund, 1972; Geller & Shaver, 1976; Insko, Worchel, Songer, & Arnold, 1973). Interestingly, also the presence of an audience increases the self-awareness of an individual (e.g., A. H. Buss, 1980; Carver & Scheier, 1978; Innes & Young, 1975; Scheier, Fenigstein, & Buss, 1974). According to Gibbons (1990), mirror presence is best for inducing an unbiased, private state of self-awareness. According to him, this is because mirrors purely focus on private self-attention, whereas other factors (e.g., cameras, tape-recordings, audience) also focus on public self-attention (see also A. H. Buss, 1980) and the external social image (Falk, 2017). We, therefore, employ mirror presence as a mechanism to induce self-awareness in our analysis.
Carver and Scheier (1978) were the first to empirically research the ‘pure’ effect of a mirror on self-awareness. In their Self-Focus Sentence Completion study (Exner, 1973), they demonstrated that mirror presence increases self-aware sentence completions. Mirror presence has shown to influence human cognition and behavior across a number of empirical studies (e.g., Brockner et al., 1985; D. M. Buss & Scheier, 1976; Carver, 1974, 1975; Froming et al., 1982). For instance, Liebling and Shaver (1973) and Plant and Ryan (1985) demonstrate that mirror presence has an effect on task performance. They found that individuals who completed a cognitive task while facing a mirror outperformed those who did not face a mirror. This is an effect similar to social facilitation (Zajonc, 1965; Zajonc & Sales, 1966). In a different study, Brockner et al. (1985) found that mirror presence increases the feeling of a strong depression among participants, possibly due to increased self-evaluation. This is consistent with research by Radell, Keneman, Adame, and Cole (2014), who show that mirrors in dance classrooms can cause a poor body image, due to increased self-evaluation.
Increased self-evaluation also applies to some studies which research the consequences of mirror presence on antinormative behavior. For instance, Diener and Wallbom (1976) observed a lower cheating rate in an anagram test among subjects who were seated in front of a mirror than to those who were not. In their puzzle task, Vallacher and Solodky (1979) found similar effects of mirror presence on cheating. More recently, Falk (2017) demonstrated that also a ‘virtual’ webcam mirror induces mirror presence. He observed that subjects, who faced a virtual mirror of themselves, showed significantly less immoral behavior, relative to subjects who did not. Moreover, he hypothesized that this effect is also applicable to dishonest behavior. All of the mentioned authors conclude that mirror presence induces self-awareness, which leads to reduced antinormative behavior. From these findings, we hypothesize that we can use mirror presence to reduce antinormative behavior in our analysis. (2) Second, we consider the factors that decrease self-awareness. On the one hand, a decrease in self-awareness might occur as a factor guides the attention away from the self and to the environment (Wicklund, 1975). Such factors include any simple visual or auditory distraction, like a captive movie (e.g., Carver & Scheier, 1981; Wicklund, 1975), a cognitively exhausting task (e.g., Gino, Schweitzer, Mead, & Ariely, 2011), or a distracting game (e.g., Zanna & Aziza, 1976). In our analysis, we can use these insights to eliminate and control for any environmental distraction that could have a decreasing effect on self-awareness.
On the other hand, a decrease in self-awareness can also happen as a factor inhibits the mere ability of an individual to discriminate between himself and his environment (Wicklund, 1975). The resulting state of lowered self-awareness is described as deindividuation (Festinger, Pepitone, & Newcomb, 1952), and leads to increased antinormative behavior (Zimbardo, 1969). Zimbardo (1969) and Diener (1977) reviewed an extensive list of factors that initiate a deindividuated state, including anonymity (Diener, 1976; Diener et al., 1976), group presence (Diener, 1976; Diener et al., 1976; Diener, Lusk, DeFour, & Flax, 1980; Westford, Diener, & Diener, 1973), and physical arousal (Westford et al., 1973). According to Diener et al. (1980), anonymity is the key antecedent of deindividuation and decreased self-awareness (see Diener, 1979). We, therefore, employ anonymity as a mechanism to inhibit self-awareness in our analysis.
Numerous studies provide evidence for the effect of anonymity and disguise on antinormative behavior (e.g., Diener et al., 1976; Ellison, Govern, Petri, & Figler, 1995; Miller & Rowold, 1979; Rehm, Steinleitner, & Lilli, 1987; Silke, 2003; Zimbardo, 1975). For instance, Silke (2003) observes that wearing a mask significantly increases aggression and violence among offenders in Northern Ireland. In their Halloween field study, Miller and Rowold (1979) also demonstrated that masked children showed significantly more stealing behavior than unmasked children. This is consistent with an earlier Halloween field study by Diener et al. (1976). From these findings, we hypothesize that we can use a mask to decrease self-awareness in our analysis.
2.3 Dishonesty
In the next part, we review the relevant literature on dishonesty. Revisiting the underlying motives and mechanisms of dishonesty, the key influencing factors on dishonesty are discussed. We specifically highlight the effect of self-awareness on lying behavior to link both strands of theory. We use these findings to further develop our hypotheses and identify potential side effects.
2.3.1 Motives for Dishonesty
Honesty is generally considered an expected normative standard (Goldsen et al., 1960), thus inherently carrying benefits for honest individuals in society (Wiltermuth, Newman, & Raj, 2015). Therefore, dishonest behavior can only occur when the individual perceives greater value from the consequences of dishonesty than from honesty (Rixom & Mishra, 2014). Under this necessary condition, we can define four different types of lies, based on the consequences arising from them, for oneself and for others (Gneezy, 2005). Each type has different consequences, acting as motives to or not to engage in the lie.
First, Pareto white lies are pro-social and create a benefit for both sides or at least others (Gneezy, 2005). Engaging in such lies builds trust (Levine & Schweitzer, 2014) and even improves the moral image others have about oneself (Levine & Schweitzer, 2015). Second, altruistic white lies are pro-social and create benefit for others, but not for oneself (Gneezy, 2005). A motive to engage in this lie might be mere altruism (Becker, 1976; Gneezy, 2005), or a preference for giving (Andreoni, 1990). Third, spiteful black lies are anti-social and do not create benefit for any side, but can create harm for others (Gneezy, 2005). A motive to engage in this lie is to decrease the reputation of the other or create costs for him as a reaction to a prior conflict (Gneezy, 2005). Fourth, selfish black lies are anti-social and only create benefit for oneself, but not for others (Gneezy, 2005). Clearly, a common motive to engage in such lies are material gains that can only be obtained by lying (DePaulo & Kashy, 1998; Steinel & De Dreu, 2004; Wiltermuth et al., 2015). However, this comes at the risk of others also lying to oneself (Cialdini, Kallgren, & Reno, 1991), a feeling of guilt about others‘ disappointed expectations (Charness & Dufwenberg, 2006; Dufwenberg & Gneezy, 2000; Weibull & Villa, 2005), a degraded the general trustworthiness of others towards oneself (Tyler, Feldman, & Reichert, 2006), and a loss of one’s moral self-image (Fischbacher & Föllmi-Heusi, 2013; Mazar et al., 2008; Shalvi & Leiser, 2013). Selfish black lies are most prevalent among society (Gneezy, 2005) and most relevant in the economics domain (Erat & Gneezy, 2012). We, therefore, focus our analysis on dishonesty of this type.
2.3.2 Major Mechanisms of Dishonesty
After having reviewed different motives behind engaging or not engaging in a lie, understanding by which mechanisms an individual may weigh and decide between these motives is the next step. Two lines of research dominate the field and provide different theories regarding the mechanisms of lying.
The first line of research refers to the classical economic model of dishonesty (Becker, 1968; Lewicki, 1984; see Rosenbaum et al., 2014). Becker (1968) first employed this line of thought to explain the rationality behind the crime. In his research, he assumes that crime is committed consciously and under a deliberate trade-off between the risk-weighted rewards and costs of the criminal activity. According to him, this presents an algebraic optimization problem with one optimal solution. Following this model, any dishonest action is a rational and economic decision of utility maximization (Lewicki, 1984). This is consistent with the notion of the homo oeconomicus (Persky, 1995). Subsequent models have adapted the basic model to better explain specific areas of dishonest behavior, for instance, income tax evasion (Allingham & Sandmo, 1972) and solidarity across international communities (Hechter, 1990). In order to mitigate dishonest behavior, strategies for policymakers suggest choosing a level of control and punishment that makes dishonest behavior economically unreasonable (Allingham & Sandmo, 1972; Becker, 1968; Tulkens & Jacquemin, 1971) and increases rewards for honesty (Somanathan & Rubin, 2004).
Despite its resonance, most empirical research is not in line with the implications of the basic model of Becker (1968) (see Jacobsen et al., 2018; Rosenbaum et al., 2014). For instance, even when there is no risk of detection and an individual faces no punishment for dishonest behavior, individuals tend to reveal truthful private information (e.g., Lewis et al., 2012) and do not lie to the maximum extent (e.g., Abeler et al., 2014; Erat & Gneezy, 2012; Gneezy, 2005; Jacobsen & Piovesan, 2016; Mazar et al., 2008). This suggests that economic incentives and dishonesty are not related in such predetermined and rational way, as it was previously assumed (Cappelen, Halvorsen, Sørensen, & Tungodden, 2017). Therefore, a different line of research is needed to explain these differing results.
The second line of research refers to the social psychology model of unconditional honesty (Abeler et al., 2014; Campbell, 1964; Henrich et al., 2001; Rosenbaum et al., 2014; Somanathan & Rubin, 2004). In general, any behavior internalized by large parts of society and considered as generally acceptable can constitute a social norm (Pruckner & Sausgruber, 2013; Rosenbaum et al., 2014). The internalization of this social norm into an individual’s psychology implies that the individual may face internalized costs of lying when engaging in dishonest behavior (Abeler et al., 2014). In other words, engaging in honest behavior is rewarded positively, whereas engaging in dishonest behavior is punished (Campbell, 1964; Henrich et al., 2001). As a result, dishonest behavior should be unattractive to the individual (Rosenbaum et al., 2014). The internalized social norm of honesty should result in ethical individuals who do not lie at all (Gibson, Tanner, & Wagner, 2013; Gneezy, 2005).
However, this line of theory appears to be too optimistic. Individuals may not always lie to the maximum extent, however, they still clearly engage in dishonest behavior (e.g., Abeler et al., 2014; Erat & Gneezy, 2012; Gneezy, 2005; Jacobsen & Piovesan, 2016; Mazar et al., 2008). Furthermore, this theory would contradict with empirical field results (Graham et al., 2002; Mazar & Ariely, 2006; Weber et al., 2003). Therefore, the internalization of norms alone fails to fully explain lying either (Rosenbaum et al., 2014).
Both lines of research cannot explain the phenomenon of incomplete dishonesty (Shalvi, Dana, Handgraaf, & De Dreu, 2011) on their own. However, a certain balancing mechanism between the economic factors and the psychological costs of lying could provide an explanation for this phenomenon and a new line of research (Rosenbaum et al., 2014). One of such mechanisms is provided by N. L. Mead, Baumeister, Gino, Schweitzer, and Ariely (2009), who hypothesize that a process of balancing occurs between competing impulses, in their case the temptation of cheating and the resistance from self-control. Another mechanism is provided by Mazar et al. (2008), who hypothesize that individuals balance between an economic gain from lying and psychological maintenance of a positive self-image (see Aronson, 1969; Harris, Mussen, & Rutherford, 1976). In six experiments about dishonesty, they consistently demonstrated that individuals indeed preferred to maintain a positive self-image of themselves. By choosing a strategy of lying partially and insignificantly, individuals can realize a slight economic gain while still maintaining their positive self-image as an honest person (Ayal & Gino, 2011; Gino, Ayal, & Ariely, 2009; Mazar et al., 2008).
Incomplete dishonesty and partial lying due to self-concept maintenance have since been demonstrated across a number of empirical studies (Fischbacher & Föllmi-Heusi, 2013; Haan & Kooreman, 2002; Mazar et al., 2008; see Rosenbaum et al., 2014 for a review; Shalvi et al., 2011; Steinberg, McDonald, & O'Neal, 1977). One common experimental design employed is the die-under-a-cup paradigm by Fischbacher and Föllmi-Heusi (2013). In this paradigm, dishonesty cannot be detected for an individual but only for the overall population, which guarantees total anonymity (see Fischbacher & Föllmi-Heusi, 2013; Mazar et al., 2008; Schweitzer, Ordóñez, & Douma, 2004). As predicted by self-concept maintenance theory, even under total anonymity, only some individuals lie to the full extent (20%) or not at all (39%), while many individuals show partial lying behavior (41%) (Fischbacher & Föllmi-Heusi, 2013).
2.3.3 Alternative Mechanisms of Dishonesty
Besides self-concept maintenance, a few alternative mechanisms of incomplete dishonesty and partial lying have been proposed, which are very similar in effect (see Jacobsen et al., 2018 for a review). We can hypothesize that these add to the mechanism of self-concept maintenance. Later, we can use them to interpret additional effects and secondary findings.
First, moral balancing provides another explanation for partial lying (Nisan, 1991). As Nisan (1991) proposes, individuals keep a balance (like an account) of their moral actions. He argues that this balance allows individuals to trade-off immediate morally incompliant behavior with recent compliant behavior. Therefore, he argues that individuals can choose to lie and still feel morally good, as long as their overall balance is positive. Empirical validation for this mechanism indeed shows that when individuals have just behaved morally right, like acting in an environmentally-friendly way (Mazar & Zhong, 2010), they are more willing to engage in morally wrong behavior subsequently (see Cojoc & Stoian, 2014; Ploner & Regner, 2013; Vincent, Emich, & Goncalo, 2013). This effect also works vice-versa. As Barkan, Ayal, Gino, and Ariely (2012) show, reminding individuals of their own morally wrong behavior increases the willingness to do morally good.
Second, Shu, Gino, and Bazerman (2011) theorize an alternative mechanism of moral disengagement. According to this mechanism, engaging in unethical behavior creates a feeling of moral disengagement and decreases the individual importance of ethical behavior. This is very close to Nisan (1991), who suggests that individuals act under limited morality and excuse themselves from high moral standards that they expect from others. For instance, religious individuals often expect much higher standards from others than they can live up to themselves (Shalvi & Leiser, 2013; Utikal & Fischbacher, 2013).
Third, self-serving justifications also provide an alternative explanation for partial lying (Jacobsen et al., 2018). Individuals may justify dishonest behavior by a flawed justification process to excuse immoral behavior ex-ante, or to prevent a deterioration of the positive moral self-image ex-post (Shalvi, Eldar, & Bereby-Meyer, 2012; Shalvi, Gino, Barkan, & Ayal, 2015). The vast extent of justifications in different situations is demonstrated by Gino and Galinsky (2012), who show that even slight cues of justification in the immediate environment are employed by individuals to engage in dishonest behavior. Shalvi et al. (2015) hypothesize that other mechanisms like positive self-image maintenance (Mazar et al., 2008) and moral balancing (Nisan, 1991) could be justification strategies themselves.
2.3.4 Influencing Factors on Dishonesty
The empirical evidence of the mechanisms reviewed suggests that honesty is a dynamic and malleable concept for individuals, susceptible to predisposition (Abeler et al., 2014) and situational factors (Mazar et al., 2008). Understanding which traits have to be controlled for, and which contextual cues can be employed to influence dishonest behavior is essential for the development of our analysis.
First, we define predisposition factors as fixed characteristics and preferences which are stable across different situations for an individual. Most demographic factors do not show to have any robust effect on dishonesty (see Rosenbaum et al., 2014), including age (e.g., Abeler et al., 2014; Childs, 2012; Conrads, Irlenbusch, Rilke, & Walkowitz, 2013) , income (e.g., Abeler et al., 2014; Haan & Kooreman, 2002), and religiousness (e.g., Abeler et al., 2014; Ruffle & Tobol, 2014; Utikal & Fischbacher, 2013). Some studies find significant differences across gender, indicating that female individuals cheat less than male (e.g., Conrads, Irlenbusch, Rilke, Schielke, & Walkowitz, 2014; Conrads et al., 2013; Houser, Vetter, & Winter, 2012; Ward & Beck, 1990). However, this effect is not robust (Abeler et al., 2014; Diener & Wallbom, 1976; Lewis et al., 2012; Piazza, Bering, & Ingram, 2011). Furthermore, individuals majoring in economics are found to be significantly more dishonest than non-economists (Childs, 2012; Lewis et al., 2012; Lundquist, Ellingsen, Gribbe, & Johannesson, 2009). We can later employ these factors as control variables in our analysis where necessary.
Second, we define situational factors as contextual variables (Mazar et al., 2008) and priming factors (Friesen & Gangadharan, 2013), specific to a situation. For instance, monitoring, or even the mere feeling of being observed, has been robustly found to have a positive effect on honesty (e.g., Fischbacher & Föllmi-Heusi, 2013; Howells, 1938; Rustagi, Engel, & Kosfeld, 2010). In the context of our methodology, we control for this factor. Furthermore, the psychological costs of lying in a particular situation also present a robust positive effect on honesty (e.g., Frey & Meier, 2004; Gneezy, 2005). These psychological costs of lying can be described as the distance of a dishonest individual to the victim (Rosenbaum et al., 2014). In our methodology, we also control for this factor of influence by maintaining a high interpersonal distance to each individual. Other factors that have been identified include cognitive demand (e.g., Baumeister, Vohs, & Tice, 2007; Gino et al., 2011; N. L. Mead et al., 2009; see Muraven, Tice, & Baumeister, 1998), ego depletion (e.g., Achtziger, Alós-Ferrer, & Wagner, 2015), and time pressure (e.g., Shalvi et al., 2012). We use these factors to include potential side effects in our discussion.
Similar in effect, a number of priming factors have been proposed to promote honesty (Jacobsen et al., 2018; Rosenbaum et al., 2014). Jacobsen et al. (2018) differentiate between explicit and implicit moral cues as promoting factors. An example of the former could be reminding an individual about the ten commandments (Mazar et al., 2008) or about his moral self (Bryan, Adams, & Monin, 2013; Pruckner & Sausgruber, 2013), both leading to increased honesty. An example of the latter could be asking an individual for his signature (Shu & Gino, 2012; Shu, Mazar, Gino, Ariely, & Bazerman, 2012) or creating a feeling of ethical dissonance (Barkan et al., 2012; see also Festinger, 1957), also leading to increased honesty. Interestingly, placing a mirror in front of an individual also provides an implicit moral cue through mirror presence, leading to more honest behavior (Diener & Wallbom, 1976; Falk, 2017; Gino & Mogilner, 2014; Vallacher & Solodky, 1979). In contrast, anonymity factors like masks blocks this implicit moral cue (Diener et al., 1976; Miller & Rowold, 1979; Silke, 2003). This empirical evidence provides the critical link between the findings of self-awareness (see Section 2.2) and the findings of dishonesty (see Section 2.3). To summarize, increasing the self-awareness of an individual, for instance through mirror presence, provides an implicit moral cue, which engages self-concept maintenance, and, therefore, promotes honest behavior (Jacobsen et al., 2018). In contrast, anonymity and deindividuation, by wearing a mask, inhibit this moral cue, which, in turn, does not promote honest behavior (Jacobsen et al., 2018).
2.4 Virtual Reality
After having built a sufficient understanding of the mechanisms of self-awareness and dishonesty, we now provide a basic overview of VR as a new environment for these effects. The general analysis is limited to the terminology, to put more emphasis on relevant psychological mechanisms that have already been examined in VR. These may support our hypotheses with evidence from VR, and provide ground for further discussion.
2.4.1 Technical Terminology
VR can be defined as “the computer-generated simulation of a three-dimensional image or environment that can be interacted with in a seemingly real or physical way by a person using special electronic equipment (…)” (Oxford Dictionaries, 2019). In research, Innocenti (2017) differentiates between two classes of VR experiences. He considers a low-immersive virtual environment (LIVE) as a simulation rendered on the 2D desktop screen of a computer. In contrast, he defines a high-immersive virtual environment (HIVE) as a 3D projection of a simulation that fills the entire field of view. This thesis only considers the latter category, as extensive economic research has already been conducted on LIVE (see Innocenti, 2017), and as HIVE are likely to become relevant in the near future (Madary & Metzinger, 2016). To deliver a HIVE experience, modern VR equipment often uses a head-mounted display (HMD), which is a headset with a small integrated screen inside (Mol, 2019). In addition, the equipment often makes use of controllers and motion sensors (Mol, 2019). Two photographic descriptions of such a VR setup can be found in Figure A1.
VEs often include virtual humans. In this context, Bailenson and Blascovich (2004) distinguish between agents that interact by algorithmic commands and avatars that interact by human commands. A VR experience can induce a feeling of presence in an individual (Bombari, Schmid Mast, Canadas, & Bachmann, 2015; Sanchez-Vives & Slater, 2005). This feeling causes an individual to perceives himself as if he actually is in the VE and actually interacts with the virtual surrounding (Bombari et al., 2015; Sanchez-Vives & Slater, 2005). The presence effect is a highly subjective illusion created by the VR experience (Mol, 2019). As a result of feeling present, VR experiences can induce a feeling of virtual embodiment, as explained by Slater and Sanchez-Vives (2016). They describe this as an illusion of body ownership, which causes an individual to forget about his body in RL and rather perceive it as displayed in VR. Adapting the terminology of Bailenson and Blascovich (2004), an individual would, therefore, identify with his avatar as if it was himself.
One of the most important influencing factors on both presence and virtual embodiment is the degree of immersion (see Slater & Sanchez-Vives, 2016). Bombari et al. (2015) define immersion as “the objective amount and quality of (…) perceptual input provided to the participant through technology” (p. 3). For instance, using photo-realistic avatars (Zhang & Hommel, 2016) and projecting the RL limb movements of an individual onto their VR limbs by motion sensors (Sanchez-Vives & Slater, 2005) increases immersion. Ultimately, the higher the immersion, the more present individuals feel in the VE, and the more likely it is that they perceive their avatar in VR as themselves (Mol, 2019; Slater, 2009). We later employ these factors in our experiment to create a highly immersive experiment, promoting natural behavior.
2.4.2 Experimental Advantages of VR
In addition the growing practical relevance, VR is also relevant for economic experiments, as naturalistic settings can be tested with high control over environmental factors (Innocenti, 2017). We suggest this gives VR experiments several advantages over traditional experimental designs, especially relevant in the fields of self-awareness and dishonesty.
First, VR experiments can test factors and stimuli of self-awareness manipulation that could not be tested in RL (Mol, 2019). For instance, disguising the outer appearance using a traditional mask only provides limited anonymity and deindividuation (Diener et al., 1976; Miller & Rowold, 1979). VEs, in contrast, enable the use of avatars to completely control for all aspects of the appearance of an individual (Thaler et al., 2018). This enables entirely new and also non-human disguises in VR (Mol, 2019). We make use of this advantage to test the influence of non-human avatars on self-awareness. Second, the VE of a VR experiment can be completely standardized and controlled in all details (Mol, 2019). This allows for the controlled manipulation of one exact variable, while all other factors remain constant across the different treatments (Peck, Seinfeld, Aglioti, & Slater, 2013). We use this principle to ensure no other factors biased the treatments. Third, VEs enable the automated logging of data that could not be captured easily in a RL experiment, such as data on the exact position and rotation of an individual (McCall & Singer, 2015; Parsons, 2015). Especially relevant for the domain of dishonesty, response times can be recorded with a precision of milliseconds, and be employed as a measure of lying (Suchotzki, Verschuere, Van Bockstaele, Ben-Shakhar, & Crombez, 2017; Walczyk, Mahoney, Doverspike, & Griffith-Ross, 2009; Walczyk, Roper, Seemann, & Humphrey, 2003). We make use of this advanced measure for additional analysis.
2.4.3 Psychological Effects in VR
To the best of our knowledge, no study has yet investigated the effect of mirror presence on dishonesty in VR. However, a number of other, possibly related psychological effects have been assessed. Therefore, a review of these effects is relevant to capture effects of self-awareness in VR.
The basic condition for any psychological effect to occur in VR is the illusion of presence (Sanchez-Vives & Slater, 2005). If salient dimensions of immersion are satisfied by the VR system, the individual experiences presence as a sense of consciousness for his virtual body (Kilteni, Groten, & Slater, 2012; Slater & Sanchez-Vives, 2016) and the environment around (Bombari et al., 2015; Sanchez-Vives & Slater, 2005). As Slater (2009) proposes, this consciousness initiates behavior in VR that corresponds to ‘realistic’ behavior, as expected in RL. For instance, Zhang and Hommel (2016) demonstrate that individuals respond to the simulated physical threat in VR with recessive behavior, as expected under real conditions. Interestingly, they also observe this recessive behavior when individuals can only see their avatar in a mirror, suggesting mirror presence is possible in VR. Realistic behavior even occurs in situations of extreme arousal in VR, as Slater et al. (2006) show. In their replica of the famous Milgram obedience experiment (Milgram, 1963), participants had to administer electric shocks to a computer-generated virtual agent in VR, yet showed similar levels of arousal, as observed in the original experiment in RL (Slater et al., 2006).
Realistic behavior can also be observed in interactions with other agents and avatars. For instance, individuals in VR keep more distance to agents representing others than to agents representing themselves, which is consistent with intimacy effects in RL (Bailenson, Blascovich, & Guadagno, 2008). Jorjafki, Sagarin, and Butail (2018) show that the behavior of other agents in VR can even trigger behavioral contagion and change one’s own behavior (cf. Gallup, Vasilyev, Anderson, & Kingstone, 2019). This gives rise to the idea that other agents, as a trigger for self-awareness (Aymerich-Franch, Kizilcec, & Bailenson, 2014), could possibly have an impact on dishonesty. Indeed, Mol, van der Heijden, and Potters (2018) identified that an active observer agent significantly reduces dishonest behavior in VR, compared to the absence of such an agent. We, therefore, hypothesize that self-awareness also exists in VR and yields similar effects on dishonest behavior, as observed in RL (see Section 2.2.4).
However, not only do other agents and avatars have implications on behavior, but also characteristics of the own avatar itself. Across a number of studies, individuals change their behavior depending on their avatar height (Yee & Bailenson, 2007), weight (Fox & Bailenson, 2009), skin color (Peck et al., 2013), or realism (Mol, 2019), irrespective of their real characteristics (Slater & Sanchez-Vives, 2016). This is also known as the Proteus effect (Yee & Bailenson, 2007). The empirical evidence, therefore, suggests that individuals become self-aware of their own avatars. Moreover, Aymerich-Franch et al. (2014) find that individuals, who dislike public speaking because of anxiety, rather choose avatars that do not resemble their own appearance. Consequently, we can hypothesize that individuals may strategically become anonymous and avoid self-awareness by choosing an avatar that acts as a mask (Miller & Rowold, 1979).
2.5 Hypotheses
An extensive review about the topics of self-awareness (see Section 2.2), dishonesty (see Section 2.3), and VR (see Section 2.4) has been given. The following part aims to summarize and link these findings to develop two central hypotheses for the upcoming analysis.
A state of self-awareness initiates a process of self-evaluation, which makes the individual aware of discrepancies between his actual behavior and his desired standards (Duval & Wicklund, 1972; Wicklund, 1975). One of the most important of such standards is honesty. Different factors have been identified which either increase or decrease self-awareness (see Section 2.2.4). In the context of honesty, therefore, they draw an individual’s attention more or less to his moral discrepancy, and, consequently, either inhibit or engage dishonest behavior (see Section 2.3.4). As VR gains in relevance (Madary & Metzinger, 2016), we may investigate to what extent these effects are valid in VR. As a number of related psychological effects have already been shown in VR (Aymerich-Franch et al., 2014; Slater et al., 2006; Yee & Bailenson, 2007; Zhang & Hommel, 2016), we assign higher relevance to our analysis. Two factors and their effects have been selected to be of primary interest for our analysis.
First, mirror presence increases self-awareness (Carver & Scheier, 1978), and, therefore, inhibits dishonest behavior in RL (Diener & Wallbom, 1976; Falk, 2017). Furthermore, an individual can become aware of their avatar in VR (Sanchez-Vives & Slater, 2005). We, therefore, hypothesize that a mirror in VR also induces mirror presence, which increases self-awareness and inhibits dishonest behavior.
Hypothesis 1 (H1): The presence of a mirror leads to reduced dishonesty in VR.
Second, masks induce anonymity and deindividuation, which decreases self-awareness (Zimbardo, 1969), and, therefore, engages dishonest behavior in RL (Diener et al., 1976; Miller & Rowold, 1979). Furthermore, avatars that do not resemble an individual can provide anonymity (Aymerich-Franch et al., 2014). Thus, we hypothesize that a fictitious, non-human avatar in VR also acts as a mask and induces deindividuation. This avatar decreases self-awareness and motivates to engage in dishonest behavior.
Hypothesis 2 (H2): A non-human avatar leads to increased dishonesty in VR.
Chapter 3: Methodology
3.1 Objective
The central hypotheses aim to understand whether mirror presence and avatar choice have an effect on dishonesty in VR. To test the hypotheses, a between-subjects experiment was conducted at WHU – Otto Beisheim School of Management. The experiment was supervised and sponsored by the IHK-Chair of Small and Medium-Sized Enterprises at WHU.
Testing the hypotheses involves two components: (1) Introducing participants to a task in VR that incentivizes dishonest behavior under anonymity, and (2) manipulating mirror presence and avatar choice while performing the task. For the first component (1), we employed the mind game paradigm (Greene & Paxton, 2009; Jiang, 2013) – a variation of the die-under-a-cup paradigm (Fischbacher & Föllmi-Heusi, 2013) – and adapted it to VR (see also Mol et al., 2018). For the second component (2), we altered the presence of a mirror in VR (H1) and introduced a human and a non-human avatar for the participant to choose (H2).
Experiment participants were randomly assigned to either a mirror treatment or a no-mirror treatment. Moreover, participants could choose between a human avatar and a non-human avatar . The effect of the factors on self-awareness, and, therefore, on dishonesty in VR, would be any difference in the sample results between the different conditions.
Moreover, we collected information about the demographics and the experience of a participant, using a questionnaire after the experiment (see Appendix E). First, this data is necessary to control for unwanted influencing factors on self-awareness and dishonesty (see Section 2.2.4; 2.3.4). Second, we also expect to find evidence for some of the mechanisms discussed earlier (see Section 2.3.2; 2.3.3).
3.2 Experimental Design
3.2.1 Participants
Participants were invited among business students who were enrolled at WHU – Otto Beisheim School of Management, campus Vallendar, at the time of the experiment. Two invitation emails were used to invite participants (see Figure B1; B2). The emails informed students about the voluntary nature of the experiment and motivated them to participate for experiencing VR and for receiving a financial reward. Neither email provided information on the specific purpose of the experiment, other than that it was an experiment in VR.
The first email was sent three weeks in advance to the experiment, on a Thursday at 11:00 am, to all enrolled Bachelor and Master students. A similar reminder email followed a week before the experiment, on Monday at 12:00 pm, to all enrolled Bachelor and Master students. In total, approximately 950 students received invitation emails. Both emails included a link to an online scheduling tool, where students could register for a time slot of 30 minutes to join the experiment (see Figure B3). Time slots were offered on five consecutive business days (Monday to Friday), from nine am to nine pm each day, resulting in 120 time slots in total. Following his registration, a participant received a confirmation email with general instructions about the experiment participation (see Figure B4).
In total, 88 students chose to participate in the experiment, out of which n = 45 were assigned to one of two experimental treatments of interest. A demographic overview of these participants of interest is provided in Table 1. Data from the remaining participants are discarded from any further analysis3.
Abbildung in dieser Leseprobe nicht enthalten
Table 1: Demographic overview of the n = 45 experiment participants.
Note: The table displays absolute numbers. Data was collected as part of a questionnaire participants had to fill out after the VR experiment (see Appendix E). Variable definitions: nationality (local if German, international otherwise); local student (local if permanently enrolled, exchange if enrolled for an exchange semester)
The experiment conducted was part of a joint collaboration experiment which aimed to test three hypotheses about influencing factors on dishonesty in VR. In total, the joint experiment involved 88 participants and four treatment conditions. Only data from n = 45 participants assigned to one of the two treatment conditions of interest is relevant for the analysis. The remaining 43 participants were assigned to either of the other two treatment conditions, which are not of interest for this analysis.
3.2.2 Procedure
Prior to the experiment, each participant was randomly assigned to one of the two treatments (see Section 3.2.3), using stratified random sampling. We stratified over the two variables (1) gender (male, female) and (2) time of day (morning, noon, afternoon, evening). First, we used (1) gender to upfront reduce the risk of unbalanced sampling between the treatments, due to an under-proportionate number of female to male participants (see Table 1). Second, we used (2) time of day to control for biases from different times of the day. The resulting treatment allocation assigned nno-mirror = 21 participants to the no-mirror treatment, and nmirror = 24 participants to the mirror treatment. A descriptive statistics test for biases in the treatment allocation was done as part of the analysis and is reported in Table 3 (see Section 4.1.1).
The experiment took place in four office rooms of WHU – Otto Beisheim School of Management, in close walking distance to the student campus of the university. The rooms were chosen, because they are isolated from the main student activities and buildings, reducing disturbance and noise from other students. Moreover, the rooms allowed for permanent installation of the VR equipment. The core technical setup consisted of a full HTC Vive© kit (HTC Corporation, Taoyuan), including one headset, two controllers, and two base stations (see Figure A1), and a standard ASUS© (ASUSTeK Computer Inc., Taipeh) with a NVIDIA GeForce© GTX 1070 Ti graphics card (NVIDIA Corporation, Santa Clara, CA). Both components were chosen due to their optimization for HIVE, which allows for smooth and high-resolution VR experiences. We identified this as a key factor to deliver a high degree of immersion in VR (see Mol, 2019). Two student research assistants (one female, one male student) from the supervising research chair conducted the experimental procedure, each in a different role. Both roles had specific instruction scripts, to ensure that every participant received exactly the same instructions (see Appendix F). Both roles were also switched after every day to reduce any potential assistant bias. The first role, the assistant, personally guided the participant through the different steps of the experiment. The second role, the analyst, controlled the experimental software and the data backup process during the experiment, from an isolated, separate room. The analyst did not leave the room or interact with the participant in any way. Therefore, participants only met the assistant and were not made aware of the analyst, to reduce any fear of observation.
The experiment was conducted over the course of five consecutive business days (Monday to Friday) to allow for maximum flexibility in students’ time preferences. Each participant had a time slot of 30 minutes. In preparation for each participant, the assistant cleaned the VR headset and controllers with disinfectant wipes and ensured that the windows and shutters were closed to prevent any distraction or observation from outside. Meanwhile, the analyst saved all data from the previous participant and initialized the experimental software for the next participant. If a participant arrived while another participant was still in the experiment, he was asked to wait in a separate waiting room to ensure that participants could not interact in any way during the experiment.
At the start of the experiment, the participant was welcomed by the assistant and guided to the briefing room. He was asked to take a seat and sign a form of informed consent (see Appendix C). The form was adapted from the Social & Behavioral Sciences Institutional Review Board from the University of Chicago. All participants were given the time to read the form and the opportunity to ask questions about their participation. After verifying his, consent, the participant had to agree not to disclose any information about the experiment before the end of the week by placing his signature. The latter measure was undertaken to prevent any spread of information about the experiment among prospect participants, who were all students from the same university campus. Neither the form nor the assistant gave any information on the specific purpose of the experiment other than that it involved VR.
Specific to his pre-allocated treatment, the participant then received an instructional text by the assistant, informing him about the planned procedure of the experiment (see Figure D1; D2). The assistant told the participant that he would leave the room and prepare the experiment until the participant had finished reading the text. Participants were left alone in the briefing room to make sure they took the time to carefully read and understand all instructions in detail.
Afterward, the participant was guided to the experiment room with the VR equipment and computer. The assistant first helped the participant to mount the headset and adjust the fit to his comfort until he confirmed he could see clearly. Then, the assistant instructed the participants into the use of the controllers and the buttons needed for the experiment. Before the experiment started, the assistant clearly stated that he would be outside the experiment room for the entire duration of the experiment. The participant was instructed to call out loud once the experiment is over so that the assistant knew when to re-enter. We explicitly emphasized this to make sure participants would not feel observed at any time. Afterward, the assistant left the room.
The core procedure of the experiment consisted of two parts, (1) a tutorial game and (2) a prediction task, both of which the participant played fully on his own in VR. First, (1) the tutorial game provided an entertaining and interactive VR experience for participants to accommodate with the VR equipment. We did not capture any data in this part, as our sole objective was to induce immersion and make the VR equipment feel natural to the participant, in preparation for the experiment following. Specifically, participants played the minigame Longbow from the VR game The Lab© (Valve Corporation, Bellevue, WA) (see Figure A2). We specifically chose this game for three reasons. First, the minigames of The Lab © are intended to showcase the possibilities of VR, and thus inherently aim to reach a high degree of immersion (Katz et al., 2016). Second, Longbow, as an archery game, required the participant to move naturally and realistically, increasing body ownership (Zhang & Hommel, 2016). Third, the game involved exactly the controller elements of aiming and clicking, which would be needed in the following task, making their use more intuitive. Participants were manually teleported from (1) the tutorial game into (2) the prediction task in VR after five minutes by the analyst.
Second, the (2) the prediction task provided the actual experimental task. As explained earlier, we employed the paradigm of a mind game (Greene, 2009; Jian, 2013) and adapted it to a VR setting (Mol et al., 2018). The prediction task was programmed in Unity 3D© (Unity Technologies Inc., San Francisco, CA). At the beginning of the task, the participant could choose to wear a male, a female, or a teddy bear avatar (see Figure 1). His preferred avatar choice of a human or non-human avatar was relevant to test the second hypothesis (see Section 3.2.3). Then, he was teleported into the treatment room corresponding to his pre-allocated treatment (see Figure 2). Only one of the treatment rooms contained a mirror to induced mirror presence and test the first hypothesis (see Section 3.2.3).
Abbildung in dieser Leseprobe nicht enthalten
Figure 1: Screenshot of the avatar selection room. From left to right: The male, the female, and the teddy bear avatar. Participants could select the avatar for the prediction task following afterward.
Abbildung in dieser Leseprobe nicht enthalten
Figure 2; Screenshots of the two treatment rooms. The left screenshot a) depicts the no-mirror treatment without any mirrors. The right screenshot b) depicts the minor treatment with two mirrors reflecting the avatar (currently male).
In the treatment room, the participant found himself standing on an elevated stage in front of four rows of empty tables and chairs. On the opposite side of the room was a large display. Above the display, a text label constantly showed his currently achieved balance, starting at EUR 0.5. First, this display briefly repeated the instructions given in the instructional text earlier. Then, the display showed a countdown of seven seconds, and the first round of the prediction task started. In each round, participants had seven seconds to predict which of two possible fruits (either an apple or a banana) would appear on the display after the countdown. The paiticipant knew from the instructions that the display only ever showed either an apple or a banana and that both fruits were equally likely to appear on the display. Moreover, he was instructed to do the prediction quietly in his mind. After the countdown, one of the two fruits randomly appeared on the display, and the paiticipant had to choose whether he had predicted correctly or not that this fruit would appear (see Figure 2).
At the start of each round, the participant also automatically received a bonus of EUR 0.5 on top of his balance. If he predicted correctly, he was instructed to click the "Correct (keep 0.5 EUR)" button to the left of the display to secure his bonus of EUR 0.5 in his balance. If he did not predict correctly, he was instructed to click the "Incorrect (lose 0.5 EUR)" button to the right of the display, and thereby lost his bonus of EUR 0.5 again from his balance. This prediction task was repeated for 20 rounds in total. After the 20 rounds, the display showed the final balance. This final balanced comprised the additional bonus earned. In total, the time spent in VR was approximately 15 minutes per paiticipant.
Throughout the task, participants were incentivized to behave dishonestly, as they could maximize their financial gains by reporting overly correct about then predictions. The paradigm of a mind game was chosen over other dishonesty paradigms for two reasons. First, it eliminates any possibility of observing or logging whether an individual paiticipant behaved dishonestly (see Fischbacher & Föllmi-Heusi, 2013; Mazar et al., 2008; Schweitzer et al., 2004). Thus, even if participants suspected being technically logged in VR, they knew that their dishonest behavior could not be detected. Second, this approach makes the use of any deception techniques unnecessary, and is, therefore, in line with modern economic research standards (see Ortmann & Hertwig, 2002). In addition, we deliberately framed bonuses under the principles of loss-aversion (see Grolleau, Kocher, & Sutan, 2016; Kahneman & Tversky, 1979). This framing was employed to further incentivize dishonest behavior and increase the spread in effect between the two treatments.
After finishing the task, the assistant re-entered the experiment room and helped the participant to demount the headset. The assistant then took the headset to check the final balance on the display himself and repeated the final balance in front of the participant. We included this cue to again give the participant the impression that he had not been observed. The participant was then brought back into the briefing room, where he was asked to fill out a questionnaire about his experiences during the prediction task in VR (see Appendix E). The questionnaire included questions about the participant’s demographics, as well as his perception of the task, of the avatar, and of the VE. In addition to the questionnaire, the participant was served chocolate cookies and bottled water, as many appeared exhausted after the experiment. The assistant left the room again to make sure participants did not feel observed or pressured in any way and to get the most accurate reflections to the questions posed. Moreover, we did not want the presence of participants’ pay-outs to have any impact on their reflection accuracy. After filling out the questionnaire, the participant received his final pay-out, consisting of the guaranteed show-up reward of EUR 7.50 plus the additional bonus gained, in cash. The assistant then reminded him about the non-disclosure agreement and thanked him for his participation in the experiment.
3.2.3 Treatments
The first hypothesis aims to test the effect of mirror presence on dishonesty in VR (H1). To test this hypothesis, we established two treatments that differed with regards to mirror presence. The first treatment (no-mirror) provided the control treatment. In this treatment, the VR room of the prediction task did not contain a mirror or other factor that could induce self-awareness (see Figure 2). In each round of the prediction task, when participants had to choose whether their prediction was correct or incorrect, they had to turn their head to either of the two walls adjacent to the display to click the corresponding button We intentionally left these walls white and blank, to prevent any kind of distraction.
In contrast, the second treatment (mirror) provided the positive treatment. In this treatment, instead of the two neutral walls, two mirrors were placed adjacent to the display4. Therefore, when participants had to make their choice and turn to either of the two mirrors to click the corresponding button, they had to face a reflection of their own avatar in the mirror (see Figure 2). This reflection was the manipulating element we employed to induce mirror presence and subsequently create a state of increased self-awareness.
For mirror presence to arise, we had to arrange that participants would intuitively recognize the reflective surface as a realistic mirror in VR. We specifically designed the mirrors to reflect every movement of the avatar without any lag. Synchrony is a key requirement for mirrors to be realistic in VR (Zhang & Hommel, 2016). To further enhance the effect of mirror presence on self-concept maintenance, we employed two principles from Falk (2017). First, similar to the ‘webcam zoom effect’ (Falk, 2017), we magnified the reflected mirror image, to fill a larger part of the visual field in VR. Second, we placed the buttons in a way that locating and clicking them required participants to look straight at their avatar. We did this to make sure participants could not avoid the mirror reflection of their avatar in any way (Falk, 2017) but had to directly face their avatar. We later asked participants about the perceived realism of the mirror, to control for this factor (see Appendix E).
The second hypothesis aims to test the effect of avatar choice on dishonesty in VR (H2). For this hypothesis, we allowed participants to choose either a human (male, female) or a non-human (teddy bear) avatar before entering the prediction task (see Figure 1). We specifically adapted the instruction text of the mirror treatment to mention the mirror in the text and in a screenshot (see Figure D2). Participants, therefore, knew the conditions they would be facing during the prediction task up front. We presume that participants could do a strategic avatar choice. We also asked participants whether they would have chosen another avatar instead in the questionnaire, to control for the factor (see Appendix E).
3.2.4 Variables and Measurement
To test the hypotheses and control for possible factors of bias, several variables were established and measured during the experimental procedure. Measurements were done by a logging algorithm, part of the experimental software. First, we chose an algorithm to eliminate the need for a manual observation of the participant. This allowed us to only capture data in an anonymized form, leaving all other details about a participant unobserved. Second, some variables of interest could not be measured manually, emphasizing the need for a logging algorithm.
The raw data output of the algorithm was in the form of a TXT file (see Figure A3). The respective variables can be retrieved from the raw data by formatting and splitting the lines into rows and columns. An overview of all variables measured by the logging algorithm is given in Table 2. The variables logged were intended as regressors for regression analysis.
Abbildung in dieser Leseprobe nicht enthalten
Table 2 : Overview of variables measured by the logging algorithm.
Note: Measurements were only taken during the prediction task in VR (see Section 3.2.2)
Besides the variables logged by the algorithm, several additional variables were established and measured by the questionnaire (see Appendix E). Most variables in the questionnaire were intended to serve as control variables for regression analysis.
Chapter 4: Results and Discussion
4.1 Results
4.1.1 Descriptive Statistics
Before testing the two hypotheses, we first test for a potential bias between the two treatments. We employ the common approach of descriptive statistics for several factors of relevance (see Table 3). Data is drawn from the experimental software and the participant questionnaires.
Abbildung in dieser Leseprobe nicht enthalten
Table 3: Descriptive statistics of the prediction task.
Note: The table reports the sample mean and the sample standard deviation in parentheses for each variable overall (total) and for each treatment (no-mirror: minor). Variable definitions: no-mirror; mirror (participant treatment) : avatar (0 if non-human. 1 if human): assistant (0 if female. 1 if male): time of day (1 if morning. 2 if noon, 3 if afternoon, 4 if evening): VR before (0 if false. 1 if true): gender (0 if female. 1 if male): age (years): international student (0 if German. 1 otherwise); academic level (0 if Bachelor, 1 if Master)
The analysis does not reveal any statistically significant differences for any factor across the two treatments. Therefore, we can assume that our sampling scheme (see Section 3.2.2) has not produced any significant bias between the two treatments.
4.1.2 Primary Results
Two factors of influence on dishonesty were analyzed, namely (1) treatment and (2) avatar choice. Both factors are binary and independent from each other, resulting in four possible treatment-avatar combinations. Individual dishonesty was measured by the balance a participant achieved in the task (see Section 3.2.4). The overall dishonesty in a particular combination can, therefore, be expressed as the sample mean of all balances achieved under this combination. The resulting sample mean balance (in EUR) of each treatment-avatar combination is reported in Table 4 (for a graphical presentation, see Figure Gl).
Abbildung in dieser Leseprobe nicht enthalten
Table 4: Analysis of balances in all treatment-avatar combinations in the experiment in a [3. 3] matrix. Note: The table reports the sample mean of balances (in EUR) and the sample standaid deviation of balances (in EUR) in parentheses, for each combination. The table also includes the total (marginal) balance in both dimensions. Variable definitions: no-mirror; mirror (participant treatment): non-human avatar; human avatar (participant avatar)
Our first hypothesis assumed that a treatment with a niirror in VR would decrease dishonesty, compared to a treatment without a mirror. Consistent with this hypothesis, the marginal mean balance of the no-mirror treatment Mno-m\rror= 6.0476 (SD = 1.6726) is slightly greater than that of the mirror treatment Mm\rror= 5.7708 (SD = 1.5947), although this difference is not significant5 (t = 0.56595, p = 0.5745, Welch two-sample t-test, two-sided). Our second hypothesis assumed that choosing a non-human avatar in VR would increase dishonesty, compared to choosing a human avatar. Consistent with the hypothesis, the marginal mean balance of the participants with non-human avatars Mnon-human = 5.9773 (SD = 1.5312) is slightly greater than that of participants with human avatars Mhuman =5.8261 (SD = 1.7295), although this difference is again not statistically significant (t = 0.3108, p = 0.7575, Welch two-sample t-test, two-sided). Analyzing both hypotheses jointly, even the two most extreme combinations Mno-mirror, non-human = 6.045 (SD = 1.604) and Mmirror, human = 5.6538 (SD = 1.7003) do not differ significantly (t = 0.5798,/> = 0.5681, Welch two-sample t-test, two-sided).
So far, the analysis potentially suffered from omitted variable bias, i.e. a bias in effect size due to correlated factors which are not controlled for. In order to test for the robustness of the effects, a multiple linear regression model of different regressors on the dependent variable balance is estimated, using the sample data. We selected variables to be included based on their theoretical background and the model improvement they provided. The OLS regression employs heteroskedasticity-robust OLS estimators for the variable coefficients6. A summary of the predicted model in four specifications is presented in Table 5. The underlying analysis is given in Appendix G.
Dependent Variable:
Abbildung in dieser Leseprobe nicht enthalten
Table 5: Multiple linear regression model of different regressors and control variables on the dependent variable balance. Note: The predicted model uses heteroskedasticity-robust OLS estimators. The table reports OLS estimator means and OLS estimator standard errors in parentheses. The adjusted R-squared and SER are given as measures of model fit. p-values are indicated as: * p < 0.1, ** p < 0.05, *** p < 0.01. Variable definitions: balance (EUR); mirror treatment (0 if no-minor. 1 if mirror): human avatar (0 if non-human, 1 if human); gender (0 if female. 1 if male): age (years): local student (0 if German. 1 otherwise): realism graphics (no. 9. Appendix E); realism movement (no. 10. Appendix E); realism immersion (no. 17, Appendix E); intuition headset (no. 3. Appendix E); intuition controllers (no. 4. Appendix E): observed (no. 12. Appendix E); pressured (no. 14. Appendix E): moral decision-making (no. 5, Appendix E).
Specification 1 (column 1) presents the core specification with only the two regressors of primary interest included, treatment and avatar (R2 = -0.038, SER = 1.649). Both regressors are modeled by a dummy variable, holding the no-mirror treatment and the non-human avatar as the default. The estimated intercept {bo = 6.1113***) can, therefore, be inteipreted as the balance by paiticipants in the no-mirror treatment with a non-human avatar. As expected, the estimated coefficients on treatment (b1 = -0.2680) and avatar (b2 = -0.1337) are both negative in value, hence are predicted to have a negative effect on balance. To control for covariates that could possibly lead to bias more data was collected as part of a questionnaire (see Appendix E).
Specification 2 (column 2) uses some of the data to control for potential biases from participant demographic factors (R2 = -0.0937, SER = 1.693). Neither of the estimated coefficients on gender (b3 = -0.3516), age (b4 = 0.087), and local student (b5 = 0.3648) status is a significant predictor, and including them in the model does not lead to a significant improvement of the overall model predicted (F = 0.2863, p = 0.835, one-way ANOVA). Hence, these factors are excluded from further specifications.
Specification 3 (column 3) includes a group of variables to control for the presence of a participant (R2 = 0.1765, SER = 1.469). Specifically, we controlled for the influence of realism (b6 = 0.4112**; b7 = -0.4051***; b8 = -0.1742), and intuition (b9 = 0.1321; b10 = 0.5701*) on balance. Including these variables yields a significant improvement of the predicted model (F = 3.1876; p = 0.0172, one-way ANOVA). However, the estimated coefficients on realism should not be interpreted directly, due to high risk of multicollinearity.
Specification 4 (column 4) includes variables of participants’ feelings of being watched to control for any bias from perceived ‘observedness’ on balance (R2 = 0.1923, SER = 1.455). The estimated coefficients on observed (b11 = 0.0169) and pressured (b12 = -0.2446) are not statistically significant. However, they lead to an improvement of the predicted model, though also not statistically significant (F = 1.3621; p = 0.2694, one-way ANOVA).
Specification 5 (column 5) controls for participants’ understanding of the experiment involving moral decision-making (R2 = 0.1689, SER = 1.476), to control for potential ethical blindness (Chugh et al., 2005; Gino et al., 2010). The estimated coefficient (b12 = 0.0138) is of minor impact, and the overall model is not significantly improved (F = 0.0146; p = 0.9044, one-way ANOVA).
Across all specifications investigated, the key regressors treatment and avatar appear robust in their decreasing effect on balance. However, neither of the regressors appears statistically significant and thus does not have a significant effect.
4.1.3 Secondary Results
The insignificance of the two key regressors asks for an extended analysis of the assumptions made by our hypotheses. A necessary assumption employed is that participants at least behave significantly dishonest at all. To test this assumption, we compare the observed sample mean balance of the no-mirror treatment Mno-mirror= 6.0476 (SD = 1.6726) and of the inirror treatment Mmirror = 5.7708 (SD = 1.5947) to the hypothetical expected mean balance under perfect honesty Mhonest = 5 (SD =1.118, assuming binomial distribution (n = 20, p = 0.5)). For both treatments, we find that the observed values differ significantly from perfect honesty (t = 2.5803, p = 0.0152; t = 2.0782,/? = 0.04499, Welch two-sample t-test, two-sided). This is further supported by a graphical analysis of the density functions (see Figure Gl). This implies that participants of both treatments at least behaved dishonestly. We conclude that our assumption is satisfied.
At this point, an alternative measure for dishonest behavior can be tested between the treatments. As we argued earlier, the response time of participants before taking a decision can be used as a proxy to detect dishonest behavior (Suchotzki et al., 2017; Walczyk et al., 2009; Walczyk et al., 2003). We now assume that longer response times imply a longer cognitive self-reflection and thus a higher reluctance towards lying. To test the two central hypotheses under this new assumption, we repeat the earlier analysis approach (see Section 4.1.2), and test for any effect of (1) treatment and (2) avatar choice on response time per round, as a proxy for dishonest behavior. The sample mean response time per round (in sec) of each treatment-avatar combination is reported in Table 6.
Abbildung in dieser Leseprobe nicht enthalten
Table 6: Analysis of response times in all treatment-avatar combinations in the experiment in a [3, 3] matrix. Note: The table reports the sample mean of response times (in sec) and the sample standard deviation of response times (in sec) in parentheses, for each combination. The table also includes the total (marginal) response time in both dimensions. Variable definitions: no-mirror; mirror (participant treatment): non-human avatar; human avatar (participant avatar)
Our first hypothesis claimed that the presence of a mirror in VR would decrease dishonesty. Assuming participants behaved equally dishonest (from Table 4), we hypothesize that the presence of a mirror in VR at least increased the time to lie.
Consistent with this alternative hypothesis, the marginal mean response time of the mirror treatment Mmirror = 3.9003 (SD = 2.1491) is indeed longer than that of the no-mirror treatment Mno-mirror = 2.973 (SD = 0.779), and the difference is significant with p < 0.1 (t = 1.9699, p = 0.0583, Welch two-sample t-test, two-sided).
Our second hypothesis claimed that choosing a non-human avatar in VR would provide a mask and increase dishonesty. Again, assuming participants behaved equally dishonest (from Table 4), we hypothesize that a non-human avatar in VR at least decreased the time to lie. Consistent with this alternative hypothesis, the marginal mean response time of the participants with non-human avatars Mnon-human = 2.9046 (SD = 0.5952) is indeed shorter than that of participants with human avatars Mhuman =4.0065 (SD = 2.2043), and the difference is significant with p < 0.05 (t = 2.3108, p = 0.0293, Welch two-sample t-test, two-sided).
4.2 Discussion
4.2.1 General Findings
The experiment results give rise to a number of possible interpretations of the implications of VR. Consistent with Carver and Scheier (1978), D. M. Buss and Scheier (1976), and Diener and Wallbom (1976), the presence of a mirror in VR also seems to induce self-awareness. Our analysis yields that individuals took significantly longer to lie in the presence of a mirror in VR (Walczyk et al., 2009; Walczyk et al., 2003). However, in our main model, the resulting effect on dishonesty does not appear statistically significant. Therefore, we suggest that the mechanisms of mirror presence in RL also apply in VR, but the effect on dishonesty might be insignificant, compared to other factors.
Consistent with Diener et al. (1976), Miller and Rowold (1979), and Zimbardo (1975), wearing a non-human avatar in VR also seems to provide a mask and thus a form of deindividuation, which reduces self-awareness. Our analysis yields that individuals took significantly less time to lie when wearing a non-human avatar (Walczyk et al., 2009; Walczyk et al., 2003). However, in our main model, this does not have a statistically significant effect on dishonesty. Therefore, we suggest that avatars in VR also apply as a deindividuation cue, but the effect on dishonesty is insignificant.
Besides these key findings, some of the theoretical lying mechanisms presented earlier (see Section 2.3.2; 2.3.3) could be observed among participants. For instance, even though dishonesty could not be detected on an individual level, only two (4.44%) participants lied to the full extent. However, at least 25 (55.56%) participants showed signs of partial lying (critical value at balance ≥ 6, t = 3.3005, p = 0.0014). We suggest that this is consistent with prior empirical results about self-concept maintenance in RL (Fischbacher & Föllmi-Heusi, 2013; Mazar et al., 2008), and shows that individuals also have a preference for self-concept maintenance under moral decision-making in VR.
Furthermore, some participants were found to be influenced by moral balancing (Barkan et al., 2012; Mazar & Zhong, 2010; Nisan, 1991) in their decision-making. After they had taken a relatively long time before deciding to report an incorrect prediction, they were observed deciding relatively fast to report correct predictions for the next rounds in a row. Although this observation is only on a subjective level, this could suggest that individuals use balancing and justification mechanisms they know from RL in a comparable way in VR. This is also in line with the strategic cheating behavior observed in the VR mind game by Mol et al. (2018).
4.2.2 Specific Findings of Self-Awareness
The insignificant primary analysis asks for a more detailed discussion about the true effect of a mirror in VR on self-awareness. We hypothesize two conditions that have to be satisfied with a mirror in VR to induce self-awareness. First, the reflective surface in VR has to be understood as a mirror, to induce mirror presence. Observations from our experiment show that this is likely the case. The majority of participants in the mirror treatment interacted and entertained themselves with their avatar reflection, for instance by dancing, nodding with their head, or waiving with their arms. This is in line with behavior typically observed in front of a mirror (Radell et al., 2014). Therefore, we regard this condition as satisfied.
Second, the individual has to identify himself with his avatar in VR, for the mirror reflection to induce self-awareness. We assumed that a non-human avatar should not induce self-awareness, whereas a human avatar should induce self-awareness (Kilteni et al., 2012; Zhang & Hommel, 2016). Participants had to rate their self-awareness towards their own avatar in two questions (see Appendix E). For both questions, the mean self-awareness score of human- avatars (M1 = 5; M2 = 4.8696) was higher than the same score of non-human avatars (M1 = 4.3182; M2 = 3.5). However, only one (2.22%) human- avatar participant in the mirror treatment reported feeling observed through their avatar reflection in the mirror.
One explanation could be that the human avatars did not look realistic enough (Thaler et al., 2018). Another explanation provided by Jin (2010) claims that individuals rather identify with an idealized avatar of themselves than with a normal avatar. It could, therefore, be critical to the methodology of any VR experiment involving self-awareness to employ realistic avatar models and let participants design their idealized avatars prior to the experiment.
4.2.3 Specific Findings of Dishonesty
The insignificant primary analysis further asks for a discussion of potential influencing factors on dishonesty in VR. Because the alternative analysis yielded a significant mirror presence effect, we hypothesize that the subsequent effect on dishonest behavior was overshadowed by factors inherent to the VR equipment.
First, the intuition in the use of the VR equipment might be a factor of influence. For instance, our questionnaire results show that none of the participants owns any VR equipment, and only 22 (48,89%) have ever experienced VR before. Some participants were observed holding their controllers in the wrong hands but not correcting the mistake. Furthermore, participants who rated the VR equipment less intuitive to use overall achieved lower balances, according to the regression model (see Table 5). We suggest this shows that individuals can be overwhelmed by the unfamiliar usage and experience of VR, such that they do not actively think about the possibility of lying. This is consistent with prosocial behavior due to cognitive depletion (see Rand, Greene, & Nowak, 2012). A general recommendation for VR experimental design could, therefore, be to ensure high intuition and familiarity of participants with the VR equipment, for instance by increasing the time participants spend in a tutorial prior to the experiment.
Second, the tutorial experience itself might also have influenced dishonesty. After the experiment, we noticed that participants repeatedly emphasized they had especially enjoyed the tutorial game. A few participants even indicated they would have participated in the experiment for no financial reward but just for the tutorial game. The tutorial game of our experiment, therefore, possibly overshadowed the financial incentives of the prediction task for some participants. We suggest that highly immersive and ‘playful’ experiences could provide a source of distraction, and make it hard for individuals to adapt to a ‘monotonic’ task directly afterward (Rand et al., 2012). Another recommendation for VR experimental design could, therefore, be to also align the level of immersion and experience of the tutorial and the experimental setting.
An alternative interpretation is given by Mazar and Zhong (2010); (Nisan, 1991). The tutorial game, involving the act of shooting human-like agents, could have negatively primed our participants in a way so that they balanced this with overly honest behavior in the prediction task afterward. Again, this adds to our recommendation of aligning the experience of all experimental stages in VR to prevent any priming effects.
Third, the VR system itself might constitute an inherent factor of influence. Interestingly, 24 (53.33%) participants indicated feeling observed by other factors than the mirror. As one factor, four (8.89%) participants had the feeling of being observed. They explained this by naming the mere fact that they could not see their real surrounding but still sometimes heard sounds outside the room. Therefore, we suggest that wearing a VR headset can induce the feeling of being observed by someone else in the room. Moreover, we suggest that a lack of auditory input provides an inconsistent experience in VR, and, therefore, has a negative effect on immersion (see Bombari et al., 2015). As another factor, eight (17.78%) participants reported the feeling of being observed by the VR system and software. Consequently, we suggest that a VR headset, linked to a computer, can create a feeling of constantly being watched or logged by the software. An important implication for dishonesty experiments in VR could be that individuals might inherently feel observed and act more honest when wearing a VR headset, consistent with the effects of observation on dishonesty in RL (e.g., Fischbacher & Föllmi-Heusi, 2013; Howells, 1938; Mol et al., 2018; Rustagi et al., 2010).
4.2.4 Implications for Managers and SMEs
As we showed earlier, dishonesty is a prevalent problem in the business environment and especially bears high costs for small and medium-sized enterprises (SMEs) (see Weber et al., 2003). As VR is a maturing technology (Gartner, 2017, 2018), businesses are projected to increasingly conduct economic transactions and information exchange in VR (Madary & Metzinger, 2016). This implies novel opportunities for managers of SMEs to employ VR as a technology. We argue that VR offers new measures for SMEs to reduce dishonesty.
First, our discussion shows that VR experiences and the VR system can temporarily prime individuals in a way that reduces dishonesty (see Section 4.2.3). Managers could deliberately exploit this mechanism to reduce dishonesty. For instance, they could initiate a dishonesty-reducing VE for all parties, before entering a negotiation. Especially in situations of short duration, like a call, this temporary mechanism could effectively to decrease dishonesty. Further research is necessary to identify which specific experiences prime against dishonesty.
Second, our discussion shows that self-awareness seems to exist in VR, and the two initially proposed factors (1) mirror presence and (2) avatar choice have an influence on it in VR (see Section 4.2.2). This suggests some best practices in designing VEs that induce self-awareness and, therefore, reduce dishonesty. (1) First, managers could place a large mirror in the virtual conference room behind their own seat. Moreover, electronic contract closing in VR could employ a reflection of one’s avatar above the signature field. (2) Second, managers could only allow for the usage of realistic human avatars in a negotiation, to inhibit masking and anonymity. Both of these conditions induce self-awareness and, therefore, decrease dishonesty. Further research could identify more factors of influence on self-awareness in VR.
Chapter 5: Limitations and Further Research
A number of suggestions have already been provided where applicable. The following part extends this discussion to summarize the methodological limitations of the experiment. Afterward, suggestions for future research are provided.
First, we note that all participants were chosen among business students from the same university, to acquire a large number of participants. Moreover, the fact that this academic institution is a private business school makes the group of participants highly homogenous across age, education, and socio-economic background. Although six exchange business students participated in the experiment, we still found little variation across the demographic traits (see Table 1). Furthermore, many participants showed their analytical capabilities and calculated the expected mean of the prediction task. These ‘extreme’ traits provide limited external validity. We should, therefore, only interpreted the results observed within the context of business school students, and call for further experiments with a broader group of participants.
Second, we did not employ the elements of a double-blind procedure in our experiment. The assistant and the analyst role were executed by two business students of the same university as the participants. Therefore, some of the participants knew the assistant personally or even were friends with them. As Ennis, Vrij, and Chance (2008) and Fischbacher and Föllmi-Heusi (2013) argue, this possibly increased image concerns of lying (Mazar et al., 2008), and led to reduced lying behavior.
Third, our questionnaire did not include any questions to measure psychological characteristics like the Big Five characteristics of our participants. This aspect was neglected due to a lack of time and to focus participants’ concentration on a few high-qualitative questions on the VR experience. However, some of these factors can have significant effects on dishonesty behavior of an individual and should be controlled for (e.g., Conrads et al., 2013). We were not able to control for these effects in our regression analysis, and, therefore, have to be aware of potential omitted variable bias when interpreting the results.
Fourth, our regression model (see Table 5) currently includes three variables which could suffer from multicollinearity (Realism graphics, Realism movement, Realism immersion). We nevertheless reported all three in the model, to account for the impact of each on presence, as suggested by theory (Sanchez-Vives & Slater, 2005). This is of no bias in the light of our exploratory focus. However, we may only interpret their individual coefficients with caution.
Contrary to our initial hypothesis, the experiment results indicate that neither human nor non-human avatars significantly influenced dishonesty in VR. Future research is, therefore, necessary to investigate the potentially more complex conditions for individuals to adapt their dishonest behavior to their own avatar in VR. Such a study could investigate a larger variety of avatars for self-awareness. Furthermore, it could test whether self-awareness is necessarily linked to human appearance, or whether non-human avatars could induce similar effects in VR. One possible starting point concerning avatar design is provided by Jin (2010) and Thaler et al. (2018).
Our discussion also shows that the VR system itself could have made individuals feel constantly observed, which has an impact on their dishonest behavior. Future research might, therefore, find the methodological approach of a field experiment, for instance in a massive multiplayer online role play game (MMORPG) more promising. An experimental task could be provided in the form of a task in the game, which incentivizes dishonest behavior. Conducting the task at home would decrease the feeling of being observed among participants. Furthermore, participants would likely be more used to the VR equipment. Potential experimental setups in a LIVE have been reviewed by Innocenti (2017) and could be adapted to a HIVE.
Chapter 6: Conclusion
The thesis investigated the effects of self-awareness on dishonest behavior in VR. An extensive body of research has already demonstrated these effects in RL. We built on these findings to test their applicability, using an experiment in VR.
We first reviewed the major theory on self-awareness and dishonesty to provide a detailed overview of the underlying mechanisms and influencing factors. The self-concept maintenance mechanism serves as the key link to explain the negative relationship between self-awareness and dishonesty. We proceeded with a review of VR to identify previous findings of psychological effects in the technology. Following this review, we hypothesized that mirror presence and avatar choice would have an effect on dishonesty in VR.
To test these hypotheses, we conducted an experiment with n = 45 participants which employed the paradigm of a mind game in VR. Consistent with our first hypothesis, we found that mirror presence reduces dishonesty in VR. Consistent with our second hypothesis, we also found evidence for a human avatar to reduce dishonesty. Both effects are robust, though lack statistical significance. In an alternative analysis, we employed response time as a proxy for dishonesty. This analysis yielded statistically significant results for both hypotheses.
To conclude, mirror presence and avatar choice indeed seem to influence self-awareness in VR, though the resulting effect on dishonesty is minor, compared to other factors. Having discussed this finding, we identified several such factors. Moreover, we inferred implications and best practices for managers and SMEs from our analysis.
Our research is among the first to investigate self-awareness and its consequences on dishonest behavior in VR. We thus contribute with exploratory findings and empirical evidence to the field of economic behavior in VR. We want to highlight the value and impetus of our experiments for both theory and practice, and call for more influencing factors to be tested in future research.
Appendices
Appendix A: VR Hardware and Software
Appendix B: Participant Emails
Appendix C: Participant Consent Form
Appendix D: Participant Instructions
Appendix E: Participant Questionnaire
Appendix F: Experimenter Roles
Appendix G: Analysis Background
Appendix A: VR Hardware and Software VR Hardware Setup
Abbildung in dieser Leseprobe nicht enthalten
Figure A1: Photos of the VR hardware setup. The photos display the empty VR room and the VR equipment as being used during the experiment. The VR equipment used is a HTC Vive© (HTC Corporation, Taoyuan). The components highlighted are a) the HMD, b) one of two signal stations, c) one of two controllers, and d) the computer running the VR software.
Tutorial Game
[The two original screenshots are not part of this publication.]
Figure A2: Screenshots from the tutorial game. The tutorial game is the minigame Longbow, part of the VR game The Lab© (Valve Corporation, Bellevue, WA). The game is a comic-style tower-defense game (Tan & Soh, 2011) in VR, in which the player needs to defend a fictitious castle against attacking comic figures, using archery. The game requires the player to realistically aim and shoot with a bow, using the controllers to insert an arrow, pull back to tense the bow, and release to shoot the arrow. This delivers a high immersion (Katz et al., 2016) and body ownership (Zhang & Hommel, 2016).
Abbildung in dieser Leseprobe nicht enthalten
Figure A3: Excerpt from the output of the logging algorithm for one particular participant during the prediction task. The log captures the participant id (line 1), date and time (line 2), treatment room (line 3), avatar choice (line 4), and individual round results for each of the 20 rounds (lines 5-24). The final balance is the balance in round 20 (line 24). The total response time is the sum of response time over all rounds (line 5-24).
Appendix B: Participant Emails
Abbildung in dieser Leseprobe nicht enthalten
Figure B1 : First participant invitation email (initial email)
Abbildung in dieser Leseprobe nicht enthalten
Figure B2 : Second participant invitation email (reminder email)
Abbildung in dieser Leseprobe nicht enthalten
Figure B4 : Individualized confirmation email sent to participants after they registered for participation. The name, date and time field were adapted according to each participant. The two pictures were included as attachments.
Abbildung in dieser Leseprobe nicht enthalten
Appendix C: Participant Consent Form
Abbildung in dieser Leseprobe nicht enthalten
RESEARCH INFORMATION
What is the purpose of this research?
You are being asked to participate in a study on Virtual Reality (VR). This research involves a VR Experiment.
Am I eligible to participate?
You have to be older than 18 years and not to have any major mental illnesses, incl. epilepsy. You are not expected to have any prior experience with VR.
Which procedure will the study follow?
You will be asked to read an instructional text about the planned activities. Afterward, an assistant will bring you to the physical experiment room, where the VR Experiment will take place. The assistant will help you put on the VR headset (HTC Vive ©), and adjust the fit to your comfort. Afterward, the VR Experiment will start. You will stay in a VR environment for approx. 15 minutes. During these 15 minutes, you will play a tutorial game and then conduct a prediction task. The prediction task will involve real money, which you can earn as an Research Subject Informed Consent Form additional bonus on top of your guaranteed show-up payment. After the VR Experiment is over, an assistant will demount the headset. You will be brought to a different room to fill in a questionnaire, and receive your overall payout (show-up payment + earned additional bonus).
The study is expected to take 30 minutes in total, and it will be completed in one session. All study procedures take place in the facilities of WHU, Campus Vallendar.
What are the possible risks and discomforts?
In rare cases, VR exposure can cause motion sickness. Motion sickness is a feeling of sickness and discomfort triggered by the perception of movement, while your body is physically at rest. It can also occur in a car, train, plane, or boat. Motion sickness immediately stops as soon as you close your eyes or take off the VR headset.
In very rare cases, VR exposure can cause nausea, dizziness, drowsiness, or light headache for up to several hours.
If you feel uncomfortable at any time during the VR Experiment, please close your eyes and say that you would like to stop the VR Experiment. An assistant will immediately stop the VR Experiment, and help you demount the headset.
What are the possible benefits for me and others?
You personally benefit from this study by experiencing a VR environment.
The study results may help researchers in the field of VR and behavioral research. As a practical implication, prolonged research may help managers to decide on the setup of future workplaces and collaboration setups.
How will you protect the information you collect about me, and how will that information be shared?
All study data relating to the experiment is completely anonymized and randomized. No personal information can be inferred from or linked to your experimental data at any time.
Results of this study may be used in publications and presentations. Your study data will be handled confidentially according to the EU-DSGVO and BDSG. If any results of this study are published or presented, personally identifiable information will not be used. We may also share the data we collect from you for future research studies and/or with other researchers. If we share data collected from you, we will remove any personally identifiable information to anonymize the data.
To minimize the risks to confidentiality, all data recorded is stored encrypted in a secure place. Access to the study records is highly restricted to the principal investigator and research assistants.
What will the compensation be?
Participation in this study will involve no cost for you.
Your overall participation payout consists of two components:
(1) Guaranteed show-up payment of EUR 7.50.
(2) An additional bonus, which is based on the outcome of the prediction task.
What are my rights as a research participant?
Participation in this study is completely voluntary. You do not have to answer any question you do not want to answer.
If at any time and for any reason, you would prefer not to participate in this study, please feel free not to. If at any time you would like to take a break or stop participation, please, tell the assistant. You may withdraw from this study at any time, and you will not be penalized in any way for deciding to stop participation.
If you decide to withdraw from this study, the assistant will ask you if the information already collected from you can be used.
Does the withdrawal from the study affect me as a student/employee?
You may choose not to participate or to stop participating in this research project at any time. This will not affect your class standing, grades, employment, or any other aspects of your relationship with the WHU - Otto Beisheim School of Management.
Whom can I contact if I have questions or concerns about this research study?
If you have questions, you are free to ask them now. If you have any questions about your rights as a participant in this research, you can contact the following office:
Abbildung in dieser Leseprobe nicht enthalten
I have duly read this form and the research study has been explained to me. I have been given the opportunity to ask questions and my questions have been answered. If I have additional questions, I have been told whom to contact. I agree to participate in the research study described above.
Moreover, I duly agree to not disclose any information related to the experiment and its contents to anyone any earlier than Monday, April 15, 2019.
Abbildung in dieser Leseprobe nicht enthalten
Appendix D: Participant Instructions
Treatment 1 (No-Mirror)
Welcome to the VR Experiment!
Thank you very much for your participation. For showing up, you will receive EUR 7.5. Depending on your behavior in this VR Experiment you are able to earn an additional bonus. How you can earn the additional bonus, will be explained in the following instructions. Therefore, please, read the instructions carefully.
The experiment will take place in the VR Physical Experiment room. You will be guided to the room by our assistant. The assistant will introduce you to the equipment (VR Headset and Controllers). Afterward, she will help you to put on the headset and adjust the fit. Make sure that the equipment suits you well and you have a clear view, before we continue. For the time of the experiment, you will stay alone in the room. However, if you feel uncomfortable at any time, please, close your eyes and say that you would like to stop the experiment. Our assistant will immediately stop the experiment and help you demount the headset. The assistant will announce when she is leaving the room.
Tutorial game
First, you will be able to accommodate yourself with the VR environment and usage of the controllers during a short tutorial game (max. 5 minutes).
The prediction task
After the tutorial game, you will have to choose the avatar: either female, or male, or a teddy bear. Pull and hold the trigger button with your index finger to flash the selection pointer. Point at the avatar and confirm your choice by releasing the trigger button. Then you will be teleported into a VR Experiment room with a virtual stage (see Fig. 1).
Abbildung in dieser Leseprobe nicht enthalten
Figure 1. The screenshot above depicts the VR Experiment room.
On the opposite side of the stage, you can see a large display (see Fig. 1). The display currently shows two pictures: a picture of an apple and a picture of a banana next to each other. When the prediction task starts, you will see a countdown on this display, counting down from 7 seconds. After the 7 seconds, the display will show either the apple or the banana. Both fruits are equally likely to appear on the display. Before one of the two fruits appears, your task is to predict whether an apple or a banana will be displayed. You are supposed to quietly estimate in your mind which fruit will appear. The prediction task will be repeated for 20 rounds.
In each round you start with EUR 0.5. If your prediction was correct, you are supposed to click “correct”, and you will keep the bonus of EUR 0.5. If your prediction was incorrect, you are supposed to click “incorrect”, and you will lose the bonus of EUR 0.5. The earnings of all 20 rounds will be summed up, and constitute your total additional bonus. Thus, your overall payout is your show-up payment plus your total additional bonus.
The end of the experiment
After the experiment, the assistant will help you to demount the headset, and you will be guided to a separate physical room. Here, you will be asked to complete a questionnaire, and you will receive your payout.
Abbildung in dieser Leseprobe nicht enthalten
Figure D1 : Participant instruction text for treatment 1 (no-mirror)
Treatment 2 (Mirror)
Welcome to the VR experiment!
Thank you very much for your participation. For showing up, you will receive EUR 7.5. Depending on your behavior in this VR Experiment you are able to earn an additional bonus. How you can earn the additional bonus, will be explained in the following instructions. Therefore, please, read the instructions carefully.
The experiment will take place in the VR Physical Experiment room. You will be guided to the room by our assistant. The assistant will introduce you to the equipment (VR Headset and Controllers). Afterward, she will help you to put on the headset and adjust the fit. Make sure that the equipment suits you well and you have a clear view, before we continue. For the time of the experiment, you will stay alone in the room. However, if you feel uncomfortable at any time, please, close your eyes and say that you would like to stop the experiment. Our assistant will immediately stop the experiment and help you demount the headset. The assistant will announce when she is leaving the room.
Tutorial game
First, you will be able to accommodate yourself with the VR environment and usage of the controllers during a short tutorial game (max. 5 minutes).
The prediction task
After the tutorial game, you will have to choose the avatar: either female, or male, or a teddy bear. Pull and hold the trigger button with your index finger to flash the selection pointer. Point at the avatar and confirm your choice by releasing the trigger button. Then you will be teleported into a VR Experiment room with a virtual stage (see Fig. 1).
Abbildung in dieser Leseprobe nicht enthalten
Figure 1. The screenshot above depicts the VR Experiment room.
On the opposite side of the stage, you can see a large display, and two mirrors reflecting your avatar (here as example the teddy bear; see Fig. 1). The display currently shows two pictures: a picture of an apple and a picture of a banana next to each other. When the prediction task starts, you will see a countdown on this display counting down from 7 seconds. After the 7 seconds, the display will show either the apple or the banana. Both fruits are equally likely to appear on the display. Before one of the two fruits appears, your task is to predict whether an apple or a banana will be displayed. You are supposed to quietly estimate in your mind which fruit will appear. The prediction task will be repeated for 20 rounds.
In each round you start with EUR 0.5. If your prediction was correct, you are supposed to click “correct”, and you will keep the bonus of EUR 0.5. If your prediction was incorrect, you are supposed to click “incorrect”, and you will lose the bonus of EUR 0.5. The earnings of all 20 rounds will be summed up and constitute your total additional bonus. Thus, your overall payout is your show-up payment plus your total additional bonus.
The end of the experiment
After the experiment, the assistant will help you to demount the headset, and you will be guided to a separate physical room. Here, you will be asked to complete a questionnaire, and you will receive your payout.
Figure D2 : Participant instruction text for treatment 2 (mirror)
Abbildung in dieser Leseprobe nicht enthalten
You just experienced a VR environment. Please, reflect on your experiences during the prediction task to fill out the following questionnaire.
Perception of the experiment task
Abbildung in dieser Leseprobe nicht enthalten
Appendix F: Experimenter Roles
Abbildung in dieser Leseprobe nicht enthalten
Appendix G: Analysis Background
Density Functions
Abbildung in dieser Leseprobe nicht enthalten
Figure G1 : Graphical presentation of the density functions of sample balances.
Note: Balance (in EUR) on the x-axis, density on the y-axis. The lines are specific to each treatment-avatar combination, as described by the legend. The density functions are approximated and smoothed to the underlying sample balances recorded during the prediction task (see Table 4). Interpretation: The distributions appear clearly skewed towards higher values of balance, with an abnormal density increase in the right-hand tail of balance. This gives evidence for partial lying (skewed towards higher values) and for full dishonesty in a few instances (heavy right-hand tails).
Diagnostic Plots for the Regression Model
Abbildung in dieser Leseprobe nicht enthalten
Figure G2 : The diagnostic plots were employed for specification 1 of the multiple linear regression model presented in the analysis to test the least-squares assumptions (LSA) and for possible non-linearity. From the plots presented, we argue that all LSA are sufficiently met by our data. Moreover, we do not find any suggestions for a non-linear model term.
Abbildung in dieser Leseprobe nicht enthalten
Residual standard error: 1.693 on 39 degrees of freedom Multiple R-squared: 0.03053, Adjusted R-squared: -0.09376 F-statistic: 0.2456 on 5 and 39 DF, p-value: 0.9395
Figure G4 : Regression Model – Specification 2 (see Table 5). The estimation of the multiple linear regression model was done in the statistical computing language R (R Foundation for Statistical Computing, Vienna).
Abbildung in dieser Leseprobe nicht enthalten
Residual standard error: 1.469 on 37 degrees of freedom Multiple R-squared: 0.3075, Adjusted R-squared: 0.1765 F-statistic: 2.347 on 7 and 37 DF, p-value: 0.04346
Figure G5 : Regression Model – Specification 3 (see Table 5). The estimation of the multiple linear regression model was done in the statistical computing language R (R Foundation for Statistical Computing, Vienna).
Abbildung in dieser Leseprobe nicht enthalten
Residual standard error: 1.455 on 35 degrees of freedom Multiple R-squared: 0.3575, Adjusted R-squared: 0.1923 F-statistic: 2.164 on 9 and 35 DF, p-value: 0.0497
Figure G6 : Regression Model – Specification 4 (see Table 5). The estimation of the multiple linear regression model was done in the statistical computing language R (R Foundation for Statistical Computing, Vienna).
Abbildung in dieser Leseprobe nicht enthalten
Residual standard error: 1.476 on 34 degrees of freedom Multiple R-squared: 0.3578, Adjusted R-squared: 0.1689 F-statistic: 1.894 on 10 and 34 DF, p-value: 0.08092
Figure G7 : Regression Model – Specification 5 (see Table 5). The estimation of the multiple linear regression model was done in the statistical computing language R (R Foundation for Statistical Computing, Vienna).
References
Abeler, J., Becker, A., & Falk, A. (2014). Representative evidence on lying costs. Journal of public economics, 113 (C), 96–104.
Achtziger, A., Alós-Ferrer, C., & Wagner, A. K. (2015). Money, depletion, and prosociality in the dictator game. Journal of Neuroscience, Psychology, and Economics, 8 (1), 1– 14.
Allingham, M. G., & Sandmo, A. (1972). Income tax evasion: A theoretical analysis. Journal of public economics, 1 (3–4), 323–338.
Andreoni, J. (1990). Impure altruism and donations to public goods: A theory of warm-glow giving. The Economic Journal, 100 (401), 464–477. Aronson, E. (1969). The theory of cognitive dissonance: A current perspective. In Advances in Experimental Social Psychology (Vol. 4, pp. 1–34). Ayal, S., & Gino, F. (2011). Honest rationales for dishonest behavior. In M. Mikulincer & P. R. Shaver (Eds.), Herzliya series on personality and social psychology. The social psychology of morality: Exploring the causes of good and evil (pp. 149–166). Washington, DC: American Psychological Association.
Aymerich-Franch, L., Kizilcec, R. F., & Bailenson, J. N. (2014). The relationship between virtual self similarity and social anxiety. Frontiers in Human Neuroscience, 8, 944– 954.
Azar, O. H., Yosef, S., & Bar-Eli, M. (2013). Do customers return excessive change in a restaurant?: A field experiment on dishonesty. Journal of Economic Behavior & Organization, 93 (C), 219–226. Bailenson, J. N., & Blascovich, J. (2004). Avatars. In Encyclopedia of Human-Computer Interaction. Citeseer: Berkshire Publishing Group. Bailenson, J. N., Blascovich, J., & Guadagno, R. E. (2008). Self-representations in immersive virtual environments. Journal of Applied Social Psychology, 38 (11), 2673–2690. Barkan, R., Ayal, S., Gino, F., & Ariely, D. (2012). The pot calling the kettle black: distancing response to ethical dissonance. Journal of Experimental Psychology: General, 141 (4), 757–773. Barnes, S. J. (2016). Understanding virtual reality in marketing: Nature, implications and potential. Available at SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract, 1–50. Batson, C. D., Kobrynowicz, D., Dinnerstein, J. L., Kampf, H. C., & Wilson, A. D. (1997). In a very different voice: Unmasking moral hypocrisy. Journal of Personality and Social Psychology, 72 (6), 1335–1348.
Batson, C. D., Thompson, E. R., Seuferling, G., Whitney, H., & Strongman, J. A. (1999). Moral hypocrisy: appearing moral to oneself without being so. Journal of Personality and Social Psychology, 77 (3), 525–537. Baumeister, R. F., & Newman, L. S. (1994). Self-regulation of cognitive inference and decision processes. Personality and Social Psychology Bulletin, 20 (1), 3–19. Baumeister, R. F., Vohs, K. D., & Tice, D. M. (2007). The strength model of self-control.
Current Directions in Psychological Science, 16 (6), 351–355. Becker, G. S. (1968). Crime and punishment: An economic approach. Journal of Political Economy, 76 (2), 169–217. Becker, G. S. (1976). Altruism, egoism, and genetic fitness: Economics and sociobiology. Journal of economic Literature, 14 (3), 817–826. Blauert, J., Lehnert, H., Sahrhage, J., & Strauss, H. (2000). An interactive virtual-environment generator for psychoacoustic research. I: Architecture and implementation. Acta Acustica united with Acustica, 86 (1), 94–102. Bombari, D., Schmid Mast, M., Canadas, E., & Bachmann, M. (2015). Studying social interactions through immersive virtual environment technology: virtues, pitfalls, and future challenges. Frontiers in psychology, 6, 869. Brockner, J., Hjelle, L., & Plant, R. W. (1985). Self-focused attention, self-esteem, and the experience of state depression. Journal of Personality, 53 (3), 425–434. Bryan, C. J., Adams, G. S., & Monin, B. (2013). When cheating would make you a cheater: Implicating the self prevents unethical behavior. Journal of Experimental Psychology: General, 142 (4), 1001–1005. Burris, C. T., & Lai, E. (2012). Through with the looking glass: Escape responses to implicit mirror exposure. Consciousness and cognition, 21 (1), 464–470. Buss, A. H. (1980). Self consciousness and social anxiety. San Francisco: W. H. Freeman. Buss, D. M., & Scheier, M. F. (1976). Self-consciousness, self-awareness, and self-attribution. Journal of Research in Personality, 10 (4), 463–468. Campbell, E. Q. (1964). The internalization of moral norms. Sociometry, 27 (4), 391–412. Cappelen, A. W., Halvorsen, T., Sørensen, E. Ø., & Tungodden, B. (2017). Face-saving or fair-minded: What motivates moral behavior? Journal of the European Economic Association, 15 (3), 540–557. Carver, C. S. (1974). Facilitation of physical aggression through objective self-awareness. Journal of Experimental Social Psychology, 10, 365–370.
Carver, C. S. (1975). Physical aggression as a function of objective self-awareness and attitudes toward punishment. Journal of Experimental Social Psychology, 11 (6), 510– 519. Carver, C. S. (1979). A cybernetic model of self-attention processes. Journal of Personality and Social Psychology, 37 (8), 1251–1281. Carver, C. S., & Scheier, M. F. (1978). Self-focusing effects of dispositional self-consciousness, mirror presence, and audience presence. Journal of Personality and Social Psychology, 36 (3), 324–332. Carver, C. S., & Scheier, M. F. (1981). Attention and self-regulation: A control theory approach to human behavior: New York: Springer. Charness, G., & Dufwenberg, M. (2006). Promises and partnership. Econometrica, 74 (6), 1579–1601. Childs, J. (2012). Gender differences in lying. Economics Letters, 114 (2), 147–149. Chugh, D., Bazerman, M. H., & Banaji, M. R. (2005). Bounded ethicality as a psychological barrier to recognizing conflicts of interest. In D. A. Moore, D. M. Cain, G. Loewenstein, & M. H. Bazerman (Eds.), Conflicts of interest challenges and solutions in business law medicine and public policy (pp. 74–95). New York, NY: Cambridge University Press. Cialdini, R. B., Kallgren, C. A., & Reno, R. R. (1991). A focus theory of normative conduct: A theoretical refinement and reevaluation of the role of norms in human behavior. Advances in Experimental Social Psychology, 24, 201–234. Cojoc, D., & Stoian, A. (2014). Dishonesty and charitable behavior. Experimental Economics, 17 (4), 717–732. Conrads, J., Irlenbusch, B., Rilke, R. M., Schielke, A., & Walkowitz, G. (2014). Honesty in tournaments. Economics Letters, 123 (1), 90–93. Conrads, J., Irlenbusch, B., Rilke, R. M., & Walkowitz, G. (2013). Lying and team incentives. Journal of Economic Psychology, 34, 1–7. Dana, E. R., Lalwani, N., & Duval, T. S. (1997). Objective self-awareness and focus of attention following awareness of self-standard discrepancies: Changing self or changing standards of correctness. Journal of Social and Clinical Psychology, 16 (1), 359–380. Davis, D., & Brock, T. C. (1975). Use of first person pronouns as a function of increased objective self-awareness and performance feedback. Journal of Experimental Social Psychology, 11 (4), 381–388.
De Paulo, B. M., & Kashy, D. A. (1998). Everyday lies in close and casual relationships. Journal of Personality and Social Psychology, 74 (1), 63–79. Dibbets, P., & Schulte-Ostermann, M. A. (2015). Virtual reality, real emotions: a novel analogue for the assessment of risk factors of post-traumatic stress disorder. Frontiers in psychology, 6 (May), 1–8. Diener, E. (1976). Effects of prior destructive behavior, anonymity, and group presence on deindividuation and aggression. Journal of Personality and Social Psychology, 33 (5), 497–507. Diener, E. (1977). Deindividuation: Causes and consequences. Social Behavior & Personality: an international journal, 5 (1), 143–156. Diener, E. (1979). Deindividuation, self-awareness, and disinhibition. Journal of Personality and Social Psychology, 37, 1160–1171. Diener, E., Fraser, S., Beaman, A., & Kelem, R. (1976). Effects of deindividuation variables on stealing among halloween trick-or-treaters. Journal ol Personality and Social Psychology, 33 (2), 178–183. Diener, E., Lusk, R., DeFour, D., & Flax, R. (1980). Deindividuation: Effects of group size, density, number of observers, and group member similarity on self-consciousness and disinhibited behavior. Journal of Personality and Social Psychology, 39 (3), 449–459. Diener, E., & Wallbom, M. (1976). Effects of self-awareness on antinormative behavior. Journal of Research in Personality, 10 (1), 107–111. Dittmann, A. T., & Llewellyn, L. G. (1969). Body movement and speech rhythm in social conversation. Journal of Personality and Social Psychology, 11 (2), 98–108. Dufwenberg, M., & Gneezy, U. (2000). Measuring beliefs in an experimental lost wallet game. Games and economic Behavior, 30 (2), 163–182. Duval, S., & Wicklund, R. A. (1972). A theory of objective self awareness. Oxford, England: Academic Press. Edelman, G. M. (2003). Naturalizing consciousness: A theoretical framework. Proceedings of the National Academy of Sciences, 100 (9), 5520–5524. Ellison, P. A., Govern, J. M., Petri, H. L., & Figler, M. H. (1995). Anonymity and aggressive driving behavior: A field study. Journal of Social Behavior and Personality, 10 (1), 265–272. Ennis, E., Vrij, A., & Chance, C. (2008). Individual differences and lying in everyday life. Journal of Social and Personal Relationships, 25 (1), 105–118. Erat, S., & Gneezy, U. (2012). White lies. Management Science, 58 (4), 723–733. Erikson, E. H. (1968). Identity: Youth and crisis. New York: Norton.
Ernst & Young. (2018). 15th Global Fraud Survey. Retrieved from https://fraudsurveys.ey.com/media/1627/global_fraud_survey_2018.pdf on May 6, 2019 Exner, J. E. (1973). The self focus sentence completion: A study of egocentricity. Journal of Personality Assessment, 37 (5), 437–455. Falk, A. (2017). Facing Yourself-A Note on Self-image. CESifo Working Paper Series No. 6428. Available at SSRN: https://ssrn.com/abstract=2965964, 1–18. Fenigstein, A., Scheier, M. F., & Buss, A. H. (1975). Public and private self-consciousness: Assessment and theory. Journal of Consulting and Clinical Psychology, 43 (4), 522– 527. Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford University Press. Festinger, L., Pepitone, A., & Newcomb, T. (1952). Some consequences of deindividuation in a group. The Journal of Abnormal and Social Psychology, 47 (2), 382–389. Fischbacher, U., & Föllmi-Heusi, F. (2013). Lies in disguise – an experimental study on cheating. Journal of the European Economic Association, 11 (3), 525–547. Fox, J., & Bailenson, J. N. (2009). Virtual self-modeling: The effects of vicarious reinforcement and identification on exercise behaviors. Media Psychology, 12 (1), 1– 25. Freeman, I. J., Salmon, J. L., & Coburn, J. Q. (2016). CAD Integration in Virtual Reality Design Reviews for Improved Engineering Model Interaction. Paper presented at the ASME 2016 International Mechanical Engineering Congress and Exposition. Freud, S. (1927). The ego and the id. London: Hogarth. Frey, B. S., & Meier, S. (2004). Pro-social behavior in a natural setting. Journal of Economic Behavior & Organization, 54 (1), 65–88. Friesen, L., & Gangadharan, L. (2013). Designing self-reporting regimes to encourage truth telling: An experimental study. Journal of Economic Behavior & Organization, 94 (C), 90–102. Froming, W. J., Walker, G. R., & Lopyan, K. J. (1982). Public and private self-awareness: When personal attitudes conflict with societal expectations. Journal of Experimental Social Psychology, 18 (5), 476–487. Gallup, A. C., Vasilyev, D., Anderson, N., & Kingstone, A. (2019). Contagious yawning in virtual reality is affected by actual, but not simulated, social presence. Nature Scientific Reports, 9, 1–10.
Gartner. (2017). Gartner Hype Cycle for Emerging Technologies, 2017. Retrieved from https://www.gartner.com/smarterwithgartner/top-trends-in-the-gartner-hype-cycle-for-emerging-technologies-2017/ on May 6, 2019 Gartner. (2018). Gartner Hype Cycle for Emerging Technologies, 2018. Retrieved from https://www.gartner.com/smarterwithgartner/5-trends-emerge-in-gartner-hype-cycle-for-emerging-technologies-2018/ on May 6, 2019 Geller, V., & Shaver, P. (1976). Cognitive consequences of self-awareness. Journal of Experimental Social Psychology, 12 (1), 99–108. Gergen, K. (1971). The concept of self. New York: Holt, Rinehart & Winston. Gibbons, F. X. (1990). Self-attention and behavior: A review and theoretical update. Advances in Experimental Social Psychology, 23, 249–303. Gibbons, F. X., & Wicklund, R. A. (1976). Selective exposure to self. Journal of Research in Personality, 10 (1), 98–106. Gibson, R., Tanner, C., & Wagner, A. F. (2013). Preferences for truthfulness: Heterogeneity among and within individuals. American Economic Review, 103 (1), 532–548. Gino, F., Ayal, S., & Ariely, D. (2009). Contagion and differentiation in unethical behavior: The effect of one bad apple on the barrel. Psychological Science, 20 (3), 393–398. Gino, F., & Galinsky, A. D. (2012). Vicarious dishonesty: When psychological closeness creates distance from one’s moral compass. Organizational Behavior and Human Decision Processes, 119 (1), 15–26. Gino, F., & Mogilner, C. (2014). Time, money, and morality. Psychological Science, 25 (2), 414–421. Gino, F., Norton, M. I., & Ariely, D. (2010). The counterfeit self: The deceptive costs of faking it. Psychological Science, 21 (5), 712–720. Gino, F., Schweitzer, M. E., Mead, N. L., & Ariely, D. (2011). Unable to resist temptation:
How self-control depletion promotes unethical behavior. Organizational Behavior and Human Decision Processes, 115 (2), 191–203. Gneezy, U. (2005). Deception: The role of consequences. American Economic Review, 95 (1), 384–394. Goldsen, R. K., Rosenberg, M., Williams, R. M. L., & Suchman, E. A. (1960). What college students think. New York, NJ: Van Nostrand. Graham, C., Litan, R. E., & Sukhtankar, S. (2002). The bigger they are, the harder they fall: an estimate of the costs of the crisis in corporate governance. Washington, DC: Brookings Institution. Greene, J. D., & Paxton, J. M. (2009). Patterns of neural activity associated with honest and dishonest moral decisions. Proceedings of the National Academy of Sciences, 106 (30), 12506–12511. Grolleau, G., Kocher, M. G., & Sutan, A. (2016). Cheating and loss aversion: Do people cheat more to avoid a loss? Management Science, 62 (12), 3428–3438. Haan, M., & Kooreman, P. (2002). Free riding and the provision of candy bars. Journal of public economics, 83 (2), 277–291. Harris, S., Mussen, P., & Rutherford, E. (1976). Maturity of moral judgment. The Journal of genetic psychology, 128 (1), 123–135. Hechter, M. (1990). The attainment of solidarity in intentional communities. Rationality and society, 2 (2), 142–155. Heider, F. (1958). The psychology of interpersonal relations. New York: Wiley. Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, H., & McElreath, R. (2001).
In search of homo economicus: behavioral experiments in 15 small-scale societies. American Economic Review, 91 (2), 73–78. Herman, T. (2005, May 30). Study suggests tax cheating is on the rise; most detailed survey in 15 years finds $250 billion-plus gap; ramping up audits on wealthy. The Wall Street Journal, D1. Heugens, P., & Lander, M. W. (2009). Structure! Agency! (And other quarrels): A meta-analysis of institutional theories of organization. The Academy of Management Journal, 52 (1), 61–85. Houser, D., Vetter, S., & Winter, J. (2012). Fairness and cheating. European Economic Review, 56 (8), 1645–1655. Howells, T. (1938). Factors influencing honesty. The Journal of social psychology, 9 (1), 97– 102. Hull, J. G., & Levy, A. S. (1979). The organizational functions of the self: An alternative to the Duval and Wicklund Model of self-awareness. Journal of Personality and Social Psychology, 37 (5), 756–768. Innes, J. M., & Young, R. F. (1975). The effect of presence of an audience, evaluation apprehension and objective self-awareness on learning. Journal of Experimental Social Psychology, 11 (1), 35–42. Innocenti, A. (2017). Virtual reality experiments in economics. Journal of Behavioral and Experimental Economics, 69 (C), 71–77. Insko, C. A., Worchel, S., Songer, E., & Arnold, S. E. (1973). Effort, objective self-awareness, choice, and dissonance. Journal of Personality and Social Psychology, 28 (2), 262–269. Jacobsen, C., Fosgaard, T. R., & Pascual-Ezama, D. (2018). Why do we lie? A practical guide to the dishonesty literature. Journal of Economic Surveys, 32 (2), 357–387. Jacobsen, C., & Piovesan, M. (2016). Tax me if you can: An artifactual field experiment on dishonesty. Journal of Economic Behavior & Organization, 124 (C), 7–14. James, W. (1890). The principles of psychology. New York: Henry Holt. Jiang, T. (2013). Cheating in mind games: The subtlety of rules matters. Journal of Economic Behavior & Organization, 93 (C), 328–336. Jin, S.-A. A. (2010). 'I feel more connected to the physically ideal mini me than the mirror-image mini me': Theoretical implications of the „malleable self“ for speculations on the effects of avatar creation on avatar-self connection in Wii. CyberPsychology, Behavior & Social Networking, 13 (5), 567–570. Jorjafki, E. M., Sagarin, B. J., & Butail, S. (2018). Drawing power of virtual crowds. Journal of Royal Society Interface, 15, 1–10. Kahneman, D., & Tversky, A. (1979). Prospect Theory: An Analysis of Decision under Risk. Econometrica, 47 (2), 263–291. Katz, K., Barnett, J., Sawyer, D., Kohli, T., Holden, D., Thuriot, P., & Charlesworth, M. (2016). Developing The Lab. Panel presented at the 2016 Steam Dev Days, Seattle, WA. Kilteni, K., Groten, R., & Slater, M. (2012). The sense of embodiment in virtual reality. Presence: Teleoperators and Virtual Environments, 21 (4), 373–387. Leary, M. R., & Tangney, J. P. (2012). Handbook of self and identity. New York: Guilford Press. Levine, E. E., & Schweitzer, M. E. (2014). Are liars ethical? On the tension between benevolence and honesty. Journal of Experimental Social Psychology, 53, 107–117. Levine, E. E., & Schweitzer, M. E. (2015). Prosocial lies: When deception breeds trust. Organizational Behavior and Human Decision Processes, 126, 88–106. Lewicki, R. J. (1984). Lying and Deception: A Behavioral Model. In M. H. B. a. R. J. Lewicki (Ed.), Negotiation in Organizations (pp. 68–90). Beverly Hills, CA: Sage Publications. Lewis, A., Bardis, A., Flint, C., Mason, C., Smith, N., Tickle, C., & Zinser, J. (2012). Drawing the line somewhere: An experimental study of moral compromise. Journal of Economic Psychology, 33 (4), 718–725. Liebling, B. A., & Shaver, P. (1973). Evaluation, self-awareness, and task performance. Journal of Experimental Social Psychology, 9 (4), 297–306. Lundquist, T., Ellingsen, T., Gribbe, E., & Johannesson, M. (2009). The aversion to lying. Journal of Economic Behavior & Organization, 70 (1–2), 81–92. Madary, M., & Metzinger, T. K. (2016). Real virtuality: a code of ethical conduct. recommendations for good scientific practice and the consumers of vr-technology. Frontiers in Robotics and AI, 3, 3. Mazar, N., Amir, O., & Ariely, D. (2008). The dishonesty of honest people: A theory of self-concept maintenance. Journal of Marketing Research, 45 (6), 633–644. Mazar, N., & Ariely, D. (2006). Dishonesty in everyday life and its policy implications.
Journal of public policy & Marketing, 25 (1), 117–126. Mazar, N., & Zhong, C.-B. (2010). Do green products make us better people? Psychological Science, 21 (4), 494–498. McCall, C., & Singer, T. (2015). Facing off with unfair others: introducing proxemic imaging as an implicit measure of approach and avoidance during social interaction. PLoS ONE, 10 (2), e0117532. Mead, G. H. (1934). Mind, self, and society. Chicago: University of Chicago Press. Mead, N. L., Baumeister, R. F., Gino, F., Schweitzer, M. E., & Ariely, D. (2009). Too tired to tell the truth: Self-control resource depletion and dishonesty. Journal of Experimental Social Psychology, 45 (3), 594–597. Meißner, M., Pfeiffer, J., Pfeiffer, T., & Oppewal, H. (in press). Combining virtual reality and mobile eye tracking to provide a naturalistic experimental environment for shopper research. Journal of Business Research, 1–14. Milgram, S. (1963). Behavioral study of obedience. The Journal of Abnormal and Social Psychology, 67 (4), 371-378. Miller, F. G., & Rowold, K. L. (1979). Halloween masks and deindividuation. Psychological Reports, 44 (2), 422–422. Mol, J. M. (2019). Goggles in the lab: Economic experiments in immersive virtual environments. Journal of Behavioral and Experimental Economics, 79, 155–164. Mol, J. M., van der Heijden, E., & Potters, J. J. J. (2018). (Not) Alone in the World: Cheating in the Presence of a Virtual Observer. Available at SSRN: https://ssrn.com/abstract=3267125. Muraven, M., Tice, D. M., & Baumeister, R. F. (1998). Self-control as a limited resource: regulatory depletion patterns. Journal of Personality and Social Psychology, 74 (3), 774–789. Nisan, M. (1991). The moral balance model: Theory and research extending our understanding of moral choice and deviation. In W. M. Kurtines & J. L. Gewirtz (Eds.), Handbook of moral behavior and development (Vol. 3, pp. 213–249).
Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Ornstein, R. E. (1972). The psychology of consciousness. San Francisco: Freeman. Ornstein, R. E. (1973). The nature of human consciousness. San Francisco: Freeman. Ortmann, A., & Hertwig, R. (2002). The costs of deception: Evidence from psychology. Experimental Economics, 5 (2), 111–131. Oxford Dictionaries. (2019). Virtual Reality. Retrieved from https://www.oxforddictionaries.com/definition/english/virtual-reality on May 6, 2019 Parsons, T. D. (2015). Virtual reality for enhanced ecological validity and experimental control in the clinical, affective and social neurosciences. Frontiers in Human Neuroscience, 9, 660. Peck, T. C., Seinfeld, S., Aglioti, S. M., & Slater, M. (2013). Putting yourself in the skin of a black avatar reduces implicit racial bias. Consciousness and cognition, 22 (3), 779– 787. Persky, J. (1995). The ethology of homo economicus. Journal of Economic Perspectives, 9 (2), 221–231. Piazza, J., Bering, J. M., & Ingram, G. (2011). “Princess Alice is watching you”: Children’s belief in an invisible person inhibits cheating. Journal of experimental child psychology, 109 (3), 311–320. Plant, R. W., & Ryan, R. M. (1985). Intrinsic motivation and the effects of self- consciousness, self-awareness, and ego-involvement: An investigation of internally controlling styles. Journal of Personality, 53 (3), 435–449. Ploner, M., & Regner, T. (2013). Self-image and moral balancing: An experimental analysis. Journal of Economic Behavior & Organization, 93 (C), 374–383. Postmes, T., & Spears, R. (1998). Deindividuation and antinormative behavior: A meta-analysis. Psychological Bulletin, 123 (3), 238–259. Pruckner, G. J., & Sausgruber, R. (2013). Honesty on the streets: A field study on newspaper purchasing. Journal of the European Economic Association, 11 (3), 661–679. Radell, S. A., Keneman, M. L., Adame, D. D., & Cole, S. P. (2014). My body and its reflection: a case study of eight dance students and the mirror in the ballet classroom. Research in Dance Education, 15 (2), 161–178. Rand, D. G., Greene, J. D., & Nowak, M. A. (2012). Spontaneous giving and calculated greed. Nature, 489 (7416), 427–430. Rehm, J., Steinleitner, M., & Lilli, W. (1987). Wearing uniforms and aggression: A field experiment. European Journal of Social Psychology, 17 (3), 357–360. Rixom, J., & Mishra, H. (2014). Ethical ends: Effect of abstract mindsets in ethical decisions for the greater social good. Organizational Behavior and Human Decision Processes, 124 (2), 110–121. Rosenbaum, S. M., Billinger, S., & Stieglitz, N. (2014). Let’s be honest: A review of experimental evidence of honesty and truth-telling. Journal of Economic Psychology, 45 (C), 181–196. Ruffle, B. J., & Tobol, Y. (2014). Honest on Mondays: Honesty and the temporal separation between decisions and payoffs. European Economic Review, 65 (C), 126–135. Rustagi, D., Engel, S., & Kosfeld, M. (2010). Conditional cooperation and costly monitoring explain success in forest commons management. Science, 330 (6006), 961–965. Sanchez-Vives, M. V., & Slater, M. (2005). From presence to consciousness through virtual reality. Nature Reviews Neuroscience, 6, 332–339. Scheier, M. F. (1976). Self-awareness, self-consciousness, and angry aggression. Journal of Personality, 44 (4), 627–644. Scheier, M. F., & Carver, C. S. (1977). Self-focused attention and the experience of emotion: attraction, repulsion, elation, and depression. Journal of Personality and Social Psychology, 35 (9), 625–636. Scheier, M. F., Fenigstein, A., & Buss, A. H. (1974). Self-awareness and physical aggression. Journal of Experimental Social Psychology, 10, 264–273. Schweitzer, M. E., Ordóñez, L., & Douma, B. (2004). Goal setting as a motivator of unethical behavior. Academy of Management Journal, 47 (3), 422–432. Seth, A. K., He, B. J., & Hohwy, J. (2015). Editorial. Neuroscience of Consciousness, 2015 (1), niv001. Shalvi, S., Dana, J., Handgraaf, M. J., & De Dreu, C. K. (2011). Justified ethicality: Observing desired counterfactuals modifies ethical perceptions and behavior. Organizational Behavior and Human Decision Processes, 115 (2), 181–190. Shalvi, S., Eldar, O., & Bereby-Meyer, Y. (2012). Honesty requires time (and lack of justifications). Psychological Science, 23 (10), 1264–1270. Shalvi, S., Gino, F., Barkan, R., & Ayal, S. (2015). Self-serving justifications: Doing wrong and feeling moral. Current Directions in Psychological Science, 24 (2), 125–130. Shalvi, S., & Leiser, D. (2013). Moral firmness. Journal of Economic Behavior & Organization, 93, 400–407. Shu, L. L., & Gino, F. (2012). Sweeping dishonesty under the rug: how unethical actions lead to forgetting of moral rules. Journal of Personality and Social Psychology, 102 (6), 1164–1177. Shu, L. L., Gino, F., & Bazerman, M. H. (2011). Dishonest deed, clear conscience: When cheating leads to moral disengagement and motivated forgetting. Personality and Social Psychology Bulletin, 37 (3), 330–349. Shu, L. L., Mazar, N., Gino, F., Ariely, D., & Bazerman, M. H. (2012). Signing at the beginning makes ethics salient and decreases dishonest self-reports in comparison to signing at the end. Proceedings of the National Academy of Sciences, 109 (38), 15197– 15200. Silke, A. (2003). Deindividuation, anonymity, and violence: Findings from Northern Ireland. The Journal of social psychology, 143 (4), 493–499. Silvia, P. J., & Duval, T. S. (2001). Objective self-awareness theory: Recent progress and enduring problems. Personality and Social Psychology Review, 5 (3), 230–241. Slater, M. (2009). Place illusion and plausibility can lead to realistic behaviour in immersive virtual environments. Philosophical Transactions of the Royal Society of Biology, 364, 3549–3557. Slater, M., Antley, A., Davison, A., Swapp, D., Guger, C., Barker, C., . . . Sanchez-Vives, M. V. (2006). A virtual reprise of the Stanley Milgram obedience experiments. PLoS ONE, 1 (1), e39. Slater, M., & Sanchez-Vives, M. V. (2016). Enhancing our lives with immersive virtual reality. Frontiers in Robotics and AI, 3, 74. Somanathan, E., & Rubin, P. H. (2004). The evolution of honesty. Journal of Economic Behavior & Organization, 54 (1), 1–17. Sperry, R. W. (1969). A modified concept of consciousness. Psychological Review, 76, 532– 536. Steinberg, J., McDonald, P., & O'Neal, E. (1977). Petty theft in a naturalistic setting: The effects of bystander presence. The Journal of social psychology, 101 (2), 219–221. Steinel, W., & De Dreu, C. K. W. (2004). Social motives and strategic misrepresentation in social decision making. Journal of Personality and Social Psychology, 86 (3), 419– 434. Suchotzki, K., Verschuere, B., Van Bockstaele, B., Ben-Shakhar, G., & Crombez, G. (2017). Lying takes time: A meta-analysis on reaction time measures of deception. Psychological Bulletin, 143 (4), 428–453. Swift, W. B., & Hedrick, J. (1917). Side tracking of stuttering by'Starters'. Journal of Applied Psychology, 1 (1), 84–88. Tan, C. T., & Soh, D. (2011). Augmented reality games: A review. Paper presented at the Proceedings of The Asian Simulation and AI in Games Conference, GAMEON-ASIA, EUROSIS. Thaler, A., Piryankova, I., Stefanucci, J. K., Pujades, S., De la Rosa, S., Streuber, S., . . . Mohler, B. J. (2018). Visual perception and evaluation of photo-realistic self-avatars from 3D body scans in males and females. Frontiers in ICT, 5, 1–18. Tulkens, H., & Jacquemin, A. (1971). The cost of delinquency: a problem of optimal allocation of private and public expenditure. (CORE Discussion Paper No. 7133). Tyler, J. M., Feldman, R. S., & Reichert, A. (2006). The price of deceptive behavior: Disliking and lying to people who lie to us. Journal of Experimental Social Psychology, 42 (1), 69–77. Utikal, V., & Fischbacher, U. (2013). Disadvantageous lies in individual decisions. Journal of Economic Behavior & Organization, 85 (C), 108–111. Vallacher, R. R., & Solodky, M. (1979). Objective self-awareness, standards of evaluation, and moral behavior. Journal of Experimental Social Psychology, 15 (3), 254–262. Vincent, L. C., Emich, K. J., & Goncalo, J. A. (2013). Stretching the moral gray zone: Positive affect, moral disengagement, and dishonesty. Psychological Science, 24 (4), 595–599. Walczyk, J. J., Mahoney, K. T., Doverspike, D., & Griffith-Ross, D. A. (2009). Cognitive lie detection: Response time and consistency of answers as cues to deception. Journal of Business and Psychology, 24 (1), 33–49. Walczyk, J. J., Roper, K. S., Seemann, E., & Humphrey, A. M. (2003). Cognitive mechanisms underlying lying to questions: Response time as a cue to deception. Applied Cognitive Psychology: The Official Journal of the Society for Applied Research in Memory and Cognition, 17 (7), 755–774. Ward, D. A., & Beck, W. L. (1990). Gender and dishonesty. The Journal of social psychology, 130 (3), 333–339. Weber, J., Kurke, L. B., & Pentico, D. W. (2003). Why do employees steal? Assessing differences in ethical and unethical employee behavior using ethical work climates. Business & Society, 42 (3), 359–380. Weibull, J. W., & Villa, E. (2005). Crime, punishment and social norms. (No. 610). SSE/EFI
Working Paper Series in Economics and Finance.
Westford, K. L., Diener, E., & Diener, C. (1973). Deindividuating effects of group presence and arousal on stealing by Halloween trick-or-treaters. Paper presented at the Proceedings of the Annual Convention of the American Psychological Association.
Wicklund, R. A. (1975). Objective self-awareness. Advances in Experimental Social Psychology, 8, 233–275.
Wicklund, R. A. (1979). The influence of self-awareness on human behavior. American Scientist, 67 (2), 187–193.
Wiltermuth, S. S., Newman, D. T., & Raj, M. (2015). The consequences of dishonesty. Current Opinion in Psychology, 6, 20–24.
Yee, N., & Bailenson, J. N. (2007). The Proteus effect: The effect of transformed self-representation on behavior. Human communication research, 33 (3), 271–290.
Zajonc, R. B. (1965). Social facilitation. Science, 149 (3681), 269–274.
Zajonc, R. B., & Sales, S. M. (1966). Social facilitation of dominant and subordinate responses. Journal of Experimental Social Psychology, 2 (2), 160–168.
Zanna, M. P., & Aziza, C. (1976). On the interaction of repression-sensitization and attention in resolving cognitive dissonance. Journal of Personality, 44 (4), 577–593.
Zhang, J., & Hommel, B. (2016). Body ownership and response to threat. Psychological Research, 80 (6), 1020–1029.
Zimbardo, P. G. (1969). The human choice: Individuation, reason, and order versus deindividuation, impulse, and chaos. In W. J. A. D. Levine (Ed.), Nebraska Symposium on Motivation (Vol. 17, pp. 237–307). Lincoln: University of Nebraska Press.
Zimbardo, P. G. (1975). Transforming experimental research into advocacy for social change. In M. D. H. A. Homstein (Ed.), Applying social psychology: Implications for research, practice and training (pp. 33–66). Hillsdale, NJ: Lawrence.
[...]
1 You cannot incur any financial loss in the prediction task. Your task performance will not affect your guaranteed show-up payment. Your task performance will not affect your own money.
2 Throughout this thesis, we use the term dishonesty to describe any form of dishonest behavior, including cheating, stealing, lying, and fraud. The terms are found to be used synonymously.
3 The experiment conducted was part of a joint collaboration experiment which aimed to test three hypotheses about influencing factors on dishonesty in VR. In total, the joint experiment involved 88 participants and four treatment conditions. Only data from n = 45 participants assigned to one of the two treatment conditions of interest is relevant for the analysis. The remaining 43 participants were assigned to either of the other two treatment conditions, which are not of interest for this analysis.
4 Technically, the reflective surface of the VR mirror played the video of an invisible camera, which constantly filmed the avatar from the position and angle of the VR mirror. This setup was used due to technical limitations and the need for a magnification of the reflection. From the visual appeal, we argue that this setup is indistinguishable from an actual reflective surface, and, induces mirror presence in the exact same way.
5 Throughout this analysis, the term significance describes statistical significance to a 95% confidence intervall (p < 0.05). unless specified otherwise.
6 All neccessary least-squares assumptions (LSAs) were tested and accepted. The assumptions were accepted using diagnostic plots (see Figure G2).
- Quote paper
- Cedric-Pascal Sommer (Author), 2019, Reflecting Yourself? The Influence of Mirrors and Avatars on Dishonest Behavior in Virtual Reality, Munich, GRIN Verlag, https://www.grin.com/document/594139
Similar texts
Publish now - it's free
Comments