SEARCH

SEARCH BY CITATION

Keywords:

  • online social networks;
  • personality;
  • interpersonal perception;
  • zero acquaintance;
  • lens model

Abstract

  1. Top of page
  2. Abstract
  3. ONLINE SOCIAL NETWORKS—A REALISTIC CONTEXT FOR UNDERSTANDING PERSONALITY EXPRESSION AND IMPRESSION FORMATION
  4. PRIOR RESEARCH ON PERSONALITY EXPRESSION AND IMPRESSION FORMATION AT ZERO ACQUAINTANCE
  5. THE PRESENT RESEARCH
  6. METHOD
  7. RESULTS
  8. DISCUSSION
  9. CONCLUSION
  10. ACKNOWLEDGEMENTS
  11. REFERENCES
  12. Supporting Information

In this paper, we investigate personality expression and impression formation processes in online social networks (OSNs). We explore whether, when and why people accurately judge others' personalities (accuracy), successfully manage the impressions that others form of them (impression management) and accurately infer others' impressions of them (meta-accuracy) at zero acquaintance. On the basis of targets' OSN profiles (N = 103), overall perceiver impressions were collected and compared with targets' self-view, desired impression and meta-perception. In addition, independent groups of thin-slice perceivers based their personality impressions solely on one of four kinds of information within the OSN profiles (profile picture, interests field, group list and notice board), and more than 300 OSN cues (e.g. attractive person and number of friends) were coded. Results showed evidence of accuracy, impression management and meta-accuracy, but their extent was moderated by the trait (e.g. Big Five and self-esteem), the kind of information and the interplay of trait and information. Findings could be explained by cue expression and cue utilization processes (lens model analyses). Future prospects for studying personality impressions in online and offline environments are discussed. Copyright © 2013 European Association of Personality Psychology.

Amsterdam, 10 am

‘We definitely should invite Amy to your birthday party—she really seems to be an open-minded girl,’ Ann suggests to her husband. ‘How would you know that? You've never met her!’ he answers, truly surprised. ‘Well, I just checked out her Facebook profile.’

Istanbul, 5:20 pm

‘Alright, I admit I changed my Facebook groups because I want to convey the impression of valuing VIPs and glamour and stuff. I'm sure it will help advance my career in the fashion business,’ Irem says as an answer to her sister's question on the phone.

Miami, 7 pm

‘You mean when people view my Facebook profile? Yeah, I guess they think I'm in search of excitement all the time. But who cares—I just mentioned some of my main interests there,’ Megan explains to her cousin on the way to the wakeboarding course.

These fictitious examples vividly demonstrate interpersonal perception phenomena in online social networks (OSNs) such as Facebook: People quickly form impressions of others (see Ann in the example above), try to influence how they come across (see Irem) and think about how they are viewed by other OSN users (see Megan). The success of these complex intertwined perception processes—accuracy, impression management and meta-accuracy—is essential for people's social functioning in online as in offline environments. Here, we aimed at understanding whether, when and why people accurately judge others' personalities, successfully manage the impressions that others form of them and accurately infer others' impressions of them on the basis of OSN profiles.

ONLINE SOCIAL NETWORKS—A REALISTIC CONTEXT FOR UNDERSTANDING PERSONALITY EXPRESSION AND IMPRESSION FORMATION

  1. Top of page
  2. Abstract
  3. ONLINE SOCIAL NETWORKS—A REALISTIC CONTEXT FOR UNDERSTANDING PERSONALITY EXPRESSION AND IMPRESSION FORMATION
  4. PRIOR RESEARCH ON PERSONALITY EXPRESSION AND IMPRESSION FORMATION AT ZERO ACQUAINTANCE
  5. THE PRESENT RESEARCH
  6. METHOD
  7. RESULTS
  8. DISCUSSION
  9. CONCLUSION
  10. ACKNOWLEDGEMENTS
  11. REFERENCES
  12. Supporting Information

Online social networks are interactive Web technologies that allow their users to present themselves with a public or semi-public profile and to establish visible connections to other users within the network (boyd & Ellison, 2007). OSNs are the important social phenomenon of our time: More than 1.2 billion people around the world use OSNs for a considerable amount of time, which equals 82% of the world's online population (comScore, 2011). As of March 2013, the world's largest OSN, Facebook, reported having 655 million daily active users (Facebook, 2013).

Aside from being ubiquitous in many people's social lives, OSNs provide a unique opportunity for studying interpersonal perception: First, there is substantial variance in possible behaviours that users can engage in. OSNs allow their users to self-present, behave, interact and communicate with each other in numerous ways (boyd & Ellison, 2007; Gosling, Augustine, Vazire, Holtzman, & Gaddis, 2011). Users can, for example, present themselves in terms of a personal profile (provide a profile picture and present a favourite quote). They can connect to and communicate with other users (e.g. via notice board entries), and they can present their social relationships and preferences (e.g. by creating friend lists, joining groups and listing interests).

Second, because OSNs are a relatively new social domain, there are many degrees of freedom concerning how to behave appropriately. The social rules and norms of interactions within this novel social sphere have yet to be negotiated among the users. Therefore, OSNs provide a perfect setting for individual differences to be expressed. Third, OSN communication processes are strongly intertwined with offline interactions (e.g. notice board entries and links to pictures) and offline social networks (e.g. friend lists and group memberships). OSNs are routinely used for communication and networking by individuals in virtually all societies and cultural settings (Ross, Orr, Sisic, Arseneault, Simmering, & Orr, 2009; Valkenburg & Peter, 2009). Thus, they capture important aspects of people's everyday behaviours and interactions.

Fourth, in contrast to most laboratory studies where targets know that they are part of a psychological experiment (Orne, 1962), analysing OSN profiles allows for the assessment of a rich set of naturally occurring behaviours. While showing these behaviours, targets are not aware of participating in a scientific research project. Finally, most of the users' OSN behaviours are registered and leave perceivable cues on the users' OSN profiles. These data are structured and remains available for a certain period of time. Thus, OSNs can easily be used as a tool for standardized empirical analyses (Lewis, Kaufman, Gonzalez, Wimmer, & Christakis, 2008; Watts, 2007).

As OSNs become more and more popular, related research is expanding rapidly (see Wilson, Gosling, & Graham, 2012, for a recent review of articles on Facebook). So far, however, there have been only a few empirical research articles on personality and interpersonal perception in OSNs. Studies often rely on self-reports or artificial manipulations of OSN profile information, and outcomes on circumscribed phenomena are described anecdotally. In the following, we provide a condensed overview of important phenomena in the interpersonal perception domain and its applications in OSN research.

PRIOR RESEARCH ON PERSONALITY EXPRESSION AND IMPRESSION FORMATION AT ZERO ACQUAINTANCE

  1. Top of page
  2. Abstract
  3. ONLINE SOCIAL NETWORKS—A REALISTIC CONTEXT FOR UNDERSTANDING PERSONALITY EXPRESSION AND IMPRESSION FORMATION
  4. PRIOR RESEARCH ON PERSONALITY EXPRESSION AND IMPRESSION FORMATION AT ZERO ACQUAINTANCE
  5. THE PRESENT RESEARCH
  6. METHOD
  7. RESULTS
  8. DISCUSSION
  9. CONCLUSION
  10. ACKNOWLEDGEMENTS
  11. REFERENCES
  12. Supporting Information

People's first impressions of one another (zero-acquaintance judgements; Albright, Kenny, & Malloy, 1988) are central for successful interpersonal functioning and decision making (Leising & Borkenau, 2011; Swann, 1984). Three aspects of the complex interpersonal perception dynamics are of particular importance: how well people know others' personalities (accuracy), how successful they are in communicating their desired impression (impression management) and how much they are aware of the impressions they evoke in others (meta-accuracy). So far mostly studied in offline environments, all of these phenomena are applicable to online contexts such as OSNs.

The amount of accuracy, impression management and meta-accuracy

Accuracy

The accuracy of personality impressions is often measured as (i) the agreement among several perceivers (consensus) and (ii) the convergence of perceiver judgements and the target's personality. The latter can be assessed by observing a target's behaviour (Funder & Sneed, 1993; Levesque & Kenny, 1993) or by collecting self-reports (Funder & Colvin, 1997) and/or informant reports of the target's personality (Naumann, Vazire, Rentfrow, & Gosling, 2009). In general, our first impressions of others are surprisingly consensual and accurate (Funder, 2012), as has been shown in numerous contexts such as face-to-face meetings (Kenny, Albright, Malloy, & Kashy, 1994; Watson, 1989), images and videotapes (Borkenau & Liebler, 1992; Hirschmüller, Egloff, Nestler, & Back, 2013; Nestler, Egloff, Küfner, & Back, 2012), texts (Holleran & Mehl, 2008; Küfner, Back, Nestler, & Egloff, 2010), music preferences (Rentfrow & Gosling, 2006) and personal environments (Gosling, Ko, Mannarelli, & Morris, 2002). Perceiver impressions formed on the basis of OSN profiles were found to be similar to those formed in offline settings (Weisbuch, Ivcevic, & Ambady, 2009; Zywica & Danowski, 2008) and considerably accurate (Back, Stopfer, et al., 2010; Buffardi & Campbell, 2008).

Impression management

The term impression management summarizes the process by which individuals control the impressions others form of them (Goffman, 1959; Leary & Kowalski, 1990; Paulhus, 1984). Although research has shown that people are motivated to engage in impression management (Leary, 1995; Paulhus & John, 1998), there are surprisingly few studies on the success of impression management (i.e. whether perceivers actually formed distorted impressions of the target's personality; see Human, Biesanz, Parisotto, & Dunn, 2012; Murphy, 2007; Vazire & Gosling, 2004, for exceptions). In OSNs, users intend to manage the impressions they evoke in others with their OSN profiles (Peluchette & Karl, 2010), but self-idealization does not seem to be successful (i.e. perceiver impressions are not distorted towards the targets' ideal self; Back, Stopfer, et al., 2010).

Meta-accuracy

People's views of how others generally see them (meta-perceptions; Laing, Phillipson, & Lee, 1966) are substantially valid (termed meta-accuracy; Kenny, 1994; Kenny & DePaulo, 1993), even at zero or short-term acquaintance (Carlson, Vazire, & Furr, 2011; DePaulo, Kenny, Hoover, Webb, & Oliver, 1987). In online contexts, meta-accuracy is clearly an understudied domain. One study (Sherman et al., 2001) showed that targets failed to differentiate between the impressions they made in face-to-face and online contexts (i.e. they underestimated the impact of the medium). It is an open empirical question whether there is meta-accuracy based on OSN profiles.

Moderators

Moderator variables provide important insights into the conditions under which certain effects emerge such as the accuracy in judging personality (Funder, 2012). There are differences in the judgeability between traits (i.e. the good trait; e.g. extraversion vs. neuroticism; Connelly & Ones, 2010) depending on their observability (i.e. how much they are expressed in observable cues) and/or desirability (Funder & Dobroth, 1987; John & Robins, 1993). OSNs and other online environments, where the expression of personal content is accentuated, seem to promote accurate impressions of openness (Back, Schmukle, & Egloff, 2008; Back, Stopfer, et al., 2010; Vazire & Gosling, 2004).

Besides traits, certain kinds of information may be more or less diagnostic of personality (i.e. good information; e.g. physical appearance vs. personal interests; Naumann et al., 2009; Rentfrow & Gosling, 2006). Kinds of information may vary with regard to the quantity (Borkenau, Mauer, Riemann, Spinath, & Angleitner, 2004) and quality (Borkenau & Liebler, 1992; DePaulo, Lanier, & Davis, 1983) of cues that they reveal. Results on online impressions denote the relevance of profile pictures (Stecher & Counts, 2008; Steele, Evans, & Green, 2009).

Finally, some traits are judged better on the basis of one kind of information, whereas other traits are judged better on the basis of other kinds of information (i.e. the interplay of trait × information, termed diagnosticity; e.g. observing openness in personal environments vs. in face-to-face meetings; Gosling et al., 2002). So far, there are few findings on moderator variables in the online impression formation domain. In addition, we do barely have insight into moderating variables of impression management and meta-accuracy in OSNs.

Mediators

Whereas the described moderator variables can specify when certain effects will hold, mediators explain how or why such effects occur. To achieve accuracy in personality judgements, perceivers must base their impressions of a target on information that is actually related to a target's personality. Analysing informational cues as mediators is a common approach for explaining the accuracy (or the lack thereof) of personality judgements (Back et al., 2008; Borkenau & Liebler, 1992; Graham & Gosling, 2012; Küfner et al., 2010). For example, the accuracy of narcissism ratings based on OSN profiles was mediated by the quantity of social interaction and self-promotion, and attractiveness in the profile picture (Buffardi & Campbell, 2008). Most OSN research, however, has focused solely on informational cues related either to the users' personality (Ivcevic & Ambady, 2013; Krämer & Winter, 2008; Mehdizadeh, 2010) or perceivers' impressions (Tong, Van Der Heide, Langwell, & Walther, 2008; Weisbuch et al., 2009). Mediators of impression management and meta-accuracy in OSNs have not been analysed.

An integrative lens model framework to understand the moderating and mediating processes of accuracy, impression management and meta-accuracy in online social networks

To understand the dynamics of personality impressions based on OSN profiles, researchers have started to consider moderating and mediating variables, thereby mainly focusing on the accuracy of personality impressions. However, we assume that similar processes underlie varying degrees of impression management and meta-accuracy in OSNs. Here, we apply an integrative lens model framework that conceptualizes accuracy, impression management and meta-accuracy as intertwined interpersonal perception phenomena that are the result of common cue-related processes.

According to Brunswik's lens model (1956), accuracy arises from the expression and utilization of valid behavioural cues (mediators). We extend this assumption in various ways: First, on the basis of Brunswik's concept of vicarious functioning (Brunswik, 1956; see Cooksey, 1996, for an overview), cues are thought to be (at least partly) substitutable for one another because of their intercorrelations (i.e. cues are redundant regarding what they indicate). Thus, we assume that accuracy occurs even if restricted (sets of) cues are utilized for impression formation; for example, when judgements are made on the basis of reduced information (thin slices of OSN profiles, e.g. profile picture and group list). Second, we hypothesize that cues can be represented as both circumscribed behaviours (e.g. a colourful picture as a cue to infer conscientiousness) and thin-slice impressions (e.g. the conscientiousness rating based on an OSN user's profile picture as a cue for an overall impression of conscientiousness based on a user's full OSN profile). This idea is borrowed from the hierarchical judgement design where judgements made within one hierarchical layer lay the foundation for the cue values for judgements in the next layer (Cooksey, 1996; Hammond, Stewart, Brehmer and Steinmann, 1975). Finally, we can also apply the lens model approach towards understanding impression management and meta-accuracy as phenomena that are based on cue-expression and cue-perception processes.

Understanding accuracy

Accuracy (operationalized as self-other agreement; see Funder & Colvin, 1997) is the relation between perceiver impressions and the target's self. Whether, when and why accuracy in personality judgements emerges can be understood consulting the integrative lens model framework (Figure 1). Let us take the example of the Facebook user Tina (target). Her self-view as highly conscientious may express in observable OSN profile cues such as a professionally shot profile picture (cue validity; see C1 to Cn in Figure 1). Coming across Tina's profile picture (see information 1 in Figure 1), perceivers may base their conscientiousness impressions among others on the cue professionalism of her shoot (cue utilization), thereby accurately inferring that Tina scores rather high on conscientiousness. If perceivers have access to different kinds of information (thin slices; e.g. profile picture and interests field), they may base their overall personality judgement (‘Tina scores high on conscientiousness’) on their various thin-slice impressions (‘Tina's profile picture suggests she's a conscientious person’). To the extent that accurate thin-slice impressions (which are based on valid cues) are strongly utilized for overall impression formation, perceivers' overall judgement will be accurate.

image

Figure 1. Understanding the processes of accuracy, impression management and meta-accuracy. C1, C2, C3 and Cn refer to behavioural cues 1 to n.

Download figure to PowerPoint

Differences in the extent of accuracy between different personality traits (good trait) may arise from the existence or absence of valid cues on the part of the targets or the success or failure to utilize valid cues for impression formation concerning certain traits on the part of the perceivers. Similarly, differences in the extent of accuracy between different kinds of information (good information) and different trait × information combinations (diagnosticity) can be explained by the availability and correct utilization of cues.

Understanding impression management

Impression management is conceptualized as the relation of perceiver judgements and the target's desired impression (i.e. how targets wish to be viewed by others). Impression management can be decomposed into two distinguishable kinds of processes. First, when targets base their desired impression on their self-view (i.e. they wish to be viewed by others as they see themselves; self-verification theory; Swann, 1983),1 successful impression management might occur simply because of the convergence of self-views with others' impressions (accuracy, see earlier discussion). A second possibility is based on the same personality expression and impression formation processes outlined for understanding accuracy. Specifically, successful impression management might occur (i) when one's desired impression, which goes above and beyond one's self-view, leads to the expression of cues in favour of the desired trait value; (ii) when others indeed base their thin-slice judgements on these cues; and (iii) when perceivers' overall impressions are based mainly on this particular distorted thin-slice impression.

For example, Tina may aim at being viewed by other Facebook users as a very conscientious girl. Her overly desired impression (appearing higher on conscientiousness than she sees herself) may be related to observable cues in her Facebook profile, for instance, the level of detail in the hobbies section (cue validity; see C1 to Cn in Figure 1). Coming across Tina's interests field, perceivers may utilize this cue to infer her level of conscientiousness. To the extent that such distorted thin-slice impressions are strongly utilized for overall impression formation, perceivers' overall judgements will be distorted towards the target's desired impression. As for accuracy, the extent of impression management may vary across the traits, the kinds of information and the interplay of the two because of the differential expression and utilization of cues.

Understanding meta-accuracy

Meta-accuracy is the relation of perceiver judgements and the target's meta-perception (Figure 1). The extent of meta-accuracy might be attributed to three distinct kinds of processes. First, meta-perceptions might be based mainly on self-perceptions (cf. Kenny, 1994; Kenny & DePaulo, 1993). Targets have theories about what they are like and assume these to be obvious to perceivers. As other perceptions indeed often converge with self-perceptions (accuracy, see earlier discussions), meta-accuracy might result. Second, and independent of how targets actually view themselves, they might assume that they are successful in communicating their desired impression to others. When other perceptions are indeed influenced by a target's desired impression component (impression management, see earlier discussion), meta-accuracy results. Third, targets might observe their own behavioural cues and infer others' impressions on the basis of this information (cf. Bem, 1967, 1972). Cue-based meta-accuracy emerges when the thin-slice impressions that perceivers may have used for inferring the target's personality converge with the target's meta-perceptions because they are based on the same behavioural cues (Figure 1).

For example, Tina wonders how she comes across with her Facebook profile. Her assumption of how conscientious she is viewed by other Facebook users may be related to (i) her self-view (‘Others view me as conscientious as I am’) (ii) her desired impression (‘Others view me as conscientious as I want to come across’) and (iii) observable OSN profile cues. That is, Tina may infer others' perceptions of her conscientiousness among other cues from the number of academic groups she has joined, a cue that may also be utilized by perceivers to infer Tina's level of conscientiousness from her group list, leading to an accurate meta-perception. As for accuracy and impression management, the extent of meta-accuracy may vary across the traits, the kind of information and the interplay of the two. Cue-based processes again might account for these moderator effects.

In sum, borrowing from the lens model (Brunswik, 1956) and the Realistic Accuracy Model (Funder, 2012) as well as the hierarchical judgement design and the concept of vicarious functioning (Cooksey, 1996), our study integrates research on the three central interpersonal perception phenomena by leading them back to the same core cue expression and cue perception processes.

THE PRESENT RESEARCH

  1. Top of page
  2. Abstract
  3. ONLINE SOCIAL NETWORKS—A REALISTIC CONTEXT FOR UNDERSTANDING PERSONALITY EXPRESSION AND IMPRESSION FORMATION
  4. PRIOR RESEARCH ON PERSONALITY EXPRESSION AND IMPRESSION FORMATION AT ZERO ACQUAINTANCE
  5. THE PRESENT RESEARCH
  6. METHOD
  7. RESULTS
  8. DISCUSSION
  9. CONCLUSION
  10. ACKNOWLEDGEMENTS
  11. REFERENCES
  12. Supporting Information

On the basis of the described integrative lens model framework, we aimed to explore the underlying processes of accuracy, impression management and meta-accuracy in OSNs. First, we investigated the basic questions of interpersonal perception: Is there evidence for accuracy, impression management and meta-accuracy in OSNs? Second, we examined the moderating processes of interpersonal perception and aimed to specify when judgements are accurate, impression management is successful and meta-accuracy occurs. Third, we explored why judgements are accurate, targets' impression management succeeds and meta-accuracy emerges. We use OSN profiles to apply our integrative lens model framework, thereby illuminating the ‘whethers’ (strength), ‘whens’ (moderators) and ‘whys’ (mediators) of accuracy, impression management and meta-accuracy in OSNs. These phenomena are explored at zero acquaintance that is when there has not been any prior interaction between targets and perceivers.

First-impression research has focused mainly on the expression and perception of broad core traits, particularly the Big Five (e.g. Funder, 2012; Gosling, 2008; Kenny, 1994). However, in social contexts, surface traits that capture more specific adaptations to circumscribed contexts (see Asendorpf & van Aken, 2003; McAdams, 1995) play an important role as well. In social networks such as OSNs, which centre around social belongingness, this includes personality aspects such as self-esteem and need for popularity (Christofides, Muise, & Desmarais, 2009; Santor, Messervey, & Kusumakar, 2000). Thus, we examined all questions for a broader range of personality traits, namely the Big Five, self-esteem facets and need for popularity, and emphasize that the integrative approach is applicable to other personality traits as well.

METHOD

  1. Top of page
  2. Abstract
  3. ONLINE SOCIAL NETWORKS—A REALISTIC CONTEXT FOR UNDERSTANDING PERSONALITY EXPRESSION AND IMPRESSION FORMATION
  4. PRIOR RESEARCH ON PERSONALITY EXPRESSION AND IMPRESSION FORMATION AT ZERO ACQUAINTANCE
  5. THE PRESENT RESEARCH
  6. METHOD
  7. RESULTS
  8. DISCUSSION
  9. CONCLUSION
  10. ACKNOWLEDGEMENTS
  11. REFERENCES
  12. Supporting Information

We applied our analyses to a rich set of data from targets for whom OSN profiles and diverse self-report measures were available. The set of OSN profiles used in this study is a subset of the data that were analysed in Back, Stopfer, et al. (2010), but the analyses of the papers do not overlap.2

Participants

Targets

One hundred three individuals (86 women, 17 men) with an average age of 18.17 years (SD = 0.51) participated in the study. They were recruited using Germany-wide advertisements for an online study on personality measurement. As compensation for taking part in the study, participants received individual feedback on their personality scores.

Perceivers

Five men and five women (mean age = 24.20 years; SD = 3.36) took part in the study as overall perceivers (i.e. they viewed the full OSN profiles). Another 46 individuals participated as thin-slice perceivers (i.e. they viewed only a thin slice of the OSN profiles). Specifically, six men and six women (mean age = 22.92 years; SD = 2.68) viewed all targets' profile pictures, another six men and seven women (mean age = 24.20 years; SD = 2.12) viewed all targets' interests fields, another five men and five women (mean age = 24.80 years; SD = 1.69) viewed all targets' group lists, and another six men and five women (mean age = 22.91 years; SD = 2.84) viewed all targets' notice boards. All perceivers were students or alumni from different subjects (e.g. psychology, business studies and political science). They reported being considerably experienced in using OSNs.

Procedure

Collection of targets' self-ratings

Targets' self-ratings were collected before the topic of OSNs was introduced. This approach reduced bias because it decreased the probability that targets had answered the questions about their self-view with their OSN self-presentation in mind. Targets were sent a link to an online survey where they rated their desired impression in OSNs as well as their meta-perception on the basis of their OSN profiles.

Collection of the stimulus material

We used targets' OSN profiles from two German OSNs (SchuelerVZ and StudiVZ) that are very similar to Facebook. To ensure that targets did not alter their OSN profiles in response to their participation in the study, we saved their OSN profiles before the subject of OSNs was mentioned. Additionally, we checked the date of last profile update for each target, and no modification of the profile information in response to taking part in the study was found. In the online survey, targets explicitly agreed that their OSN profiles could be used for research purposes. When targets had joined both SchuelerVZ and StudiVZ, targets indicated their primarily used OSN, and the respective OSN profile was the subject of analyses. In sum, 103 OSN profiles (84 from SchuelerVZ: 82%; 19 from StudiVZ) comprised the stimuli in the present study and were stored on disks. In order to make the amount of accessible information manageable to the perceivers, only targets' main OSN profiles were saved.

The respective thin slice of the OSN profiles (profile picture, interests field, group list and notice board) was extracted from the full OSN profiles and saved on disks. The stimulus materials in the profile picture condition consisted of the target's profile picture in its original size combined with the target's sex. The stimulus material in the interests field condition was composed of the sections hobbies, clubs, favourite music, favourite books, favourite movies and favourite quote. Thirty-one targets had filled out all of the six sections, whereas four targets had completed only one section (mean number of completed sections = 4.67; SD = 1.34). In the group list condition, the stimulus material consisted of the targets' OSN group list including one to 97 groups (mean number of groups = 38.73, SD = 25.35). The stimulus material in the notice board condition was composed of the total number of notice board entries and the latest up to 10 entries. Those included the text created by the notice board writer as well as his or her profile picture, name, school/university and a time stamp. The number of visible notice board entries ranged from two to 10, averaging 9.76 (SD = 1.08).

Collection of perceiver ratings

After agreement to data protection regulations, each perceiver was handed a disk and instructed to provide personality ratings based solely on targets' OSN profiles or thin slices of OSN profiles for all 103 targets. Concretely, perceivers viewed (thin slices of) OSN profiles on a screen and simultaneously filled out paper and pencil or digital rating forms. To ensure external validity, perceivers were allowed to peruse (thin slices of) OSN profiles without time restrictions while forming impressions of targets. Rating forms were answered at home rather than in a standardized lab situation. Perceivers were not provided with any information about their perceiver colleagues and were instructed not to discuss their ratings with other individuals. The sequence in which the (thin slices of) OSN profiles were rated was counterbalanced across perceivers to rule out order effects. To warrant zero acquaintance between targets and perceivers, all perceivers were asked to rate their level of acquaintance with all targets on a scale ranging from 1 (not at all acquainted) to 6 (strongly acquainted). All overall perceivers indicated an acquaintance level of ‘1’ (i.e. no prior acquaintance); three thin-slice perceivers in the profile picture condition indicated knowing one target each (i.e. they answered ‘3’ on the rating scale). Those rating data were excluded from further analysis.

Measures

Targets' self-ratings

For personality ratings of targets' self (‘What kind of person are you?’), items were presented in a first person form (e.g. ‘I am outgoing, sociable’). For personality ratings of targets' desired impression (‘Imagine that users come across your OSN profile—how do you wish to be viewed by others on the basis of your OSN profile?’) and meta-perceptions (‘Imagine that users come across your OSN profile—how do you think you are viewed by others on the basis of your OSN profile?’),3 items were displayed in a third-person form (e.g. ‘This person is outgoing, sociable’).

Perceiver ratings

All perceiver items were presented in a third-person version (e.g. ‘This person is outgoing, sociable’). Ratings ranged from 1 (disagree strongly) to 6 (agree strongly) except for the Big Five (1 to 5).

Instruments

Big Five

Big Five ratings (neuroticism, extraversion, openness to experience, agreeableness and conscientiousness) were collected using the German version of the BFI-10 (Rammstedt & John, 2007), which is a 10-item version of the well-established Big Five Inventory (BFI; John, Donahue, & Kentle, 1991). The BFI-10 shows test–retest reliabilities above .70 across 6 weeks and convergent validity with the NEO Personality Inventory. We used the BFI-10 because it serves as an adequate personality measure in research settings with time restrictions. Perceivers in the present study provided personality judgements for 103 targets. Hence, it was reasonable to apply short scales. In addition, the BFI-10's short-phrase items were easily adaptable to the different rating versions required for the present purpose (i.e. desired impression, meta-perception and perceiver ratings). Thereby, it also allowed for optimal comparability between different kinds of self- and other-perception measures.

Self-esteem

Targets' self-ratings of self, desired impression and meta-perception as well as perceiver ratings were assessed for general self-esteem (‘I am/this person is self-confident’), performance-related self-esteem (‘I am/this person is confident with my/his/her school performance’) and physical attractiveness-related self-esteem (‘I/this person often feel/s attractive’). Items were extracted from the Multidimensional Self-Esteem Scale (Schütz & Sellin, 2006; an adaptation of the Multidimensional Self-Concept Scale; Fleming & Courtney, 1984).

Need for popularity

Targets' self-ratings of self, desired impression and meta-perception as well as perceiver ratings were assessed using two items each (‘It is important to me/him/her to be known by many people’ and ‘It is important to me/him/her to look attractive’).

Collection of online social network profile cues

Twenty-seven independent coders counted, rated and categorized all information available on the OSN profiles. The assessment of cues was inspired from the literature (e.g. Borkenau & Liebler, 1992; Buffardi & Campbell, 2008; Funder, Furr, & Colvin, 2000; Holland, 1973; Krämer & Winter, 2008; Rentfrow & Gosling, 2003; Stangl, 1991) and from previous work (Back et al., 2008; Back, Schmukle, & Egloff, 2009, Back et al., 2010; Küfner et al., 2010) and discussions in the research group. Cues were assessed from the molar level (e.g. rating of funniness) to the molecular level (e.g. counting of words). Conceptually overlapping cues (e.g. stylishness of hair and stylishness of dress) were coded by different pairs of coders in order to reduce biases. The sequence in which the cues were coded was counterbalanced across coders. All coders were trained before starting the coding. Coders were research assistants or undergraduates receiving monetary compensation or course credit for their participation in the study. Altogether, more than 300 cues were assessed. For cues with extremely skewed distributions, the logarithmized scores were used for all analyses (e.g. number of links to pictures ranged from 1 to 293, averaging 48.27; SD = 53.47; median = 36 across targets). Within each thin slice, single cues were combined into cue aggregates on the basis of theoretical and reliability considerations. Details on the single cues contained in each cue aggregate are provided in the supplemental online material.

Profile picture cues

Four female and two male coders assessed 59 profile picture cues by means of ratings on a scale from 1 (not at all) to 6 (very much; e.g. the person is smiling), categorizations (e.g. indoor photo: no/yes) and objective assessments (e.g. height × length of the profile picture in square millimetre). Profile picture cues were combined into seven cue aggregates with each composed of six to 12 single cues (αs averaged .65 across cue aggregates). Attractiveness of the person (assessed by two coders; α = .66) was used as a single cue.

Interests field cues

Two male and two female coders assessed 63 cues by means of ratings on a scale from 1 (not at all) to 6 (very much; e.g. the entries in the hobbies section are diverse), categorizations (e.g. the favourite quote is about fun) and objective measures (e.g. counting of the number of entries in each section). In addition, we objectively collected text-based information using Linguistic Inquiry and Word Count (LIWC; Pennebaker, Booth, & Francis, 2007; Pennebaker, Francis, & Booth, 2001). This text analysis program assesses linguistic aspects such as pronouns, prepositions, verbs and psychological processes (e.g. positive feelings). The German and English dictionaries were applied. We combined the interests field cues to five cue aggregates with each composed of seven to 18 single cues (αs averaged .78 across cue aggregates).

Group list cues

Five coders (four women) assessed 55 group list cues by ratings from 1 (not at all) to 6 (very much; e.g. the group list is provocative) and categorizations (e.g. the group is about appearance) with each group being assigned to one or more categories. Coders also listed the total number of groups. After the coding was complete, the number of groups in each category was related to the total number of groups per target in order to account for the varying number of groups between targets. Again, we used LIWC to objectively code text-based information. Furthermore, for each of the targets' groups, the StudiVZ system-generated group category was registered (e.g. fun and nonsense). Group list cues were combined into five cue aggregates with each composed of four to 15 single cues (αs averaged .78 across cue aggregates). Number of groups (counted), ratings of funniness (α = .84) and ratings of gender typicality (α = .90) were used as single cues.

Notice board cues

Six female coders assessed 32 notice board cues in three groups of two coders each. Coders in Groups 1 and 2 made assessments on the basis of the full notice board by categorizations referring to the content of the text (e.g. the entry is about making an appointment) and referring to the style of the text (e.g. the entry is self-ironic). No more information than the posted texts were shown to these coders (i.e. the writer's picture, name and school/university as well as the entry's time stamp were removed). Each text could be assigned to one or more categories. Coders in Group 3 counted the number of the writers' pictures fitting into the respective category (e.g. the writer is attractive; minimum: 0, maximum: 10) for each notice board. Again, we used LIWC to objectively code text-based information. Such information that was not assessable by LIWC (e.g. names) and other objective features (e.g. number of notice board entries, date of oldest/latest entry and number of entries from the opposite sex) were counted by another coder. Notice board cues were combined into five cue aggregates with each composed of two to 10 single cues (αs averaged .60 across cue aggregates). Activity of the writers, attractiveness of the writers and number of words/all entries/entries from the opposite sex (counted) were used as single cues.

Separate cues

Five female coders assessed the additional information displayed on OSN profiles. Ninety-two separate OSN profile cues were coded by ratings on a scale from 1 (not at all) to 6 (very much; e.g. the about me text is salacious), categorizations (e.g. the user's last name is abbreviated) and objective assessments (e.g. number of photo albums). All text-based information was coded using LIWC. Cues were combined into five cue aggregates with each being composed of five to 33 single cues (αs averaged .69 across cue aggregates). Early member, in a relationship, right-wing political orientation, type of school and number of links/photo albums/total friends/friends at the same university/at other universities (counted) were used as single cues.

RESULTS

  1. Top of page
  2. Abstract
  3. ONLINE SOCIAL NETWORKS—A REALISTIC CONTEXT FOR UNDERSTANDING PERSONALITY EXPRESSION AND IMPRESSION FORMATION
  4. PRIOR RESEARCH ON PERSONALITY EXPRESSION AND IMPRESSION FORMATION AT ZERO ACQUAINTANCE
  5. THE PRESENT RESEARCH
  6. METHOD
  7. RESULTS
  8. DISCUSSION
  9. CONCLUSION
  10. ACKNOWLEDGEMENTS
  11. REFERENCES
  12. Supporting Information

Accuracy

Question 1.1: Do perceivers form accurate personality impressions of online social network users, and do results vary across traits, kinds of information or both?

The first support for the accuracy of personality impressions based on full OSN profiles was provided by the agreement among overall perceivers, which was determined by computing the pairwise intraclass correlation (ICC; Shrout & Fleiss, 1979). Average-rater correlations (ICC (2, k)) ranged from .61 for neuroticism to .86 for need for popularity and averaged .78 across traits. Accuracy computed as self-other agreement (i.e. the correlation between aggregated overall perceiver ratings and the target's self-view) varied substantially across traits, averaging .25. Specifically, the strongest accuracy emerged for openness and attractiveness-related self-esteem, followed by extraversion and conscientiousness (Figure 2(A)). Accuracy was much weaker for judgements of need for popularity and global self-esteem. There was no accuracy for neuroticism judgements. These good trait findings are mirrored in the thin-slice results: There was accuracy for openness in all thin-slice conditions and no accuracy for neuroticism in any condition (Figure 2(A)).

image

Figure 2. Accuracy, impression management and meta-accuracy in OSNs. N = neuroticism, E = extraversion, O = openness, A = agreeableness, C = conscientiousness, SE_g = self-esteem, global, SE_p = self-esteem, performance, SE_a = self-esteem, attractiveness, Nfp = need for popularity. For each trait, the first column shows the correlation between targets' self-ratings and overall perceiver judgements (viewing the full OSN profiles). Columns 2 to 4 refer to the correlation between targets' self-ratings and thin-slice perceiver ratings (based on the profile picture, interests field, etc.). Accuracy (A) was determined by correlating aggregated perceiver ratings with targets' self-view. Impression management (B) was determined by the correlation between perceiver ratings and targets' residual scores from the regression of desired impression onto self-view (rres). Meta-accuracy (C) was determined by the correlation between perceiver ratings and targets' residual scores from the regression of meta-perception onto self-view and desired impression (rres). For optimal comparability, all correlations are Fisher's Z transformed. Significance is indicated as follows: *p < .05, + p < .10. Number of targets varies from 97 to 103.

Download figure to PowerPoint

Accuracy was further determined by the kind of information: Mean self-other agreement across personality traits was strongest for the profile picture (r = .22) and almost absent for the interests field (r = .09) and the notice board (r = .08). Interestingly, consensus was also low for the notice board (mean ICC (2, k) = .31 across traits), but for the interests field, judgements were highly consensual (mean ICC (2, k) = .80 across traits), despite being inaccurate. Finally, the extent of accuracy depended on the trait × information combination: Extraversion, agreeableness, and conscientiousness were judged accurately on the basis of the profile picture, self-esteem domains were judgeable on the basis of either the profile picture or the group list, and need for popularity could be detected on the basis of either the interests field or the notice board.

Question 1.2: What are the mediating processes of accurate personality impressions?

Why were personality impressions accurate? Users' OSN profiles displayed behavioural cues that were valid indicators of personality (e.g. creative profile picture indicating the level of openness; Figure 3), and thin-slice perceivers utilized targets' behavioural cues to form impressions of the targets' personalities (e.g. creative profile picture for judging openness; see column 2 of Table 1). In those cases in which perceivers utilized behavioural cues according to their validity (using valid cues such as cultural interests and creative picture and ignoring invalid cues such as number of groups), accuracy of thin-slice personality impressions emerged (e.g. openness impressions based on the profile picture were accurate; Figure 2(A)). An overview of cues related to targets' self-view and perceiver judgements, respectively, is presented in Figure 3 and Table 1 for all personality traits.

image

Figure 3. Correlations of targets' self-ratings and exemplary cue aggregates. N = neuroticism, E = extraversion, O = openness, A = agreeableness, C = conscientiousness, SE_g = self-esteem, global, SE_p = self-esteem, performance, SE_a = self-esteem, attractiveness, Nfp = need for popularity. PP = profile picture, IF = interests field, GL = group list, NB = notice board, S = separate OSN profile cue. The size of the respective field reflects the number of significant correlations between targets' self-ratings and cue aggregates. The dark field displays the correlation between cues and targets' self-view. The lighter field shows the correlation between cues and the residual scores of the regression of targets' desired impression onto targets' self-view. The lightest field displays the correlation between cues and the residual scores of the regression of targets' meta-perception onto targets' self-view and desired impression. Correlations are controlled for target sex and arranged according to the strengths of the correlation (highest correlation at the top of the respective field). Only significant correlations are presented (p < .05 and p < .10, two-tailed). R refers to the multiple Rs of the regression of self-ratings (self-view, residual scores of desired impression and residual scores of meta-perception) onto valid cues. Number of targets varies from 93 to 103 for the profile picture, from 85 to 95 for the interests field, from 95 to 103 for the group list, from 96 to 103 for the notice board, and from 41 to 103 for the separate OSN profile cues.

Download figure to PowerPoint

Table 1. Correlations of perceiver judgements and exemplary cue aggregates
 Kind of informationExemplary cue aggregatePerceiver judgement
OverallThin slice
  1. Note: Column 1 shows the correlation between cue aggregates and ratings of perceivers who viewed the full OSN profiles (overall perceiver judgements). Column 2 shows the correlation between cue aggregates and ratings of perceivers who viewed thin slices of OSN profiles (thin-slice perceiver judgements). Correlations are controlled for target sex. Significant correlations are in bold (p < .05) or bold and italicized (p < .10, two-tailed). Total CU (cue utilization) refers to the multiple Rs of the regression of overall perceiver judgements onto utilized cues. The numbers in parentheses denote the total number of utilized cues. Number of targets varies from 93 to 103 for the profile picture, from 85 to 95 for the interests field, from 95 to 103 for the group list, from 96 to 103 for the notice board, and from 41 to 103 for the separate OSN profile elements.

  2. N = neuroticism; E = extraversion; O = openness; A = agreeableness; C = conscientiousness; SE_g = self-esteem, global; SE_p = self-esteem, performance; SE_a = self-esteem, attractiveness; Nfp = need for popularity; PP = profile picture; IF = interests field; GL = group list; NB = notice board; S = separate OSN profile cue.

NSIn a relationship−.35 
SNumber of links−.31 
NBRelatedness to partner−.24 
GLPositive outlook −.30
NBSocial contact in real life −.19
Total CU  .88 (15) 
ESNumber of friends at other universities.62 
SPage size.47 
NBNumber of entries.46.29
NBRelatedness to partner.38.39
SIn a relationship.36 
PPDressed-up person.31.17
SParty.30 
PPRebellious picture.29.34
GLRealistic−.28−.48
GLNumber of groups.27.75
SEarly member.25 
 IFWell-structured style−.23−.21
IFAmount of information.21.42
GLPositive outlook.21.57
PPExpressive picture.18.27
IFDetailed information .45
Total CU  .90 (30) 
OIFCultural interests.41.58
SRight-wing political orientation−.40 
PPCreative picture.35.45
GLMedia.34.49
SCreative.33 
IFDetailed information.27.41
SIn a relationship−.21 
GLPositive outlook .34
NBFunny text .16
NBNumber of words .14
Total CU  .77 (12) 
APPFriendly expression.41.49
PPRebellious picture−.35−.53
SProvocative−.27 
 GLCriticism−.26−.22
IFFun −.24
NBFew activity of the writers −.17
Total CU  .78 (16) 
CGLRealistic.48.59
IFWell-structured style.44.38
PPRebellious picture−.33−.55
GLCriticism−.24−.55
PPAcademic context.21.34
IFDetailed information .35
IFAmount of information .25
NBRelatedness to flirting −.18
Total CU  .81 (20) 
SE_gSNumber of friends at other universities.66 
SNumber of links.53 
STotal number of friends.52 
GLSelf-promotion.40.69
PPAttractive person.38.52
 NBNumber of entries from the opposite sex.31.20
SNumber of photo albums.27 
IFAmount of information.24.50
IFDetailed information .53
Total CU  .90 (29) 
SE_pSProvocative−.46 
GLRealistic.39.43
IFWell-structured style.34.31
PPRebellious picture−.25−.51
SWell-behaved.22 
Total CU  .63 (10) 
SE_aPPAttractive person.66.68
PPDressed-up person.53.40
SIn a relationship.37 
GLSelf-promotion.36.74
SNumber of photo albums.27 
NBFunny text.18 
GLGender typical .61
GLNumber of groups .51
Total CU  .88 (20) 
NfpSTotal number of friends.76 
NBNumber of entries.58.26
PPAttractive person.47.65
PPDressed-up person.46.45
GLSelf-promotion.42.79
SNumber of photo albums.39 
NBRelatedness to partner.35.42
SParty.31 
GLPositive outlook.31.50
GLGender typical.27.76
NBSocial contact in real life.23.17
IFAmount of information .54
IFDetailed information .54
Total CU  .93 (24) 

To more formally test the outlined process model, we computed a multiple-mediator model using Mplus (Muthén & Muthén, 2010; for multiple-mediator models, see Kenny, Kashy, & Bolger, 1998; Preacher & Hayes, 2008).4 Openness judgements based on profile pictures served as an example. Specifically, a path model was defined in which a path from targets' openness to thin-slice perceivers' openness impressions was specified (i.e. the direct effect) and in which separate paths from targets' openness to each behavioural cue (including target sex) assessed for the profile picture and from each cue to the thin-slice perceiver impressions were defined (i.e. the indirect effects). Also, cues were allowed to covary with each other. The accuracy in judging openness on the basis of the profile picture (i.e. the total effect) was .36, p < .01 (Figure 2(A)). The total indirect effect of all cues amounted to .22, p < .01, whereas the direct effect amounted to .14, p = .09. The cue aggregate creative picture was the strongest mediator to account for the relation between targets' openness and the openness impressions of thin-slice perceivers (specific indirect effect: .10, p = .02).

Further, we simultaneously tested the relevance of single cues and thin-slice impressions for the formation of an overall personality judgement. For example, overall perceivers' openness impressions were strongly related to impressions on the basis of thin slices of OSN profiles (line 1 of Table 2). Because thin-slice openness impressions were based on valid cues (column 2 of Table 1), and thus, were accurate (Figure 2(A)), overall perceivers might have formed accurate impressions of the targets' openness (Figure 2(A)) by integrating their thin-slice impressions. Column 1 of Table 1 additionally shows correlations between all cue aggregates and the overall judgements.

Table 2. Intercorrelations of perceiver ratings based on full OSN profiles and based on thin slices of OSN profiles
 NEOACSE_gSE_pSE_aNfp
 PPIFGLNBPPIFGLNBPPIFGLNBPPIFGLNBPPIFGLNBPPIFGLNBPPIFGLNBPPIFGLNBPPIFGLNB
  1. Note: Bold correlations are significant (p < .05, two-tailed); bold and italicized correlations are marginally significant (p < .10; two-tailed). Total CU (cue utilization) refers to the multiple Rs of the regressions of overall perceiver judgements onto the four thin-slice perceiver judgements for each personality trait. Number of targets varies from 93 to 103.

  2. N = neuroticism; E = extraversion; O = openness; A = agreeableness; C = conscientiousness; SE_g = self-esteem, global; SE_p = self-esteem, performance; SE_a = self-esteem, attractiveness; Nfp = need for popularity; F = full OSN profile; PP = profile picture; IF = interests field; GL = group list; NB = notice board.

F.57.18.14.22.68.38.39.34.62.72.49.34.51.39.52.21.54.59.57.24.63.29.26.32.38.45.38.18.79.14.34.37.68.26.45.48
PP.17.04.18.15.06.26.38.32.30.18.25.20.24.25.25.10−.04.21.20.24.18.08.19.35.17.21.33
IF.20.18.21.24.36.14.33.08.33.18.08.03.10.13.22.13.30.21
GL.05.29.32.08.23−.00.16.17.25
Total CU.59.80.83.68.79.75.61.82.78

Using thin-slice and overall impressions, we again performed a mediation analysis with openness as an example. Specifically, we computed a multiple-step multiple-mediator model (Hayes, 2009) in which a direct path from targets' openness to overall perceivers' openness impressions, and separate paths from targets' openness to 10 behavioural cues were specified (Figure 4). Of these cues, eight were tied to a specific thin slice (e.g. profile picture), each with two cues related to one of four thin slices, and two aggregates reflected general OSN profile features (i.e. creative and right-wing political orientation). To reduce the computational complexity of the path analysis, not all cues were included in the model. The selection of cues was based on whether the respective cue was a valid indicator of targets' openness and/or whether it was used by perceivers for impression formation.

image

Figure 4. A path model representing the indirect effect of targets' self-view on accuracy for targets' openness via perceivers' utilization of separate OSN profile features (creative and right-wing political orientation) and thin-slice perceiver impressions (profile picture judgement, interests field judgement, etc.). The latter were in turn influenced by behavioural cues available in the respective thin slice (e.g. creative picture and unique posture). The value outside the parentheses represents the total effect of self-view on personality impressions prior to the inclusion of the mediating variables; the value in parentheses represents the direct effect. N = 103.

Download figure to PowerPoint

For the two separate OSN profile features, we specified paths from each feature to the overall openness impression. For the other cues, by contrast, we specified paths to the thin-slice impressions and from these impressions to the overall openness judgement. Target sex was included as a cue in all mediational analyses to assess the contribution of each behavioural cue independent of target sex. Also, the respective two thin-slice cues and target sex were allowed to be interrelated, and the two OSN profile features and the thin-slice perceiver impressions were allowed to covary. Finally, note that beyond these suggested paths, the model also included paths—not shown in Figure 4—from the targets' openness directly to the thin-slice impressions and from thin-slice cues directly to the overall openness impression.

The path model was fitted to the data using Mplus. The results showed that the effect of targets' openness on overall openness impressions was completely mediated by the perceivers' utilization of OSN profile cues and thin-slice impressions, direct effect: −.01, ns, total indirect effect: .41, p < .01, total effect: .41. Subsequent analyses showed that whereas the sum of the specific indirect effects of the paths (Figure 4) was significantly different from zero, .26, p < .01, this was not the case for the indirect effect not specified in the model, .15, p = .11. Thus, the association between openness and overall perceivers' openness impressions indeed seemed to be mediated by (i) the expression of valid behavioural openness cues, (ii) the utilization of these cues for thin-slice openness judgements and (iii) the utilization of these thin-slice judgements to form an overall openness impression (see Figure 4 for the full mediational model).

Impression management

Question 2.1: Do targets successfully engage in impression management, and are there differences between traits, kinds of information, or both?

In a first step, we correlated aggregated overall perceiver ratings and targets' desired impression for each personality trait. Results revealed significant correlations for all traits except global self-esteem, agreeableness and neuroticism, averaging .20 across traits.

One possible explanation for these initial impression management findings is that targets just wish to be seen by others as they see themselves (correlations between self-view and desired impression ranged from .20 for neuroticism to .50 for openness, averaging .36, SD = 0.13, across traits), particularly as there was actually a relation between targets' self-view and perceiver ratings (i.e. accuracy). Thus, in a second step, we explored whether the congruence of targets' self-view and desired impression accounted for impression management. We performed regression analyses of targets' desired impression on targets' self-view for each personality trait. The resulting residual scores (i.e. the targets' desired impression components that go beyond the targets' self-view) were saved as new variables and used in all subsequent analyses. We correlated aggregated perceiver ratings with the residual scores in order to investigate impression management that was not accounted for by self-view. The results, presented in Figure 2(B), provide little evidence for impression management above and beyond accurate self-portrayals for most traits (mean impression management correlation was .11 across traits). Unique impression management was, however, successful for need for popularity and attractiveness-related self-esteem. Similarly, in all thin slices besides profile pictures, evidence for unique impression management emerged for attractiveness-related self-esteem and need for popularity (Figure 2(B)). Thus, for these traits, impression management cannot be explained merely by the congruence of targets' self-view and desired impression. Other cue-related processes might play roles instead.

Impression management was further influenced by the kind of information: A comparison of thin slices across personality traits revealed that mean unique impression management was significant in the interests field (r = .12) but absent in the profile picture (r = .03). Finally, there were some combinations between the particular thin slices and the specific trait being judged that influenced the extent of impression management: Unique impression management was somewhat stronger for global self-esteem based on the interests field and for extraversion based on the group list. Moreover, although impression management for attractiveness-related self-esteem and need for popularity tended to be successful across kinds of information (especially for the group list), it was uniquely unsuccessful for the profile picture.

Question 2.2: What are the mediating processes of successful impression management?

Why was targets' impression management successful for certain traits in certain kinds of information? As can be seen in Figure 3, the users' desired impression component (i.e. how profile owners wish to be seen by other users above and beyond how they see themselves) is manifested in their OSN profiles in terms of behavioural cues (e.g. self-promotional groups indicating desired impression concerning need for popularity; Figure 3). Moreover, thin-slice perceivers utilized targets' behavioural cues to form impressions of the targets' personalities (e.g. self-promotion in the group list for inferring the level of need for popularity; column 2 of Table 1). In those cases in which thin-slice perceivers utilized behavioural cues that were indicators of the targets' overly desired impression, successful impression management emerged (e.g. need for popularity impressions based on the group list were distorted; Figure 2(B)). For all traits, an overview of cue aggregates related to targets' desired impression component is provided in Figure 3.

Testing this idea more formally, we again performed a multiple-mediator analysis and specified separate paths from targets' desired impression concerning need for popularity to the behavioural cues (e.g. self-promotion), including target sex, and from the cues to group list perceivers' impressions concerning need for popularity. Also, a direct path was specified from the targets' desired impression to the impressions of group list perceivers, and the cues were allowed to covary. The results showed that the direct effect concerning need for popularity was partially mediated by the behavioural cues. Specifically, the direct effect of the targets' desired impression concerning need for popularity was significant, .12, p = .03, as was the total indirect effect of the behavioural cues, .20, p = .02. Furthermore, the strongest specific indirect effect was found for the cue aggregate self-promotion, .09, p = .04.

As was outlined for the explanation of accuracy, perceivers' thin-slice impressions are thought to be integrated into an overall personality impression. For example, for judging need for popularity, overall perceiver ratings were strongly based on group list and notice board impressions (line 1 of Table 2). Those were influenced by successful impression management (Figure 2(B)), leading overall perceiver judgements of the targets' need for popularity to be distorted (Figure 2(B)), To examine these assumptions more formally, we again fit a multistep multiple-mediator model to the data using the example need for popularity (Figure 5). Similar to the model displayed in Figure 4, this model contained paths from the targets' desired need for popularity to nine behavioural cues (plus target sex). Eight of these cues were thin slice specific, and one was a general OSN profile feature (i.e. total number of friends). Whereas the overall need for popularity impression was regressed on the general OSN profile feature, we specified separate paths for the eight thin-slice-specific cues to the thin-slice impressions and from these to the perceivers' overall impressions. Furthermore, a direct path was included from the targets' desired impression to the impressions of the perceivers, and the cue aggregates as well as the thin-slice impressions were allowed to intercorrelate. Finally, the path model that we calculated included also paths—not shown in Figure 5—from the targets' desired impression concerning need for popularity directly to the thin-slice impressions and from thin-slice-specific cues directly to the overall need for popularity impression.

image

Figure 5. A path model representing the indirect effect of targets' desired impression component on impression management for targets' overly desired need for popularity via perceivers' utilization of a separate OSN profile feature (total number of friends) and thin-slice perceiver impressions (profile picture judgement, interests field judgement, etc.). The latter were in turn influenced by behavioural cues available in the respective thin slice (e.g. dressed-up person and attractive person). The value outside the parentheses represents the total effect of targets' desired impression component on personality impressions prior to the inclusion of the mediating variables; the value in parentheses represents the direct effect. N = 102.

Download figure to PowerPoint

The results of the mediational analysis revealed that the effect of targets' desired need for popularity on perceivers' overall need for popularity impressions was fully mediated by the OSN profile feature, the behavioural cues and the thin-slice impressions. The direct effect of overly desired need for popularity was .07, p = .15, the total indirect effect was .26, p < .01, and the total effect was .33, p < .01. Further analysis showed that the sum of the indirect effects specified in Figure 5 differed significantly from zero, .21, p < .01, but the sum of the other indirect effects not shown in the model did not, .06, p < .01. Thus, impression management for need for popularity was indeed mediated by (i) the expression of one's desired impression for need for popularity by means of observable cues, (ii) the utilization of these cues for thin-slice judgements and (iii) the use of these judgements as a basis for the overall need for popularity impression (see Figure 5 for the full mediational model).

Meta-accuracy

Question 3.1: Do targets know how they are viewed by others, and are there differences between traits, kinds of information, or both?

In a first step, we correlated aggregated overall perceiver ratings and targets' meta-perception. Results reveal significant correlations for all traits except neuroticism, averaging .31 across traits. These findings provide initial evidence for meta-accuracy as it is commonly operationalized in the literature (Carlson & Kenny, 2012; DePaulo et al., 1987).

Two possible explanations for these meta-accuracy findings hold that targets base their meta-perceptions on their self-theories (Kenny, 1994; Kenny & DePaulo, 1993). Specifically, they might assume that they are seen by others (i) as they see themselves (correlations between meta-perception and self-view ranged from .27 for general self-esteem to .66 for openness, averaging .41, SD = 0.14, across traits) or (ii) as they wish to be seen by others (correlations between meta-perception and desired impression ranged from .27 for neuroticism to .63 for need for popularity, averaging .52, SD = 0.16, across traits). Thus, we compared perceiver ratings with targets' meta-perception components that go beyond targets' self-view and desired impression. For each personality trait, regression equations were computed for predicting targets' meta-perception by targets' self-view and desired impression. The resulting residual scores were then correlated with aggregated perceiver judgements. By this procedure, targets' meta-perception components that were not accounted for by self-view and desired impression were isolated in their contribution to the prediction of perceiver ratings. The correlations averaged .20 across traits and are presented in Figure 2(C). Meta-accuracy for openness and conscientiousness seemed to be primarily due to the targets' use of self-views. By contrast, clear evidence for unique meta-accuracy was found for extraversion, agreeableness, global and domain-specific self-esteem, and need for popularity. Across thin slices of OSN profiles, unique meta-accuracy emerged only for extraversion (Figure 2(C)).

Meta-accuracy was further influenced by the kind of information: Mean unique meta-accuracy across personality traits was strongest for the group list (r = .14) and the interests field (r = .11) and lowest for the notice board (r = .03). Finally, the extent of meta-accuracy depended on the particular thin slice based on which a specific trait was judged: Unique meta-accuracy was stronger for agreeableness based on the interests field and for conscientiousness based on the group list. Moreover, meta-accuracy for need for popularity was substantial for the profile picture and group list but absent for the interests field and notice board. In sum, for certain traits based on certain kinds of information, targets knew how they were viewed by others, and their meta-perceptions were not explained by relying solely on how they viewed themselves or wished to be seen by others. Instead, further mechanisms need to be considered to explain meta-accuracy, particularly cue processes.

Question 3.2: What are the mediating processes of meta-accuracy?

Why were targets' meta-perceptions accurate? When forming meta-perceptions, targets may—in addition to utilizing self-theories—have used behavioural cues that were available in their OSN profiles to form impressions of how they were seen by others. For instance, when inferring how extraverted they are perceived, targets seem to not only rely on their self-theories but also to use OSN profile cues such as the detailedness of the information provided in the interests field (Figure 3), which was also utilized by the interests field perceivers to form an impression of the targets' level of extraversion (column 2 of Table 1). Thus, targets formed accurate perceptions of perceivers' extraversion impressions on the basis of the behavioural cues emerging in the interests field (Figure 2(C)). We computed a path analysis5 to test this assumption. A path model was fitted to the data in which perceiver impressions as well as targets' meta-perceptions were regressed on the behavioural cues (including target sex). The cue aggregates were allowed to intercorrelate as were perceivers' impressions and targets' meta-perceptions. This path model allowed for testing whether the correlation between the latter two variables was significantly reduced when behavioural cues were taken into account. We found that the original correlation between targets' meta-perceptions and perceiver impressions, r = .48, was reduced to r = .27 when considering the influence of the cues. Results of a bootstrapping analysis using 5000 bootstrapping samples showed that the difference of Δr = .21 was highly significant, z = 2.54, p < .01. Figure 3 provides an overview of cue aggregates related to targets' meta-perception component for all traits.

Targets may utilize cues available on the basis of different kinds of information for the formation of meta-perceptions. When perceivers' thin-slice judgements are based on the same cues and perceivers base their overall personality impressions on the thin-slice impressions (particularly on those where they used the same cues as the targets), targets' meta-perceptions will match overall perceiver impressions (i.e. will be correct). For example, regarding extraversion, targets and perceivers used similar cues with respect to all thin slices. As a result, targets were able to accurately infer how extraverted they seemed to be in all thin slices. Because overall perceiver impressions of targets' extraversion were strongly based on all thin-slice judgements (line 1 of Table 2), meta-accuracy emerged. For conscientiousness, by contrast, targets' meta-perceptions in the profile picture and the interests field were not accurate (Figure 2(C)). Overall perceiver judgements were strongly based on the profile picture and the interests field impressions (line 1 of Table 2), leading targets to be less accurate in inferring how conscientious they seemed to others in general (Figure 2(C)).

We conducted a path analysis to more formally test these assumptions using extraversion as an example. A path model was fitted to the data in which we tested whether meta-accuracy for extraversion was significantly reduced when the utilization of thin-slice impressions and behavioural cues were considered. In this model (Figure 6), targets' extraversion meta-perception component was regressed on the nine most important behavioural cues and target sex; eight of these cues were tied to a thin slice and one cue was a general OSN profile feature (i.e. number of friends at other universities). Furthermore, perceiver impressions were regressed on the OSN profile feature and the thin-slice impressions; the latter were themselves regressed on the respective two behavioural cues. Finally, thin-slice-specific cues were allowed to intercorrelate, as were the thin-slice impressions and the general OSN profile feature. As expected, the original correlation between targets' meta-perceptions of extraversion and perceivers' extraversion impressions of r = .35 was reduced to r = −.10. Bootstrapping analyses using 5000 bootstrap samples again showed that the reduction of Δr = .46 was highly significant, z = 8.20, p < .01.

image

Figure 6. A path model representing the relations between targets' meta-perception component, overall perceiver impressions, thin-slice perceiver impressions and behavioural cues for extraversion. The value outside the parentheses represents the correlation of perceiver impressions and targets' meta-perception component prior to the inclusion of further variables; the value in parentheses represents the correlation of the two measures after an OSN profile feature, behavioural cues and thin-slice perceiver impressions were included. N = 103.

Download figure to PowerPoint

Thus, cues accounted for the association between overall perceivers' extraversion impressions and targets' extraversion meta-perceptions. These results show that meta-accuracy for extraversion was indeed explained by (i) the targets' use of their own behavioural cues to infer others' perceptions of them, (ii) the utilization of the same cues by perceivers for their thin-slice judgements and (iii) the reliance on these thin-slice judgements for the overall extraversion impression (see Figure 6 for the full path model).

DISCUSSION

  1. Top of page
  2. Abstract
  3. ONLINE SOCIAL NETWORKS—A REALISTIC CONTEXT FOR UNDERSTANDING PERSONALITY EXPRESSION AND IMPRESSION FORMATION
  4. PRIOR RESEARCH ON PERSONALITY EXPRESSION AND IMPRESSION FORMATION AT ZERO ACQUAINTANCE
  5. THE PRESENT RESEARCH
  6. METHOD
  7. RESULTS
  8. DISCUSSION
  9. CONCLUSION
  10. ACKNOWLEDGEMENTS
  11. REFERENCES
  12. Supporting Information

Understanding accuracy, impression management and meta-accuracy in online social networks

This paper addressed the processes of personality expression and impression formation at zero acquaintance in OSNs. Specifically, we aimed to provide an understanding of whether, when and why people accurately judge others' personality on the basis of OSN profiles, manage the impressions they make on others and know how they are viewed by others. We (i) included broad (Big Five) as well as more contextualized personality traits (self-esteem facets and need for popularity); (ii) used natural instead of artificial OSN profiles of users as stimulus material; (iii) combined a naturalistic and experimental design by showing perceivers either the full OSN profiles or thin slices of the OSN profiles; (iv) measured the whole range of OSN perceptions including self-perceptions of one's self and desired impressions, meta-perceptions, and overall and thin-slice other perceptions; and (v) provided an exhaustive assessment of the multitude of behavioural cues included in OSN profiles by independent coders. This allowed for fine-grained investigations of the underlying processes of distinct but intertwined interpersonal perception phenomena based on an integrative lens model framework.

Moderating variables

Our results underscore the relatively strong accuracy that can be found in OSNs (e.g. Buffardi & Campbell, 2008; Waggoner, Smith, & Collins, 2009). People formed accurate impressions of four of the Big Five traits as well as self-esteem facets. These judgements were mostly aided by an individual's profile picture—except for openness, attractiveness-related self-esteem and need for popularity, which were also accurately judged on the basis of information that stressed text-based self-expression (interests fields, group lists and notice boards). Neuroticism was the only trait that was not accurately judged on the basis of neither kind of information (cf. Connelly, & Ones, 2010).

In line with prior empirical research (e.g. Vazire & Gosling, 2004), only slight evidence for impression management was found. However, concerning traits that are conceptually associated with a motivation for social admiration at zero acquaintance (need for popularity and attractiveness-related self-esteem; cf. desirability; John & Robins, 1993), OSN users were successful in communicating their desired impression. Pointing to a trait × information interaction, impression management for these good trait candidates was uniquely unsuccessful when perceiver judgements were based on pictures of the targets. In addition, impression management seemed to be somewhat stronger for judgements based on the description of the OSN users' interests, suggesting that the written expression of one's personality allowed for more successful impression management than other kinds of information.

For most personality traits, OSN users held accurate assumptions of how they were seen by others. Meta-accuracy was strongest for extraversion (cf. observability; Funder & Dobroth, 1987) and judgements based on group memberships, and a number of unique trait × information combinations resulted in elevated meta-accuracies. Compared with our results on accuracy, overall, meta-accuracy tended to be somewhat stronger. That is, targets' guesses of how they were viewed by others were even more accurate than others' guesses of targets' personalities.

Comparing good information moderators across the three phenomena of interpersonal perception bears the implications that (i) perceivers should focus on the users' profile pictures to catch an accurate personality impression; (ii) OSN users may try to be judged on the basis of their interests fields instead of their profile pictures in order to communicate their desired impression; and (iii) users' guesses of others' perceptions of them should better be based on their group memberships instead of their notice board entries in order to form accurate meta-perceptions.

Mediating variables

We unravelled the underlying cue expression and cue perception processes of accuracy, impression management and meta-accuracy in OSNs. In line with our process-related framework, personality impressions were accurate when (i) users' self-views expressed in observable OSN profile cues (e.g. quantifiable cues such as the number of friends and links; cues related to creative self-expression) and (ii) perceivers utilized those valid cues and ignored invalid cues for impression formation (e.g. focusing on cues related to creativity; profile picture cues).

Impression management was successful for certain traits based on certain kinds of information when (i) users expressed their unique desired impression (over and above self-view) in observable cues (e.g. cues related to self-praises) and (ii) perceivers utilized those distorted cues for impression formation (e.g. focusing on cues related to social admiration; interests field cues). However, most OSN profile cues that were influenced by targets' desired impression were ignored by perceivers, resulting in a general low extent of successful impression management.

The differentiated results concerning meta-accuracy were also explainable by distinct cue expression and cue perception processes. However, in line with self-theories of meta-perceptions (Kenny, 1994; Kenny & DePaulo, 1993), a certain amount of meta-accuracy in OSNs was due to the fact that users simply assumed they were seen by others as they see themselves. For example, meta-perceptions for openness and conscientiousness were no longer accurate when controlling for targets' self-views. For all other traits, however, meta-accuracy remained significant, even when controlling for people's self-view and additionally for their desired impression. Thus, at least for some traits, people indeed ‘seem to have some genuine insight into their reputation and do not achieve meta-accuracy only by capitalizing on the fact that others see them similarly to how they see themselves’ (Carlson et al., 2011, p. 831). Here, we were able to unravel the processes by which people achieve this form of meta-insight: utilizing such cues that were also used by perceivers for impression formation (see e.g. group list cues). Thus, OSN users seemed to be able to take the perspective of naïve perceivers and utilize OSN profile cues accordingly.

In sum, a number of specific cue-based personality expression and perception processes were revealed. This is important, because the discussed interpersonal perception phenomena could have resulted from mechanisms other than cue processes as well. Accuracy of impressions might simply be a result of perceivers' utilization of valid stereotypes (instead of perceivable cues; see, for example, Gosling et al., 2002). Impression management might arise from targets relying on their self-views only (if targets wish to be viewed by other OSN users as they view themselves); this means that there would not be cues related specifically to impression management. Also, meta-accuracy might simply result from targets' self-views or desired impressions; again, we would not find specific cues that targets utilized to infer how they are seen by others. Instead, our research showed that targets (i) expressed their self-view and desired impression in cues and (ii) targets and perceivers utilized perceivable cues for their judgements.

An integrative lens model framework

Although accuracy, impression management and meta-accuracy are strongly intertwined interpersonal perception phenomena, they have largely been investigated in an isolated way (cf. Kenny, 1994, for an exception). We aim to understand them as closely related phenomena. To this end, we combined concepts of different research traditions, including the description of mediating processes underlying accuracy (Borkenau & Liebler, 1992; Brunswik, 1956; Gosling et al., 2002), the hierarchical judgement design (Cooksey, 1996; Hammond et al., 1975), the idea of vicarious functioning (Brunswik, 1956), research on the effects of different thin slices of observable behaviour on accuracy (Ambady, Bernieri, & Richeson, 2000; Ambady & Rosenthal, 1992) and on other moderating factors (Funder, 2012), as well as self-theories of meta-perceptions (Kenny, 1994; Kenny & DePaulo, 1993), self-perception theory (Bem, 1972) and models of impression management (Leary & Kowalski, 1990; Paulhus, 1984; Schlenker, 1980).

In our integrative lens model framework, the degree of accuracy was conceptualized as the result of (i) the expression of valid behavioural cues in a given kind of information (thin slice of OSN profile), (ii) the utilization of these cues for thin-slice personality judgements and (iii) the formation of an overall personality impression based on an integration of all thin-slice judgements. Impression management (expression and perception of distorted cues; integration of distorted thin-slice impressions) and meta-accuracy (utilization of cues by both targets and perceivers) were theorized in a similar manner. We used a realistic setting—OSNs—as a research context that allowed for the simultaneous consideration of accuracy, impression management and meta-accuracy; a differentiated investigation of trait, kind of information and trait × information moderators; and an exhaustive analysis of the mediating behavioural expression and impression formation processes. An application of this lens model framework to offline environments seems highly welcome.

Beyond online social networks: Application to the offline context

Consider everyday interpersonal perceptions in an offline environment: For example, Ted (target) attends a friend's dinner party where he encounters a number of strangers. What Ted is like (self-view), how he wishes to be seen by other party guests (desired impression) and how he assumes he is viewed by the other guests (meta-perception) are more or less related to observable cues (e.g. his shyness expresses itself by behaving in an awkward manner). At the party, Ted might like to come across as an extraverted, laid-back type of guy, leading him to tell jokes from time to time. Moreover, his assumptions about others' extraversion impressions of him might be based on his observations of his own behaviour. Paula (perceiver), who did not know Ted before the party, has access to different kinds of information (e.g. Ted's physical appearance when she first sees him, his nonverbal expressions while he is talking with another guy, the political interests he outlines during this conversation and others' reactions when Ted tells a joke), which provide different amounts and qualities of these behavioural cues. On the basis of each informational slice, Paula forms an impression on the basis of the observable cues. When she first sees him entering the kitchen, for instance, his nervous expression and unfashionable clothing make her think of an introvert. Similarly, she gets the impression that he is not very sociable from observing his frozen expression during a conversation and other guests raising their eyebrows and leaving the conversation when he tries to tell a joke. Then again, his energetic and positive way of talking about his latest paintball experience give her a rather extraverted picture of Ted. When thinking about the party guests the next morning, Paula bases her overall impression of Ted (relatively introverted, unsociable) on the variety of her specific impressions of him (physical appearance, social interactions and interests), which are integrated to infer Ted's personality.

As this example demonstrates, the present approach can be transferred to face-to-face settings. Future research may apply the suggested lens model framework offline, even though it might be challenging to comprehensively code concrete behavioural cues in face-to-face interactions. We expect that our basic assumption—the explanation of all involved interpersonal perception phenomena by means of specific cue expression and cue utilization processes—replicate across online and offline environments. The concrete behavioural cues involved may, however, vary depending on the context being studied (Gosling, 2008).

Limitations and future prospects for the study of personality expression and impression formation at zero acquaintance

Broadening sample and context

Our target sample was composed of relatively young and mainly female OSN users. It will be important to replicate this line of research with a more representative sample regarding sex and age in order to generalize findings to the general population. Moreover, besides the good trait, good information and diagnosticity, further potential moderator variables such as the good target (e.g. low self-monitoring for being accurately judged vs. high self-monitoring for being successful at impression management; Schlenker, 1980) and the good judge (e.g. extraverts being better at judging others; Funder, 2012) should be included in future studies. Research may further look at the consistency of these interindividual differences in accuracy, impression management and meta-accuracy across traits, contexts, life domains and time as well as the role of the outlined cue expression and cue perception processes therein.

Broadening the analytical approach

Analysing data on interpersonal perception could also be accomplished by determining profile similarities: An OSN user's personality can be viewed as (i) similar to the personality profile of an average person across a series of traits (i.e. normative accuracy) or (ii) different from the average person on a particular trait (i.e. distinctive accuracy; Biesanz, 2010; Furr, 2008). Future research might reveal how this profile approach can be applied to the study of impression management and meta-accuracy.

It may be especially fruitful to additionally examine all outlined interpersonal perception phenomena and processes on a dyadic level by combining the process approach outlined here with componential approaches to interpersonal perceptions (Back, Baumert, et al., 2011; Back & Kenny, 2011; Kenny, 1994). Are people particularly good at judging the traits/impressions of a specific other person (dyadic accuracy/dyadic meta-accuracy; Carlson & Furr, 2009; Kenny, 1994; Kenny & DePaulo, 1993). Is impression management particularly successful in a dyad composed of a certain target and a certain perceiver? Besides, it would be interesting to apply the integrative lens model framework to other domains of interpersonal perceptions such as interpersonal attraction (Back, Penke, Schmukle, & Asendorpf, 2011; Back, et al., 2011; Back, Schmukle, & Egloff, 2011).

Zooming in on the cue integration processes

In our integrative lens model framework, the process of impression formation was theorized to proceed in several steps from lower hierarchy judgements (using behavioural cues to form thin-slice impressions) to higher hierarchy judgements (using thin-slice impressions for overall impression formation). Moreover, the substitutability of behavioural cues (vicarious functioning; see Cooksey, 1996) was assumed so that judgements based on different kinds of information could all potentially be accurate. We tested these assumptions using (i) several groups of thin-slice perceivers who formed impressions based on cues and (ii) a group of overall perceivers who were thought to integrate several thin-slice impressions and behavioural cues into an overall impression of the target's personality. Although this described paramorphic representation (i.e. a mathematical description of judgement; Hoffman, 1960) seems plausible and was empirically supported by the present data, an alternative process of impression formation can be imagined: Instead of forming multiple separate impressions of a person, each based on a certain kind of information, perceivers may accommodate their initial impression as soon as additional information becomes available. Testing this idea experimentally requires perceivers who are successively given more information about a target and asked to provide personality ratings at any stage of information provision. It is an open empirical question whether first presenting targets' profile pictures and then more OSN profile information results in different personality impressions as compared with first presenting information on the targets' interests fields and then other OSN profile information.

Longitudinal approaches

Conducting longitudinal studies may provide important insights into the development and interdependencies of interpersonal perceptions over time. Entering a social interaction, people strive to establish ‘who is who’. This identity negotiation process usually stabilizes identities but can also lead to identity change under specific circumstances (Swann & Bosson, 2008). Positive feedback from an interaction partner, for example, can encourage a person to internalize a new self-view (Bollich, Johannet, & Vazire, 2011; Jones, Gergen, & Davis, 1962). But does this lead to an adapted self-presentation (which expresses itself in concrete behavioural cues) and in turn cause an altered impression by perceivers? Thus, an extension of the present study design may incorporate multiple feedback loops between all target and perceiver variables—how do, in the long run, self-view, desired impression and meta-perception influence each other and the perceiver impressions dynamically? Can meta-perception be seen as a barometer to infer the success of one's desired impression presentation? And may targets then update their desired impression behaviours depending on their meta-perception? Likewise, research may gather more insights into the question of how interpersonal perception processes change over time. Biesanz, West, and Millevoi (2007) showed that perceivers first use stereotype information when forming impressions about others and later on distinctive information (cf. Kenny, 1994). Can these findings hold true for the formation of meta-perceptions as well? Finally, considering the present findings and the suggested future studies, it is of immense relevance to examine whether individual differences in self-expression and impression formation processes are predictive of intra- and interpersonal outcomes such as health, well-being, popularity, relationship satisfaction or professional success.

CONCLUSION

  1. Top of page
  2. Abstract
  3. ONLINE SOCIAL NETWORKS—A REALISTIC CONTEXT FOR UNDERSTANDING PERSONALITY EXPRESSION AND IMPRESSION FORMATION
  4. PRIOR RESEARCH ON PERSONALITY EXPRESSION AND IMPRESSION FORMATION AT ZERO ACQUAINTANCE
  5. THE PRESENT RESEARCH
  6. METHOD
  7. RESULTS
  8. DISCUSSION
  9. CONCLUSION
  10. ACKNOWLEDGEMENTS
  11. REFERENCES
  12. Supporting Information

This paper outlined an integrative lens model framework that considered moderating and mediating variables when investigating three complex intertwined interpersonal phenomena: accurately judging others' personalities (accuracy), successfully communicating one's desired impression (impression management) and knowing others' impressions of oneself (meta-accuracy) in OSNs. This seems to be a promising new way of looking at the underlying dynamics of the variety of interpersonal perception phenomena in general. We are looking forward to future work that will apply the integrative lens model framework in different offline and online environments and research contexts. To move back to the example from the beginning of this paper, the present study hopefully provided new insights into whether, when and why Ann accurately judged Amy's openness; Irem successfully conveyed her desired impression to others; and Megan knew how extraverted she seemed to others.

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. ONLINE SOCIAL NETWORKS—A REALISTIC CONTEXT FOR UNDERSTANDING PERSONALITY EXPRESSION AND IMPRESSION FORMATION
  4. PRIOR RESEARCH ON PERSONALITY EXPRESSION AND IMPRESSION FORMATION AT ZERO ACQUAINTANCE
  5. THE PRESENT RESEARCH
  6. METHOD
  7. RESULTS
  8. DISCUSSION
  9. CONCLUSION
  10. ACKNOWLEDGEMENTS
  11. REFERENCES
  12. Supporting Information

Preparation of the manuscript was supported by Research Unit Media Convergence of Johannes Gutenberg University Mainz. We would like to thank Kathrin Metzler, Tim Kuhlmann, Janine Müller and Anna Lena Weil for their help with data collection.

  1. 1

    We assume that targets might be but not necessarily are aware of engaging in impression management. The same assumption holds true for targets' congruence of self-view and desired impression.

  2. 2

    The two studies differ with regard to (i) the sample (the present analyses are based on the German subset of the full target sample used in Back, Stopfer, et al., 2010), (ii) the instrument for measuring the Big Five (in the present study, the BFI-10 was used for targets and perceivers to guarantee the comparability of all research questions) and (iii) the perceiver judgments (in addition to four new samples of thin-slice perceivers, the German subset of the full perceiver sample used in Back, Stopfer, et al., 2010, was used here). Furthermore, the analyses of the present paper do not overlap with those of Stopfer, Egloff, Nestler, and Back (2013).

  3. 3

    An anonymous reviewer was concerned about whether there was enough variance in the desired impression ratings. Thus, we compared the variances in the ratings of desired impression and meta-perception for each trait and did not find significant differences (cf. Wilcox, 1990; Method M, mean Spearman correlation across traits = 0.01, ns).

  4. 4

    Simulation studies have shown that the sampling distributions of the indirect effects are often not normal but skewed, particularly in small samples (Hayes, 2009; Mackinnon, Lockwood, & Williams, 2004). The statistical tests that are employed in all reported mediational analyses are therefore based on bootstrap standard errors (Efron & Tibshirani, 1993). These standard errors were always calculated on the basis of 5000 bootstrap samples (Preacher & Hayes, 2008).

  5. 5

    We did not compute a mediation model here as we assume that both targets and perceivers utilize behavioral cues: Targets use cues to infer perceiver impressions, and perceivers use cues to judge the targets' personality.

REFERENCES

  1. Top of page
  2. Abstract
  3. ONLINE SOCIAL NETWORKS—A REALISTIC CONTEXT FOR UNDERSTANDING PERSONALITY EXPRESSION AND IMPRESSION FORMATION
  4. PRIOR RESEARCH ON PERSONALITY EXPRESSION AND IMPRESSION FORMATION AT ZERO ACQUAINTANCE
  5. THE PRESENT RESEARCH
  6. METHOD
  7. RESULTS
  8. DISCUSSION
  9. CONCLUSION
  10. ACKNOWLEDGEMENTS
  11. REFERENCES
  12. Supporting Information
  • Albright, L., Kenny, D. A., & Malloy, T. E. (1988). Consensus in personality judgments at zero acquaintance. Journal of Personality and Social Psychology, 55, 387395.
  • Ambady, N., Bernieri, F. J., & Richeson, J. A. (2000). Toward a histology of social behavior: Judgmental accuracy from thin slices of the behavioral stream. Advances in Experimental Social Psychology, 32, 201271.
  • Ambady, N., & Rosenthal, R. (1992). Thin slices of expressive behavior as predictors of interpersonal consequences: A meta-analysis. Psychological Bulletin, 111, 256274.
  • Asendorpf, J. B., & van Aken, M. A. G. (2003). Personality-relationship transaction in adolescence: Core versus surface personality characteristics. Journal of Personality, 71, 629666.
  • Back, M. D., Baumert, A., Denissen, J. J. A., Hartung, F. M., Penke, L., Schmukle, S. C., Wrzus, C. (2011). PERSOC: A unified framework for understanding the dynamic interplay of personality and social relationships. European Journal of Personality, 25, 90107.
  • Back, M. D., & Kenny, D. A. (2010). The Social Relations Model: How to understand dyadic processes. Social and Personality Psychology Compass, 4, 855870.
  • Back, M. D., Penke, L., Schmukle, S. C., & Asendorpf, J. B. (2011). Knowing your own mate value: Sex-specific personality effects on the accuracy of expected mate choices. Psychological Science, 22, 984989.
  • Back, M. D., Penke, L., Schmukle, S. C., Sachse, K., Borkenau, P., & Asendorpf, J. B. (2011). Why mate choices are not as reciprocal as we assume: The role of personality, flirting, and physical attractiveness. European Journal of Personality, 25, 120132.
  • Back, M. D., Schmukle, S. C., & Egloff, B. (2008). How extraverted is honey.bunny77@hotmail.de? Inferring personality from e-mail addresses. Journal of Research in Personality, 42, 11161122.
  • Back, M. D., Schmukle, S. C., & Egloff, B. (2009). Predicting actual behavior from the explicit and implicit self-concept of personality. Journal of Personality and Social Psychology, 97, 533548.
  • Back, M. D., Schmukle, S. C., & Egloff, B. (2010). Why are narcissists so charming at first sight? Decoding the narcissismpopularity link at zero acquaintance. Journal of Personality and Social Psychology, 98, 132145.
  • Back, M. D., Schmukle, S. C., & Egloff, B. (2011). A closer look at first sight: Social relations lens model analyses of personality and interpersonal attraction at zero acquaintance. European Journal of Personality, 25, 225238.
  • Back, M. D., Stopfer, J. M., Vazire, S., Gaddis, S., Schmukle, S. C., Egloff, B., & Gosling, S. D. (2010). Facebook profiles reflect actual personality, not self-idealization. Psychological Science, 21, 372374.
  • Bem, D. J. (1967). Self-perception: An alternative interpretation of cognitive dissonance phenomena. Psychological Review, 74, 183200.
  • Bem, D. J. (1972). Self-perception theory. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 6, pp. 162). New York, NY: Academic Press.
  • Biesanz, J. C. (2010). The social accuracy model of interpersonal perception: Assessing individual differences in perceptive and expressive accuracy. Multivariate Behavioral Research, 45, 853885.
  • Biesanz, J. C., West, S. G., & Millevoi, A. (2007). What do you learn about someone over time? The relationship between length of acquaintance and consensus and self-other agreement in judgments of personality. Journal of Personality and Social Psychology, 92, 119135.
  • Bollich, K. L., Johannet, P. M., & Vazire, S. (2011). In search of our true selves: Feedback as a path to self-knowledge. Frontiers in Psychology, 2, 312.
  • Borkenau, P., & Liebler, A. (1992). Trait inferences: Sources of validity at zero acquaintance. Journal of Personality and Social Psychology, 62, 645657.
  • Borkenau, P., Mauer, N., Riemann, R., Spinath, F. M., & Angleitner, A. (2004). Thin slices of behavior as cues of personality and intelligence. Journal of Personality and Social Psychology, 86, 599614.
  • boyd, D. M., & Ellison, N. B. (2007). Social network sites: Definition, history, and scholarship. Journal of Computer-Mediated Communication, 13, 210230.
  • Brunswik, E. (1956). Perception and the representative design of experiments. Berkeley, CA: University of California Press.
  • Buffardi, L. E., & Campbell, W. K. (2008). Narcissism and social networking web sites. Personality and Social Psychology Bulletin, 34, 13031314.
  • Carlson, E. N., & Furr, R. M. (2009). Evidence of differential meta-accuracy: People understand the different impressions they make. Psychological Science, 20, 10331039.
    Direct Link:
  • Carlson, E. N., & Kenny, D. A. (2012). Do we know how others see us? In S. Vazire, & T. D. Wilson (Eds.), Handbook of self-knowledge (pp. 212257). New York, NY: Guilford Press.
  • Carlson, E. N., Vazire, S., & Furr, R. M. (2011). Meta-Insight: Do people really know how others see them? Journal of Personality and Social Psychology, 101, 831846.
  • Christofides, E., Muise, A., & Desmarais, S. (2009). Information disclosure and control on Facebook: Are they two sides of the same coin or two different processes? Cyberpsychology & Behavior, 12, 341345.
  • comScore (2011). It's a social world: Top 10 need-to-knows about social networking and where it's headed. Retrieved June 14, 2012, from http://www.comscore.com/Press_Events/Presentations_Whitepapers/2011/it_is_a_social_world_top_10_need-to-knows_about_social_networking.
  • Connelly, B. S., & Ones, D. S. (2010). An other perspective on personality: Meta-analytic integration of observers' accuracy and predictive validity. Psychological Bulletin, 136, 10921122.
  • Cooksey, R. W. (1996). Judgment analysis: Theory, methods, and applications. San Diego, CA: Academic Press.
  • DePaulo, B. M., Kenny, D. A., Hoover, C., Webb, W., & Oliver, P. V. (1987). Accuracy of person perception: Do people know what kind of impressions they convey? Journal of Personality and Social Psychology, 52, 303315.
  • DePaulo, B. M., Lanier, K., & Davis, T. (1983). Detecting the deceit of the motivated liar. Journal of Personality and Social Psychology, 45, 10961103.
  • Efron, B., & Tibshirani, R. J. (1993). An introduction to the bootstrap. New York, NY: Chapman & Hall/CRC.
  • Facebook (2013). Newsroom. Retrieved May 6, 2013, from http://newsroom.fb.com/Key-Facts.
  • Fleming, J. S., & Courtney, B. E. (1984). The dimensionality of self-esteem: II. Hierarchical facet model for revised measurement scales. Journal of Personality and Social Psychology, 46, 404421.
  • Funder, D. C. (2012). Accurate personality judgment. Current Directions in Psychological Science, 21, 177182.
  • Funder, D. C., & Colvin, C. R. (1997). Congruence of self and others' judgments of personality. In R. Hogan, J. Johnston, & S. Briggs (Eds.), Handbook of personality psychology (pp. 617647). San Diego, CA: Academic Press.
  • Funder, D. C., & Dobroth, K. M. (1987). Differences between traits: Properties associated with interjudge agreement. Journal of Personality and Social Psychology, 52, 409418.
  • Funder, D. C., Furr, R. M., & Colvin, C. R. (2000). The Riverside Behavioral Q-sort: A tool for the description of social behavior. Journal of Personality, 68, 451489.
  • Funder, D. C., & Sneed, C. D. (1993). Behavioral manifestations of personality: An ecological approach to judgmental accuracy. Journal of Personality and Social Psychology, 64, 479490.
  • Furr, M. R. (2008). A framework for profile similarity: Integrating similarity, normativeness, and distinctiveness. Journal of Personality, 76, 12671316.
  • Goffman, E. (1959). The presentation of self in everyday life. Garden City, NY: Doubleday/Anchor Books.
  • Gosling, S. D. (2008). Snoop: What your stuff says about you. New York, NY: Basic Books.
  • Gosling, S. D., Augustine, A. A., Vazire, S., Holtzman, N., & Gaddis, S. (2011). Manifestations of personality in online social networks: Self-reported Facebook-related behaviors and observable profile information. Cyberpsychology, Behavior and Social Networking, 14, 483488.
  • Gosling, S. D., Ko, S. J., Mannarelli, T., & Morris, M. E. (2002). A room with a cue: Personality judgments based on offices and bedrooms. Journal of Personality and Social Psychology, 82, 379398.
  • Graham, L. T., & Gosling, S. D. (2012). Impressions of World of Warcraft players' personalities based on their usernames: Interobserver consensus but no accuracy. Journal of Research in Personality, 46, 599603.
  • Hammond, K. R., Stewart, T. R., Brehmer, B., Steinmann, D. O. (1975). Social judgment theory. In M. Kaplan & S. Schwartz (Eds.), Human judgment and decision processes (pp. 271312). New York, NY: Academic Press.
  • Hayes, A. F. (2009). Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication Monographs, 76, 408420.
  • Hirschmüller, S., Egloff, B., Nestler, S., & Back, M. D. (2013). The dual lens model: A comprehensive framework for understanding self-other agreement of personality judgments at zero acquaintance. Journal of Personality and Social Psychology, 104, 335353.
  • Hoffman, P. J. (1960). The paramorphic representation of clinical judgment. Psychological Bulletin, 57, 116131.
  • Holland, J. L. (1973). Making vocational choices: A theory of careers. Englewood Cliffs, NJ: Prentice-Hall.
  • Holleran, S. E., & Mehl, M. R. (2008). Let me read your mind: Personality judgments based on a person's natural stream of thought. Journal of Research in Personality, 42, 747754.
  • Human, L. J., Biesanz, J. C., Parisotto, K. L., & Dunn, E. W. (2012). Your best self helps reveal your true self: Positive self-presentation leads to more accurate personality impressions. Social Psychological and Personality Science, 3, 2330.
  • Ivcevic, Z., & Ambady, N. (2013). Face to (Face)book: The two faces of social behavior? Journal of Personality, 81, 290301.
  • John, O. P., Donahue, E. M., & Kentle, R. (1991). The Big Five Inventory—Versions 4a and 54 (Technical Report). Berkeley: Institute of Personality and Social Research, University of California.
  • John, O. P., & Robins, R. W. (1993). Determinants of interjudge agreement on personality traits: The Big Five domains, observability, evaluativeness, and the unique perspective of the self. Journal of Personality, 61, 521551.
  • Jones, E. E., Gergen, K. J., & Davis, K. E. (1962). Some determinants of reactions to being approved or disapproved as a person. Psychological Monograph: General and Applied, 76, Whole No. 521.
  • Kenny, D. A. (1994). Interpersonal perception: A social relations analysis. New York, NY: Guilford Press.
  • Kenny, D. A., Albright, L., Malloy, T. E., & Kashy, D. A. (1994). Consensus in interpersonal perception: Acquaintance and the Big Five. Psychological Bulletin, 116, 245258.
  • Kenny, D. A., & DePaulo, B. M. (1993). Do people know how others view them? An empirical and theoretical account. Psychological Bulletin, 114, 145161.
  • Kenny, D. A., Kashy, D. A., & Bolger, N. (1998). Data analysis in social psychology. In D. Gilbert, S. Fiske & G. Lindzey (Eds.), The handbook of social psychology (4th ed., Vol. 1, pp. 233265). Boston, MA: McGraw-Hill.
  • Krämer, N. C., & Winter, S. (2008). Impression Management 2.0. The relationship of self-esteem, extraversion, self-efficacy and self-presentation within social networking sites. Journal of Media Psychology, 20, 106116.
  • Küfner, A. C. P., Back, M. D., Nestler, S., & Egloff, B. (2010). Tell me a story and I will tell you who you are! Lens model analyses of personality and creative writing. Journal of Research in Personality, 44, 427435.
  • Laing, R. D., Phillipson, H., & Lee, A. R. (1966). Interpersonal perception: A theory and a method of research. New York, NY: Springer.
  • Leary, M. R. (1995). Self-presentation: Impression management and interpersonal behavior. Madison, WI: Brown & Benchmark.
  • Leary, M. R., & Kowalski, R. M. (1990). Impression management: A literature review and two-component model. Psychological Bulletin, 107, 3447.
  • Leising, D., & Borkenau, P. (2011). Person perception, dispositional inferences, and social judgment. In L. M. Horowitz & S. Strack (Eds.), Handbook of interpersonal psychology: Theory, research, assessment, and therapeutic interventions (pp. 157170). Hoboken, NJ: John Wiley & Sons.
  • Levesque, M. J., & Kenny, D. A. (1993). Accuracy of behavioral predictions at zero acquaintance: A social relations analysis. Journal of Personality and Social Psychology, 65, 11781187.
  • Lewis, K., Kaufman, J., Gonzalez, M., Wimmer, A., & Christakis, N. (2008). Tastes, ties, and time: A new social network dataset using Facebook.com. Social Networks, 30, 330342.
  • Mackinnon, D. P., Lockwood, C. M., & Williams, J. (2004). Confidence limits for the indirect effect: Distribution of the product and resampling methods. Multivariate Behavioral Research, 39, 99128.
  • McAdams, D. P. (1995). What do we know when we know a person. Journal of Personality, 63, 365396.
  • Mehdizadeh, S. (2010). Self-presentation 2.0: Narcissism and self-esteem on Facebook. Cyberpsychology, Behavior and Social Networking, 13, 357364.
  • Murphy, N. A. (2007). Appearing smart: The impression management of intelligence, person perception accuracy, and behavior in social interaction. Personality and Social Psychology Bulletin, 33, 325339.
  • Muthén, L. K., & Muthén, B. O. (2010). Mplus user's guide (6th ed.). Los Angeles, CA: Muthén & Muthén.
  • Naumann, L. P., Vazire, S., Rentfrow, P. J., & Gosling, S. D. (2009). Personality judgments based on physical appearance. Personality and Social Psychology Bulletin, 35, 16611671.
  • Nestler, S., Egloff, B., Küfner, A. C. P., & Back, M. D. (2012). An integrative lens model approach to bias and accuracy in human inferences: Hindsight effects and knowledge updating in personality judgments. Journal of Personality and Social Psychology, 103, 698717.
  • Orne, M. T. (1962). On the social psychology of the psychological experiment: With particular reference to demand characteristics and their implications. American Psychologist, 17, 776783.
  • Paulhus, D. L. (1984). Two-component models of socially desirable responding. Journal of Personality and Social Psychology, 46, 598609.
  • Paulhus, D. L., & John, O. P. (1998). Egoistic and moralistic biases in self-perception: The interplay of self-deceptive styles with basic traits and motives. Journal of Personality, 66, 10251060.
  • Peluchette, J., & Karl, K. (2010). Examining students' intended image on Facebook: “What were they thinking?!” The Journal of Education for Business, 85, 3037.
  • Pennebaker, J. W., Booth, R. J., & Francis, M. E. (2007). Linguistic Inquiry and Word Count: LIWC 2007. Austin, TX: LIWC.
  • Pennebaker, J. W., Francis, M. E., & Booth, R. J. (2001). Linguistic Inquiry and Word Count: LIWC 2001. Mahwah, NJ: Erlbaum.
  • Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behavior Research Methods, 40, 879891.
  • Rammstedt, B., & John, O. P. (2007). Measuring personality in one minute or less: A 10-item short version of the Big Five Inventory in English and German. Journal of Research in Personality, 41, 203212.
  • Rentfrow, P. J., & Gosling, S. D. (2003). The do re mi's of everyday life: The structure and personality correlates of music preferences. Journal of Personality and Social Psychology, 84, 12361256.
  • Rentfrow, P. J., & Gosling, S. D. (2006). Message in a ballad: The role of music preferences in interpersonal perception. Psychological Science, 17, 236242.
    Direct Link:
  • Ross, C., Orr, E. S., Sisic, M., Arseneault, J. M., Simmering, M. G., & Orr, R. R. (2009). Personality and motivations associated with Facebook use. Computers in Human Behavior, 25, 578586.
  • Santor, D. A., Messervey, D., & Kusumakar, V. (2000). Measuring peer pressure, popularity, and conformity in adolescent boys and girls: Predicting school performance, sexual attitudes, and substance abuse. Journal of Youth and Adolescence, 29, 163182.
  • Schlenker, B. R. (1980). Impression management: The self-concept, social identity, and interpersonal relations. Monterey, CA: Brooks/Cole.
  • Schütz, A., & Sellin, I. (2006). Die multidimensionale Selbstwertskala (MSWS) [The multidimensional self-esteem scale]. Göttingen: Hogrefe.
  • Sherman, R. C., End, C., Kraan, E., Cole, A., Campbell, J., Klausner, J., & Birchmeier, Z. (2001). Metaperception in cyberspace. Cyberpsychology & Behavior, 4, 123129.
  • Shrout, P. E., & Fleiss, J. L. (1979). Intraclass correlations: Uses in assessing rater reliability. Psychological Bulletin, 86, 420428.
  • Stangl, W. (1991). Der Freizeit-Interessen-Test (FIT) [Leisure interests test]. Zeitschrift für Differentielle und Diagnostische Psychologie, 4, 231244.
  • Stecher, K. B., & Counts, S. (2008). Thin slices of online profile attributes. Proceedings of the International Conference on Weblogs and Social Media (pp. 1–9). Seattle, WA: Association for the Advancement of Artificial Intelligence.
  • Steele, F., Jr., Evans, D. C., & Green, R. K. (2009). Is your profile picture worth 1000 words? Photo characteristics associated with personality impression agreement. Proceedings of the International Conference on Webblogs and Social Media (pp. 327330). Seattle, WA: Association for the Advancement of Artificial Intelligence.
  • Stopfer, J. M, Egloff, B., Nestler, S., & Back, M. D. (2013). Being popular in online social networks: How agentic, communal, and creativity traits relate to judgments of status and liking. Journal of Research in Personality, 47, 592598.
  • Swann, W. B., Jr. (1983). Self-verification: Bringing social reality into harmony with the self. In J. Suls & A. G. Greenwald (Eds.), Social psychological perspectives on the self (Vol. 2, pp. 3366). Hillsdale, NJ: Erlbaum.
  • Swann, W. B., Jr. (1984). Quest for accuracy in person perception: A matter of pragmatics. Psychological Review, 91, 457477.
  • Swann, W. B., Jr., & Bosson, J. K. (2008). Identity negotiation: A theory of self and social interaction. In O. P. John, R. W. Robins & L. A. Pervin (Eds.), Handbook of personality: Theory and research (pp. 448471). New York, NY: Guilford Press.
  • Tong, S. T., Van Der Heide, B., Langwell, L., & Walther, J. B. (2008). Too much of a good thing? The relationship between number of friends and interpersonal impressions on Facebook. Journal of Computer-Mediated Communication, 13, 531549.
  • Valkenburg, P. M., & Peter, J. (2009). Social consequences of the internet for adolescents: A decade of research. Current Directions in Psychological Science, 18, 15.
    Direct Link:
  • Vazire, S., & Gosling, S. D. (2004). E-perceptions: Personality impressions based on personal websites. Journal of Personality and Social Psychology, 87, 123132.
  • Waggoner, A. S., Smith, E. R., & Collins, E. C. (2009). Person perception by active versus passive perceivers. Journal of Experimental Social Psychology, 45, 10281031.
  • Watson, D. (1989). Strangers' ratings of the five robust personality factors: Evidence of a surprising convergence with self-report. Journal of Personality and Social Psychology, 57, 120128.
  • Watts, D. J. (2007). A twenty-first century science. Nature, 445, 489.
  • Weisbuch, M., Ivcevic, Z., & Ambady, N. (2009). On being liked on the web and in the “real world”: Consistency in first impressions across personal webpages and spontaneous behavior. Journal of Experimental Social Psychology, 45, 573576.
  • Wilcox, R. R. (1990). Comparing the variances of two dependent groups. Journal of Educational Statistics, 15, 237247.
  • Wilson, R. E., Gosling, S. D., & Graham, L. T. (2012). A review of Facebook research in the social sciences. Perspectives on Psychological Science, 7, 203220.
  • Zywica, J., & Danowski, J. (2008). The faces of Facebookers: Investigating social enhancement and social compensation hypotheses; predicting Facebook and offline popularity from sociability and self-esteem, and mapping the meanings of popularity with semantic networks. Journal of Computer-Mediated Communication, 14, 134.

Supporting Information

  1. Top of page
  2. Abstract
  3. ONLINE SOCIAL NETWORKS—A REALISTIC CONTEXT FOR UNDERSTANDING PERSONALITY EXPRESSION AND IMPRESSION FORMATION
  4. PRIOR RESEARCH ON PERSONALITY EXPRESSION AND IMPRESSION FORMATION AT ZERO ACQUAINTANCE
  5. THE PRESENT RESEARCH
  6. METHOD
  7. RESULTS
  8. DISCUSSION
  9. CONCLUSION
  10. ACKNOWLEDGEMENTS
  11. REFERENCES
  12. Supporting Information
FilenameFormatSizeDescription
per1935-sup-0001-tablesS1-S5.docxWord 2007 document80KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.