فریب آشنایی: جنسیت، دوستیابی آنلاین و خودارائه گری اغراق آمیز

Abstract This study examined how differences in expectations about meeting impacted the degree of deceptive self-presentation individuals displayed within the context of dating. Participants filled out personality measures in one of four anticipated meeting conditions: face-to-face, email, no meeting, and a control condition with no pretense of dating. Results indicated that, compared to baseline measures, male participants increased the amount they self-presented when anticipating a future interaction with a prospective date. Specifically, male participants emphasized their positive characteristics more if the potential date was less salient (e.g., email meeting) compared to a more salient condition (e.g., face-to-face meeting) or the control conditions. Implications for self-presentation theory, online social interaction, and online dating research will be discussed.

1. Introduction “All the world’s a stage, and all the men and women merely players” – William Shakespeare. “On the Internet, no one knows you’re a dog” – Peter Steiner. In the words of Erving Goffman (1959, p. 9), life is akin to a performance: “When an individual plays a part, he [sic] implicitly requests his observers to take seriously the impression that is fostered before them.” Self-presentation describes this process or “performance” wherein individuals attempt to control the impressions others form of them ( Goffman, 1959, Jones, 1990, Rosenfeld et al., 2005 and Schlenker, 1980). As part of the self-presentation process, individuals seek to create specific impressions – to be liked, perceived as competent and perceived to be high in status – among their varying audiences. The context of the situation may influence the salience of certain self-presentational goals. For example, a person at a job interview may be more interested in appearing competent than likable, whereas a person on a date may have the opposite goal. With regard to communication mode, other contextual factors such as the availability of physical appearance and nonverbal cues may influence the ways in individuals self-present ( McKenna & Bargh, 2000). For instance, in a text-based context such as an email, an email signature rather than one’s clothing, briefcase, or car, may be the best way to convey status. Additionally, the combination of physical distance and relative anonymity inherent in communication via the Internet may facilitate a trend toward more deceptive self-presentation. The present study investigates these issues by examining the use of deceptive self-presentation in dating profiles as a function of the mode of communication in which a person will interact with a potential date. 1.1. Deceptive self-presentation Self-presentation is usually aimed toward achieving strategic goals (Leary, 1995). People tend to present and sometimes exaggerate or fabricate their characteristics in an attempt to create their desired impression. The present investigation focused on the type of self-presentation that is deceptive in nature. Research indicates that the likelihood of deceptive self-presentation increases as a function of the pressure to engage in self-presentation (Baumeister, 1992 and Feldman et al., 2002). For instance, interacting with a member of the opposite sex, particularly if they are attractive, increases motivation to self-present. A threat to one’s self-image also increases motivation to engage in self-presentation. Moreover, both factors have been shown to increase deceptive self-presentation (Rowatt et al., 1999 and Tyler and Feldman, 2005). Similar patterns of deceptive self-presentation have been shown in dating contexts. When presenting themselves to desirable potential dates, men are more likely than women to engage in deceptive self-presentation (Rowatt, Cunningham, & Druen, 1998). Additionally, the literature indicates that men and women use deceptive self-presentation to enhance different traits. Took and Camire (1991) surveyed male and female college students and asked them to indicate their willingness to engage in deceptive self-presentation to attract a mate. Men reported being more willing to use deception to appear more dominant, more resourceful, and more kind than they actually were. Conversely, women reported being willing to use deception to present their physical appearance as more favorable than it actually was. Similar findings have been reported in classic research on self-presentation. Specifically, Zanna and Pack (1975) demonstrated that women changed their self-reported sex role attitudes to match the gender role values (either traditional or non-traditional) of a perceiver when they believed the perceiver was a desirable male. Overall, the literature on deceptive self-presentation suggests that both the context of the interaction and gender of the interactants matter. 1.2. Deception online There are marked differences between face-to-face and computer-mediated communication (see Bargh and McKenna, 2004 and McKenna and Bargh, 2000 for reviews); the majority of these differences fall into four categories: relative anonymity, reduced importance of physical appearance, attenuation of physical distance, and greater control over the time and pace of interactions. Most germane to the present investigation is the relative anonymity inherent in many forms of computer-mediated interactions. The ability to be relatively anonymous in a social interaction online reduces accountability and leads to the depersonalization and deindividuation of the interactants (Postmes, Spears, & Lea, 2002). Therefore, anonymity increases the potential for antisocial behavior such as deception. Also relevant to the present study is the decreased importance of physical appearance inherent in online interactions (McKenna & Bargh, 2000). This visual anonymity may also increase the likelihood of deception. Similarly, the decreased emphasis on nonverbal cues relative to face-to-face interactions may also foster greater use of deception as the lack of nonverbal cues produce a feeling of anonymity in interactants (Sproull & Kriesler, 1986). Anonymous online social interaction has been described as limited, depleted, less rich, and impoverished (Hiltz et al., 1989, Siegel et al., 1986 and Sproull and Kriesler, 1986) due to its absence of nonverbal cues. Finally, the greater control over time and pace of interactions is also relevant as this feature of online communication contributes to the selective self-presentation that often occurs online (Gibbs, Ellison, & Heino, 2006). Taken together, the features of online communication reviewed above may all contribute to a greater propensity for individuals to engage in deception in online contexts. 1.3. Deceptive self-presentation in online dating While there are many ways to find potential dates online, the use of dating websites – websites specifically oriented toward helping people looking for romance – is increasing rapidly and has become a widely used means of finding potential romantic partners (Ellen, 2009, February 12). To participate in most online dating web sites, an individual registers by filling out a profile indicating desired mate preferences, providing demographic information, and sometimes completing personality measures. A photograph provided by the individual may or may not be provided with the profile. Initial contact between online daters usually takes the form of messages exchanged through the dating website, and, if communications continue, telephone or face-to-face contact may follow. Research examining the behavior of individuals using online dating web sites indicates that some online daters present an unrealistic or deceptive image of themselves (Brym and Lenton, 2001, Ellison et al., 2006, Hitsch et al., 2009 and Toma et al., 2008). For instance, Hitsch and colleagues reported that online daters exaggerated information about themselves and that men and women enhanced different characteristics – men emphasized their status; women emphasized their physical attractiveness (Hitsch et al., 2009 and Schmitt, 2002). In another study, men lied about height while women lied about weight (Toma et al., 2008). Furthermore, participants in this study reported being accurate in the photographs they posted and when reporting relationship details. Thus, the literature emerging from online dating research indicates that online daters do engage in deceptive self-presentation but may balance their deception against the constraints set in place by the promise of a future interaction. 1.4. Gender differences in mate selection Men and women are similar in that they both want mates that are kind, reliable, outgoing and smart (Botwin, Buss, & Shackelford, 1997). However, there are also notable differences in the mate preferences of men and women. Owing to the differences in men and women’s parental investment, human mate selection is one of female choice (Darwin, 1871). This is illustrated by the gender difference in the proportion of men vs. women who get approached through their online dating profiles. Specifically, men approach women through online dating sites more than women approach men. For instance, once study of online daters reported that 57% of men vs. 23% of women never got a single email from a prospective date (Hitsch et al., 2009). Moreover, contact from prospective dates varied as a function of the content of participants’ profiles in a manner predicted by the evolutionary psychological framework on mate selection (Buss, 1989 and Kenrick et al., 1990). In the profiles of actual online daters, Hitsch et al. (2009) reported that for men, income – an indicator of status – was most predictive of getting approached by potential daters through the website, with higher earners getting more emails. For women, physical appearance – an indicator of fertility – garnered the most emails from potential suitors. Both short men and overweight women were the least likely to get emails through the dating site. These data are consistent with the deceptive self-presentational practices of men and women reviewed above (Toma et al., 2008). Men and women who are searching for a mate are aware of what potential mates consider attractive and the evidence indicates that they will alter their profiles to reflect these characteristics. The research on mate selection also indicates that there may be gender differences in the preferred personality characteristics of a mate. One way in which personality preferences in mate selection have been examined is in terms of the five-factor model, also called the “Big 5” (McCrae & Costa, 1986). This model of personality consists of the following dimensions: neuroticism (emotional stability), extraversion, openness to new experience, conscientious, and agreeableness (kind and helpful). In an experimental setting, Jensen-Campbell, Graziano, and West (1996) demonstrated that female participants preferred to date men who were helpful, particularly when they were both helpful and dominant. Similarly, research on personality and mate preferences indicates that newly married women perceive their husbands as being high in agreeableness. (Botwin et al., 1997). As such, the limited evidence suggests that women may differentiate on personality more so than will men when selecting a mate.

4. Results A series of 2 (participant gender: men vs. women) × 4 (meeting condition: face-to-face vs. no meeting vs. questionnaire control) × 2 (time: pretest vs. posttest) mixed design analysis of variances were conducted to assess the impact of gender and meeting condition on participants’ Big 5 scores and self-reported attractiveness rating. We examined any resulting interactions between time and other variables using a difference score between pretest and posttest. For analyses using this difference score, positive means indicate that participants reported a higher mean at posttest whereas negative means indicate that participants reported a lower mean at posttest. Pre- and post-test means and standard deviations for all our dependent measures by experimental condition are presented in Table 1 and Table 2. Table 1. Means and standard deviations for the Big 5 items at pre- and post-test for men. Online dating context Email Face-to-face No meeting Questionnaire N = 15 N = 15 N = 15 N = 15 Time 1 Agreeableness 5.20 (1.10) 5.09 (0.91) 5.10 (0.93) 5.38 (0.88) Extraversion 4.50 (1.68) 5.31 (0.89) 4.68 (1.64) 4.20 (1.47) Openness 5.25 (1.24) 5.35 (1.14) 5.21 (1.08) 4.93 (0.94) Conscientiousness 4.57 (0.83) 4.62 (0.93) 4.73 (1.16) 4.66 (1.23) Neuroticism 4.10 (1.38) 3.52 (1.53) 3.26 (1.06) 2.89 (1.21) Time 2 Agreeableness 5.89 (0.80) 5.75 (1.00) 5.21 (0.96) 5.14 (0.69) Extraversion 4.60 (0.99) 4.97 (0.68) 4.67 (0.79) 4.15 (1.28) Openness 5.38 (1.05) 5.21 (1.20) 5.32 (0.88) 4.83 (0.95) Conscientiousness 4.85 (0.56) 5.06 (1.09) 4.97 (1.03) 4.82 (0.93) Neuroticism 3.08 (1.56) 2.87 (1.29) 3.52 (1.48) 3.17 (1.25) Table options Table 2. Means and standard deviations for the Big 5 items at pre- and post-test for women. Online dating context Email Face-to-face No meeting Questionnaire N = 23 N = 23 N = 23 N = 21 Time 1 Agreeableness 4.75 (1.12) 4.78 (0.66) 4.83 (0.74) 5.68 (1.19) Extraversion 4.96 (1.39) 5.53 (1.54) 5.16 (1.19) 5.10 (1.52) Openness 4.70 (1.14) 4.81 (1.03) 4.99 (0.87) 4.69 (1.58) Conscientiousness 4.93 (1.52) 4.62 (1.16) 4.39 (1.25) 4.52 (1.50) Neuroticism 3.63 (1.11) 3.63 (1.00) 3.91 (1.26) 3.82 (1.06) Time 2 Agreeableness 5.07 (1.01) 4.76 (0.74) 4.74 (0.79) 6.07 (1.03) Extraversion 4.75 (0.94) 4.82 (0.61) 4.78 (1.03) 5.35 (1.41) Openness 4.58 (0.87) 4.76 (0.65) 4.69 (0.68) 4.54 (1.31) Conscientiousness 4.92 (0.99) 4.49 (0.74) 4.59 (0.92) 4.62 (1.49) Neuroticism 3.71 (0.86) 3.90 (0.85) 3.97 (0.80) 3.89 (1.41) Table options 4.1. Changes in Big 5 personality Prior to analyzing the data, we calculated the Big 5 personality subscales and found them to be reliable at both time points. The pre- and post-test reliabilities for each subscale were as follows: agreeableness α = .67 at pretest and α = .72 at posttest; extraversion α = .88 at pretest and α = .73 at posttest; openness α = .79 and α = .79 at posttest; conscientiousness α = .72 at pretest and α = .79 at posttest; and neuroticism α = .76 at pretest and posttest. Additionally, with regard to any interactions, the means reported below indicate a difference score between pretest and posttest score for each participant. 4.1.1. Agreeableness Participants reported being significantly more agreeable at posttest than pretest (M = 5.10, SD = .98 vs. M = 5.32, SD = .99), Wilk’s λ = .927, F (1, 152) = 11.37, p = .001, View the MathML sourceηp2=.073. In addition, a significant time by meeting condition interaction emerged, Wilk’s λ = .939, F (3, 152) = 3.13, p = .028, View the MathML sourceηp2=.061. Simple effects indicated that participants expecting to meet their dates via email (M = .46, SD = .75) reported being significantly more agreeable at posttest than participants in the no meeting (M = −.01, SD = .82) or the questionnaire control conditions (M = .09, SD = .93). This was further qualified by a significant three-way interaction between participant gender, meeting condition, and time, Wilk’s λ = .917, F (3, 152) = 4.33, p = .006, View the MathML sourceηp2=.083. Simple effects revealed that men in the face-to-face (M = .65, SD = .80) and email (M = .69, SD = .79) conditions reported being more agreeable at posttest than did men in the questionnaire control (M = −.24, SD = .77) and no meeting conditions (M = .10, SD = .94), F (3, 144) = 5.17, p = .002, View the MathML sourceηp2=.097. For women, there were no significant changes in agreeableness in any of the experimental conditions (see Fig. 2 for a display of means by condition). Change in agreeableness over time broken down by condition and gender. (For ... Fig. 2. Change in agreeableness over time broken down by condition and gender. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Figure options 4.1.2. Neuroticism There was a significant time by participant gender by meeting condition interaction on neuroticism scores, Wilk’s λ = .942, F (3152) = 2.95, p = .035, View the MathML sourceηp2=.058. Simple effects revealed that men in the face-to-face condition (M = −.65, SD = 1.71) and in the email condition (M = −1.03, SD = 1.44) reported that they were significantly less neurotic than men in the questionnaire control (M = .26, SD = 1.27) and the no meeting conditions (M = .28, SD = 1.61), F (3, 144) = 4.29, p = .006, View the MathML sourceηp2=.082. For women, neuroticism was not significantly different in any of the experimental conditions (see Fig. 3 for a display of means by condition). Change in neuroticism over time broken down by condition and gender. (For ... Fig. 3. Change in neuroticism over time broken down by condition and gender. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Figure options No main effects or interactions on extraversion, openness to new experience, or conscientiousness emerged. 4.2. Changes in attractiveness rating We collected a measure of self-reported attractiveness at both time points and also photographed participants when they came into the lab for phase 2 of the study. Two independent raters (both female) blind to condition evaluated the physical attractiveness of each participant on a scale ranging from 1 (not at all attractive) to 7 (extremely attractive). These ratings were averaged to create an index of participant attractiveness, with an inter-rater reliability of α = .66. An ANOVA on these ratings revealed no significant main effects or interactions. Thus, the perceivers’ evaluation of participants’ attractiveness indicated that there was no difference by condition as would be expected by successful randomization to condition. See Table 3 for a display of means and standard deviations of the coders’ averaged ratings. Table 3. Self-reported attractiveness at pre- and post-test and independent judges’ attractiveness ratings by condition. Online dating context Email Face-to-face No meeting Questionnaire Time 1 Male 4.67 (1.91) 5.80 (1.37) 5.53 (1.46) 5.47 (1.70) Female 5.87 (1.91) 5.65 (1.99) 5.62 (1.44) 5.95 (1.99) Time 2 Male 6.00 (1.41) 6.20 (0.94) 6.00 (1.77) 5.41 (1.06) Female 6.26 (1.18) 5.87 (1.36) 5.62 (0.92) 5.55 (1.71) Judges Male 4.23 (1.25) 3.50 (1.18) 3.90 (0.93) 3.22 (1.09) Female 3.81 (1.19) 3.57 (1.15) 3.30 (1.33) 3.86 (1.23) Note: To assess self-reported attractiveness, participants rated their attractiveness compared to other college students using a scale of 1 (much less than average) to 10 (much more than average). Two independent raters blind to condition evaluated the physical attractiveness of each participant on a scale ranging from 1 (not at all attractive) to 7 (extremely attractive). Table options However, when we evaluated participants’ self-reports, a different pattern emerged. Specifically, there was a main effect for self-reported attractiveness across time indicating that participants reported being more attractive at posttest than pretest (M = 5.57, SD = 1.75 vs. M = 5.88, SD = 1.31), Wilk’s λ = .960, F (1, 145) = 6.11, p = .015, View the MathML sourceηp2=.040. This was qualified by a significant interaction between meeting condition and time, Wilk’s λ = .943, F (3, 145) = 3.92, p = .036, View the MathML sourceηp2=.057. Post hoc tests, using Fisher’s LSD, revealed that participants in the email condition (M = .76, SD = 1.58) reported being significantly more attractive at posttest than participants in questionnaire control condition (M = −.18, SD = 1.43). Participants in the face-to-face (M = .29, SD = 1.71) and no meeting (M = .18, SD = 1.32) conditions also enhanced their attractiveness self-reports overtime but the difference was not significant from the other meeting conditions. Next, we examined the data to determine whether the above findings persisted when controlling for participants’ objective physical attractiveness as assessed by our independent coders. A repeated measure ANOVA using the coder’s average ratings as a covariate revealed a significant two-way interaction between self-reported attractiveness and meeting condition, Wilk’s λ = .943, F (3, 139) = 2.80, p = .042, View the MathML sourceηp2=.057. Post hoc tests using Fisher’s LSD indicated the nature of the interaction was identical to that reported above. However, this analysis also revealed a significant participant gender by time interaction indicating that men (M = .56, SD = 1.39) reported being more attractive at posttest than at pretest compared to women (M = .027, SD = 1.59), Wilk’s λ = .968, F (1, 139) = 4.54, p = .035, View the MathML sourceηp2=.032. As displayed in Fig. 4, this interaction was driven by men in the email condition – they reported being more attractive at posttest than participants in any other condition. Change in attractiveness over time broken down by condition and gender and ... Fig. 4. Change in attractiveness over time broken down by condition and gender and controlling for objective attractiveness. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)