1. Introduction In the last decade, interactions on Social Networking Sites (from now on: SNSs) have become part of many people's daily routines. Although many platforms allow their users to customize their online experience or to provide space and context for useful and enjoyable exchanges, not all SNS's interac-tions are pleasant ones. The risk of being a target of aggressive behaviors is significant, especially for women and users belonging to minorities or dis-criminated groups (Rainie et al., 2012; Duggan, 2014). In this project, we test whether online incivility on SNSs plays a detri-mental role in lay people's pro-social attitudes, especially trust, and whether women are generally more vulnerable to this effect. In order to seek evidence on this, we tested 254 participants' reactions to different Facebook interac-tions concerning four emotionally charged topics, and their interactions in a trust game. Preliminary results show that priming participants with online interactions positive correlate with trust impairment and emotional worsen-ing, especially on female participants. Moreover, female participants were generally more likely to avoid trustworthy commitments overall, which led us to think that the use of Facebook may per se prompt a negative emotional response in women. 2. Theoretical perspectives The quantity of disclosures and the quality of interactions in our offline routines are determined by cultural frames, norms, politeness rules and miti-gation habits. Moreover, our offline lives hardly ever expose us to huge masses of diverse and randomly reunited interlocutors who share no com-monalities. Apparently, this does not correspond to all online social environ-ments: the more the online networks are open, wide, public, generalist, un-personalized and complex (i.e.: the more their parameters do not resemble those of preferred or usual offline interactions), the more the occurrence of non-cooperative and flaming behaviors becomes more likely. (Currarini et al., 2009; Econometrica, 2014; Dall'Asta et al., 2013, PNAS). In fact, expo-sure to massive diversity can weaken trust and pro-social attitudes. The expe-rience of diversity on generalist, wide and poorly customizable online con-texts (e.g.: those of newspapers' comment sections or message boards) may lead SNSs users to experience a lower degree of perceived control on the platforms, which may as well result in a sense of higher vulnerability and in the need of enacting self-defensive behaviors (e.g.: the use of anonymity or a higher disposition to react or even overreact to emotionally charged issues). At the same time, online asynchronous exchanges may lead to a higher de-gree of perceived control on the interaction and in a lower need of smoothing the interactional management by means of mitigating or face-saving strate-gies, resulting in displays of higher directness, intensity and higher general assertiveness if compared to offline attitudes. When these factors collide, irrational and hostile outcomes arise more easily. Online incivility, which includes aggressive or disrespectful behaviors, vile comments, online harassment, and hate speech is a widespread and rising phenomenon (Rainie et al., 2012; Duggan, 2014). In our project, we are try-ing to seek evidence on what kind of effects it may have on social trust, espe-cially on women. 3. Data and choice of methods Our experimental design finds itself at the intersection of cognitive psy-chology and behavioral economics. A total of 254 participants (134 identified as male and 120 identified as female) responded to a multi-scaled anonymous survey where socio-demographic data, trust attitudes, personality features and SNSs experience were tested. Collected socio-demographic data included information on participants' gender, age, origin and educational qualification. Initial information on their trust attitude was collected via Yamagishi's 6 items scale (Yamagishi, 1986; Yamagishi & Yamagishi, 1994). Personality features were collected via the abbreviated BIG 5 test. Social network experi-ence was tested by means of a dedicated multi-item scale, which investigated the participants' familiarity, competence, attitudes and concerns about the platforms. The participants thus had to interact with four Facebook top-ic/comment scenarios on distinct emotionally charged issues. Manipulations on the same topics were divided into uncivil/civil and anonymous/non anon-ymous tokens. Subsequently, we tested the participants' trust dispositions by means of a trust game and their emotional status by means of a visual ana-logue scale (VAS). In the trust game (TG) a "trustor" (player A) is randomly matched with a "trustee" (player B). Both players are endowed with 10 ECU (1 ECU=0.50EUR) in order to avoid inequity/inequality concerns. Player A can send to player B any amount "X" between 0 and 10 ECU. The amount sent by player A is then multiplied by a factor of 3 by the experi-menters and player B gets X*3 on her account. At this point, player B can send any amount between 0 and X*3 back to player A. In order to elicit the full strategy set, responses by player B are elicited implementing the strategy-method. 4. Preliminary results 127 participants played the game in the role of trustor A (57 males and 70 females) and the remaining 127 participants played the game in the role of trustee B (77 males and 50 females). Results currently showed no gender effects related to the treatment variations. However, we detected an inde-pendent gender effect on the trust game itself, insofar as on average female participants tended to provide remarkably smaller (-30%) amounts of ECU in the role of trustor (male: 6.19, female: 4.13; MWU p=0.0004). This gender effect holds in all the four different factorial configurations, by gender (civ-il/uncivil, anonymous/ non-anonymous). On the other hand, male participants tended to send smaller amounts only after being exposed to uncivil stimuli (civil: 6.88, uncivil: 5.32; MWU p=0.11) whereas they tended to send conspicuous amounts after being ex-posed to the anonymous configuration (anonymous: 7, non-anonymous 5.6; MWU p=0.07). 5. General discussion Studies advancing pessimistic claims on the role of SNSs highlight that individual well-being may be directly affected by online networking due to the inherent characteristics of Internet-mediated interactions (de Vries and Kühne, 2015; Lim and Yang, 2015; Tandoc et al., 2015; Kross et al., 2013; Krasnova et al., 2013; Chou and Edge, 2012; Muise et al., 2009). However, it is important to look at these phenomena also in terms of gender differences: recent findings reveal that males and females differ in how they use and per-ceive SNSs (Chan, Cheung, Shi, & Lee, 2015; Nesi & Prinstein, 2015; Choi & Kim, 2014; Junco, 2013). In fact, the very structure of generalist SNSs seems likely to frustrate the most common social motivations that prompt women to use them, i.e. communicating with others and managing relation-ships. Insofar as most SNSs, including Facebook, are more a broadcast tool than a discussion tool (Kirschner, 2015), they may not be best suited to pro-vide a satisfactory social environment for female users, whom, in turn, may have adverse effects on their social perception of other users.
The effects of online interaction on trust: An experimental study with Facebook primes
Paglieri F;Bonelli;
2015
Abstract
1. Introduction In the last decade, interactions on Social Networking Sites (from now on: SNSs) have become part of many people's daily routines. Although many platforms allow their users to customize their online experience or to provide space and context for useful and enjoyable exchanges, not all SNS's interac-tions are pleasant ones. The risk of being a target of aggressive behaviors is significant, especially for women and users belonging to minorities or dis-criminated groups (Rainie et al., 2012; Duggan, 2014). In this project, we test whether online incivility on SNSs plays a detri-mental role in lay people's pro-social attitudes, especially trust, and whether women are generally more vulnerable to this effect. In order to seek evidence on this, we tested 254 participants' reactions to different Facebook interac-tions concerning four emotionally charged topics, and their interactions in a trust game. Preliminary results show that priming participants with online interactions positive correlate with trust impairment and emotional worsen-ing, especially on female participants. Moreover, female participants were generally more likely to avoid trustworthy commitments overall, which led us to think that the use of Facebook may per se prompt a negative emotional response in women. 2. Theoretical perspectives The quantity of disclosures and the quality of interactions in our offline routines are determined by cultural frames, norms, politeness rules and miti-gation habits. Moreover, our offline lives hardly ever expose us to huge masses of diverse and randomly reunited interlocutors who share no com-monalities. Apparently, this does not correspond to all online social environ-ments: the more the online networks are open, wide, public, generalist, un-personalized and complex (i.e.: the more their parameters do not resemble those of preferred or usual offline interactions), the more the occurrence of non-cooperative and flaming behaviors becomes more likely. (Currarini et al., 2009; Econometrica, 2014; Dall'Asta et al., 2013, PNAS). In fact, expo-sure to massive diversity can weaken trust and pro-social attitudes. The expe-rience of diversity on generalist, wide and poorly customizable online con-texts (e.g.: those of newspapers' comment sections or message boards) may lead SNSs users to experience a lower degree of perceived control on the platforms, which may as well result in a sense of higher vulnerability and in the need of enacting self-defensive behaviors (e.g.: the use of anonymity or a higher disposition to react or even overreact to emotionally charged issues). At the same time, online asynchronous exchanges may lead to a higher de-gree of perceived control on the interaction and in a lower need of smoothing the interactional management by means of mitigating or face-saving strate-gies, resulting in displays of higher directness, intensity and higher general assertiveness if compared to offline attitudes. When these factors collide, irrational and hostile outcomes arise more easily. Online incivility, which includes aggressive or disrespectful behaviors, vile comments, online harassment, and hate speech is a widespread and rising phenomenon (Rainie et al., 2012; Duggan, 2014). In our project, we are try-ing to seek evidence on what kind of effects it may have on social trust, espe-cially on women. 3. Data and choice of methods Our experimental design finds itself at the intersection of cognitive psy-chology and behavioral economics. A total of 254 participants (134 identified as male and 120 identified as female) responded to a multi-scaled anonymous survey where socio-demographic data, trust attitudes, personality features and SNSs experience were tested. Collected socio-demographic data included information on participants' gender, age, origin and educational qualification. Initial information on their trust attitude was collected via Yamagishi's 6 items scale (Yamagishi, 1986; Yamagishi & Yamagishi, 1994). Personality features were collected via the abbreviated BIG 5 test. Social network experi-ence was tested by means of a dedicated multi-item scale, which investigated the participants' familiarity, competence, attitudes and concerns about the platforms. The participants thus had to interact with four Facebook top-ic/comment scenarios on distinct emotionally charged issues. Manipulations on the same topics were divided into uncivil/civil and anonymous/non anon-ymous tokens. Subsequently, we tested the participants' trust dispositions by means of a trust game and their emotional status by means of a visual ana-logue scale (VAS). In the trust game (TG) a "trustor" (player A) is randomly matched with a "trustee" (player B). Both players are endowed with 10 ECU (1 ECU=0.50EUR) in order to avoid inequity/inequality concerns. Player A can send to player B any amount "X" between 0 and 10 ECU. The amount sent by player A is then multiplied by a factor of 3 by the experi-menters and player B gets X*3 on her account. At this point, player B can send any amount between 0 and X*3 back to player A. In order to elicit the full strategy set, responses by player B are elicited implementing the strategy-method. 4. Preliminary results 127 participants played the game in the role of trustor A (57 males and 70 females) and the remaining 127 participants played the game in the role of trustee B (77 males and 50 females). Results currently showed no gender effects related to the treatment variations. However, we detected an inde-pendent gender effect on the trust game itself, insofar as on average female participants tended to provide remarkably smaller (-30%) amounts of ECU in the role of trustor (male: 6.19, female: 4.13; MWU p=0.0004). This gender effect holds in all the four different factorial configurations, by gender (civ-il/uncivil, anonymous/ non-anonymous). On the other hand, male participants tended to send smaller amounts only after being exposed to uncivil stimuli (civil: 6.88, uncivil: 5.32; MWU p=0.11) whereas they tended to send conspicuous amounts after being ex-posed to the anonymous configuration (anonymous: 7, non-anonymous 5.6; MWU p=0.07). 5. General discussion Studies advancing pessimistic claims on the role of SNSs highlight that individual well-being may be directly affected by online networking due to the inherent characteristics of Internet-mediated interactions (de Vries and Kühne, 2015; Lim and Yang, 2015; Tandoc et al., 2015; Kross et al., 2013; Krasnova et al., 2013; Chou and Edge, 2012; Muise et al., 2009). However, it is important to look at these phenomena also in terms of gender differences: recent findings reveal that males and females differ in how they use and per-ceive SNSs (Chan, Cheung, Shi, & Lee, 2015; Nesi & Prinstein, 2015; Choi & Kim, 2014; Junco, 2013). In fact, the very structure of generalist SNSs seems likely to frustrate the most common social motivations that prompt women to use them, i.e. communicating with others and managing relation-ships. Insofar as most SNSs, including Facebook, are more a broadcast tool than a discussion tool (Kirschner, 2015), they may not be best suited to pro-vide a satisfactory social environment for female users, whom, in turn, may have adverse effects on their social perception of other users.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
