Identity Processes in Face-to-Face and Digitally Mediated Environments

Abstract

In this study, we investigate how identity prominence influences individuals to use face-to-face or digitally mediated settings when communicating with significant others, and whether emotional reactions to nonverifying feedback regarding an identity vary depending upon the prominence of the identity and interaction setting. Results show that the more prominent an identity, the more likely individuals are to communicate with significant others related to the identity in face-to-face rather than digitally mediated settings. However, when controlling for identity prominence, we find that nonverifying feedback from significant others produces a similar a degree of negative emotions in both face-to-face and digitally mediated environments.

Keywords

identity prominence digitally mediated settings emotions

Over the past few decades, social interaction rituals have changed dramatically with the advancement of information technology. Interactions that traditionally occurred in face-to-face encounters are now occurring in digitally mediated environments: interaction settings where communication between individuals occurs via a computer, mobile phone, or similar electronic device. Indeed, the pervasiveness of communication technology in the digital age has created a new social reality—a “networked era” (boyd 2010) in which “networked individuals” (Rainie and Wellman 2012) interact (Davis 2016:137).

The advancement of communication technology raises questions about the nature of the self in the digital age. Some argue that digitally mediated environments allow individuals to present themselves differently compared with how they present themselves in face-to-face interactions, while others suggest that such environments simply provide an alternative venue for presenting the authentic self. This debate continues, but it raises additional questions for identity scholars: When communicating with others, does the nature of one’s identities influence one to choose either a face-to-face or a digitally mediated interaction setting? How do identity verification processes operate in each context? Similarly, differently, or at all? In this study, we apply identity theory and address these questions.

Identity and Social Setting

Over the years, identity scholars have produced a variety of theoretical and empirical studies that address how identity processes operate in different social contexts. Early theoretical work addressed how the self operates in specific roles and social situations (Foote 1951; Stone 1962). Others then examined how situations influence individuals to enact specific types of identities (Burke and Franzoi 1988); how individuals place varying levels of importance on identities at home, in their neighborhood, in public places, and at work (Jaret and Reitzes 1999); and how different elements of the environment shape subjective definitions of the self, self-evaluations, and social interactions (Deaux and Martin 2003). More recent research has addressed how proximate social structures (e.g., families, clubs, or social events that celebrate a subculture) operate as alternative contexts where individuals can enact and play out identities (Merolla et al. 2012; Stryker, Serpe, and Hunt 2005; Yarrison 2016), how total institutions influence identity processes (Asencio 2011), and how identities motivate behavior differently in individual versus group contexts (Carter 2013).

In this study, we add to aforementioned literature by investigating how identity processes relate to social context/setting in two ways. First, we examine the relationship between the prominence (importance) of an identity and one’s choice of interaction setting (digitally mediated settings as opposed to face-to-face settings) when interacting with others in the social structure. Second, we examine whether individuals’ emotional reactions to nonverifying feedback that pertains to an identity performance vary depending upon the interaction setting. We first review the literature on identity in the digital age and then discuss how identity theory can help us understand identity processes as they relate to face-to-face and digitally mediated social settings.

Identity in the Digital Age

The “virtual self” is defined as the configuration of the meaningful characteristics of a person in a digitally mediated space (Ruyter and Conroy 2002). An abundance of research has emerged in the past decades that examines self-processes in digitally mediated environments. Work in this area has examined how identities motivate social movement participation in real versus digitally mediated space (Diani 2000), how individuals with stigmatized and secret identities use Internet groups to find greater self-acceptance and strength to divulge an identity to others (McKenna and Bargh 1998), and how diasporic immigrants use the Internet to create a common identity and community in their new, alien culture (Wenjing 2005).

Together, research on identity in the digital age has generated a debate among scholars concerning the nature of real versus virtual “selves” (Suh 2013). One view sees the creation of the virtual self as a direct reflection of the real self (Schau and Gilly 2003); the other view focuses on how individuals intentionally create virtual selves to be different from one’s real-self counterpart (Kim and Sherman 2007; Turkle 1995). These perspectives have led to a need to understand the degree to which individuals’ presentation of self operates differently in face-to-face versus digitally mediated environments. Applying questions raised by these opposing perspectives to identity theory, two questions emerge: Do identities influence individuals to choose face-to-face settings over digitally mediated settings when they want to interact with others and verify an identity, and do identity verification processes operate similarly or differently (or at all) when individuals interact in face-to-face compared with digitally mediated settings? Identity theory can help answer these questions.

Identity Theory

Identities are “internal positional designations” (Stryker 1980), or meanings that define an individual as a unique person, role player, or group member. Identities characterize and classify individuals as similar or different to others in the social structure. Identity meanings are formed through a combination of shared cultural agreements and individual experiences related to a particular identity, all of which are all influenced by one’s social class, age, and racial/ethnic background (Stryker 1980; Stryker et al. 2005). Identity theory is a social psychological theory of the self that explains how identity processes influence behavior and emotions in social situations. The theory has developed as a cumulative research program over the past half century, with a variety of empirical studies supporting its tenets (Burke and Stets 2009; Stets and Serpe 2013).

Identity Verification

Identity theory addresses how identities motivate behavior in social settings, positing that individuals are motivated to behave among others in ways aligned with identity meanings to receive feedback that matches one’s identity standard (or how one sees oneself). Others in a social setting communicate this feedback via a presentation of significant gestures, which the self perceives as “reflected appraisals.” Reflected appraisals represent the self’s perceptions of others’ views with respect to how the self-in-situation is performing a particular identity.

As a form of significant feedback, reflected appraisals can be positive or negative. Identity theory posits that reflected appraisals are compared with identity standards (i.e., meaning structures that represent an identity) for congruence. When one perceives that congruence exists between self-in-situation meanings and identity standard meanings, identity verification occurs and positive emotions emerge. When self-in-situation meanings are discrepant from identity standard meanings, identity nonverification occurs, which signals an error that produces a negative emotional response. In short, identity theory predicts that when something occurs in a social situation that disconfirms one’s identity, they will feel bad (see Burke 1991 for a more detailed summary of the identity verification process). Identity verification has been examined in many empirical studies that employ an identity theory framework (Burke and Stets 1999; Burke, Stets, and Cerven 2007; Riley and Burke 1995; Stets and Cast 2007).

Identity Prominence and Salience

An individual’s multiple identities are hierarchically ranked according to prominence (importance) and salience (likelihood of being invoked; McCall and Simmons 1978; Stryker and Serpe 1994). Prominence is considered the subjective ranking of identities according to their relative worth or value to an individual, and has been likened to Morris Rosenberg’s (1979) concept of “psychological centrality” (Brenner, Serpe, and Stryker 2014; Ervin and Stryker 2001). Salience represents more of an objective ranking in the sense that it reflects the number of opportunities one has to engage in a particular identity (Stryker 1968; Stryker and Serpe 1982). In many cases, prominence and salience are highly correlated, for example, when a student who highly values being a student also engages the student identity each day by attending classes. However, the two are distinct concepts and may produce different identity rankings as in the case of a person who highly values being a student, but is not able to attend class because of other obligations such as work.

Recent research by Philip S. Brenner et al. (2014) has refined the distinction between prominence and salience by describing prominence as reflective of the aspirational or ideal self and salience as reflective of the actual self. Brenner and his colleagues closely examined the causal relationship between the prominence and salience of the “scientist” identity. To capture prominence, student participants responded to a series of four questions reflecting how important being a scientist was to them. In a separate measure of salience, participants reported how likely they were to tell different people about their desire to be a scientist. After conducting a series of structural equation models, the authors drew two important conclusions—that prominence and salience of identities remain relatively stable over time, and that greater prominence of an identity predicts greater salience of an identity (but not the reverse). The results of Brenner et al.’s study suggest the importance of subjective ranking of identities in predicting and guiding behavioral outcomes.

George J. McCall and J. L. Simmons (1978) suggested that identity prominence influences behavioral and emotional outcomes, and that identities that rank higher in prominence are more likely to guide behavior and emotions than lower ranked identities. According to McCall and Simmons, individuals are also more likely to seek out interaction partners and situations that allow for verification of more prominent identities than less prominent identities. Furthermore, Peter J. Burke and Jan E. Stets (2009) suggested that nonverification of identities that rank higher in prominence results in a more intense negative emotional response than nonverification of identities that rank lower in prominence. We return to these predictions later.

Identity Processes in Face-to-Face versus Digitally Mediated Environments

Past work in identity theory has established that role enactment and behavior is motivated by identities, but the relationship between identities and behavior has not previously been examined in the context of digitally mediated social environments. When an identity becomes activated and motivates an individual’s behavior, such behavior must be initiated and carried out in a particular social setting to receive reflected appraisals that indicate how successful one is at performing the identity. It is likely that some settings offer more opportunities for successful identity verification than others. For example, when seeking to verify a prominent identity, individuals may choose face-to-face social settings over digitally mediated settings because face-to-face settings offer more in terms of communicating nonverbal symbolic gestures. As Erving Goffman (1959) explained, the presentation of self to others takes place in specific social settings that are mediated by an exchange of language and significant gestures. When actors “perform” in these settings, they gauge the success of their performance by the “audience’s” response. Using the identity standard as a script, the performance is the behavior component of the identity process we described previously, and the audience response is the reflected appraisal component. Importantly, Goffman emphasized that the setting in which an interaction occurs is part of a performance as it also conveys symbolic meanings. He would remind us that the setting in which identity verification processes occur is as important to consider as actors’ performances themselves.

Previous studies in identity theory have generally examined social interactions in geographic locations where others are present (or assumed to be present). However, the theory does not posit that a social setting must be a physical locale. In the digital age, individuals have choices regarding the social setting they use to invoke identities and communicate with others. Nowadays, one can (often) choose whether to contact and communicate with others in some digitally mediated forum rather than in a face-to-face interaction. In other words, it is likely that identities are now, more than ever, enacted and verified in varying social contexts.

Digitally mediated environments provide a unique setting for identity formation and maintenance in that the exchange of symbols does not occur in the same physical space with interaction partners as it does in face-to-face interactions. As actors are not necessarily in the same space at the same time in a digitally mediated environment, it is possible that an actor’s performance is not received and responded to in the moment as it is during a face-to-face interaction. There could be a delay between actors that may lead to different kinds of performances, and subsequent differences in feedback from an audience. For example, an angry student who is dissatisfied with a grade and confronts a professor face-to-face right after class may emit a different performance than if the student sends an e-mail to the professor a few hours after class. The student’s method of delivery may very well influence the professor’s response (a reaction to an in-person confrontation may differ from a response to an e-mail inquiring about a grade). Using a digitally mediated communication method such as e-mail allows the audience to more carefully craft a response conveying the desired feedback than there is time for in a face-to-face communication setting. However, it does not guarantee that the recipient of the e-mail will have the desired response, or interpret the message as intended. In this way, e-mail may actually provide an actor with less control over the interpretation of the performance than face-to-face interaction as there is no opportunity for simultaneous performance adjustments as needed to insure interaction partners are interpreting the situation similarly. It is also true that the nature of digitally mediated communication makes it generally easier to “turn off” or exit an interaction that is not satisfying than it is to leave a face-to-face interaction setting. In some cases, it may be more desirable to have a “captive audience” in the sense that others witness one’s identity performance immediately, and feedback is in the moment; in other cases, the opposite might be true.

The fact that individuals now commonly interact—and enact identities—in different social contexts raises questions. How do individuals decide on how to communicate, that is, how do they choose the forum they use to enact identities and engage with others? Surely, physical proximity and convenience are factors. But is there more to it than that? Do people choose a type of social setting in the digital age depending on the identity that links themselves to those in which they seek to communicate? For example, if one has a highly prominent friend identity, are they more likely to communicate with friends in person (a face-to-face environment) than by text messaging (a digitally mediated environment)? If one has a highly prominent worker identity are they more likely to contact or communicate with their boss in person rather than over e-mail? Are students who have more prominent student identities more likely to contact their professors in person rather than using some electronic means of communication? These questions raise a general question regarding how identities motivate behavior in the digital age: Does the importance one places on an identity determine (or influence) the context one uses to engage in behavior related to the identity?

Identity Prominence and Identity Performance: Face-to-Face and Digitally Mediated Social Settings as Opportunity Structures

Earlier we noted that the prominence of an identity represents how central the identity is within the overall self-concept (McCall and Simmons 1978). Together, an individual’s hierarchy of prominent identities represents the self they aim to portray to others during interaction. The prominence of an identity is determined by several factors, including the amount of social support received for enacting/verifying the identity, the level of commitment to the identity, and the rewards gained by enacting/verifying the identity (McCall and Simmons 1978), rewards which often equate to positive emotions.

Included in determining the prominence of an identity is the degree to which one’s view of self is connected to the opinions of relevant others (e.g., one’s friends, parents, or boss; McCall and Simmons 1978:75). To receive support and rewards, the identity must be enacted and performed among others in a social situation; receiving a continual stream of support and rewards in the form of verifying feedback, or reflected appraisals, requires an identity to be performed continually over time. Individuals choose social settings that are most likely to provide support and rewards (in the form of verifying feedback from interaction partners associated with an identity), and avoid settings where receiving each is in question.

Contexts individuals use to enact identities aimed at receiving support and rewards are known as perceived opportunity structures (McCall and Simmons 1978:81). In considering what opportunity structure to use when enacting an identity, an individual appraises others in terms of what potential profit they offer. Thus, in an opportunity structure, others connected to an identity are potential sources of support for the identity that is in need of verification. In each encounter, individuals strive for social support of their identities, and they seek gratifications that emerge from successful role-performances as indicated by reflected appraisals consistent with the identity standard for the identity enacted. Thus, individuals seek to maximize rewards, or opportunities for verification in any given encounter.

Traditionally in identity theory, scholars addressed the process of role-performance leading to gaining support and rewards in settings where others were physically present. In these settings in which a prominent identity becomes enacted, participants in the encounter “bargain and haggle” over the kinds and quantities of rewards that are given between self and others (McCall and Simmons 1978:78). While digitally mediated social environments offer much regarding communicating with others who are physically distant, they have limitations regarding the manner in which information is transferred between individuals. The process of “bargaining and haggling” can be constrained. Even the most sophisticated electronic communication technology cannot emulate all facets of a face-to-face encounter. Past research has noted that up to 60 percent of the communication process during a face-to-face interaction is provided with body language (Mehrabian 1972). Many social cues are missed in digitally mediated communication, and the interpretation of gestures can be problematic, particularly if there are no visual cues. Individuals’ role-performances are further constrained in digitally mediated environments compared with face-to-face settings because identity performances are often separate in time and space from the audience response, increasing the chance of miscommunication or misunderstandings. Thus, despite the increased ability to control the setting in a digitally mediated environment (in terms of interruptions and distractions), as an opportunity structure, digitally mediated environments tend to be more limiting than face-to-face environments, making identity performances more difficult. Face-to-face environments allow for clearer communication, significant gestures, body language, and unspoken communication—all of which individuals (often) use when performing identities among others.

To recap, identity theory predicts that individuals choose opportunity structures that offer the best chance of receiving support and rewards from others associated with an identity. And, face-to-face settings provide advantages that digitally mediated settings lack. As face-to-face environments are opportunity structures that do not contain the same communicative limitations as digitally mediated environments (and hence provide a greater likelihood for obtaining support and rewards for identity enactment/verification), we offer an initial hypothesis regarding the relationship between identity prominence and face-to-face/digitally mediated social settings:

Hypothesis 1: The more prominent an identity, the more likely an individual will communicate with significant others related to the identity in a face-to-face social setting rather than a digitally mediated social setting.

Identity Performance, Nonverifying Feedback, and Negative Emotions in Face-to-Face and Digitally Mediated Environments

Research that examines how individuals respond to nonverifying feedback regarding an identity in different social contexts is lacking, though some does exist. Among the research that has addressed how one experiences disconfirming information in digitally mediated space and how it affects one’s sense of identity, some studies have applied self-discrepancy theory and regulatory focus theory. Self-discrepancy theory posits that discrepancies between one’s actual (experiential) and ideal/ought (aspirational) selves result in psychological distress (Higgins 1987, 1989). Chuan Hu, Li Zhao, and Jiao Huang (2015) used self-discrepancy theory to show that online environments allow individuals to reduce discrepancies between actual and ideal/ought selves. Self-discrepancy reduction in digitally mediated environments is accomplished by multiple ways: by creating a virtual self that more resembles one’s ideal self (which reduces the discrepancy between the actual and ideal self) and by acting anonymously toward others so as to deflect nonverifying feedback that would normally be attached to the actual self (reducing the discrepancy between the actual and ought self).

Self-regulatory theory posits that individuals are motivated to align their behaviors with their ideal-self and ought-self guides that are influenced by a promotion or prevention focus (Arazy and Gellatly 2012; Brockner and Higgins 2001; Scholer and Higgins 2008). Hu et al.’s (2015) work revealed that people reconstruct their identity in anonymous social network communities due to a desire to access new social networks (motivated by one’s promotion focus), or due to privacy concerns, a desire to escape from old networks, or to avoid being disturbed by others in the digitally mediated environment (all motivated by one’s prevention focus).

Nonverifying feedback regarding the performance of an identity has been studied extensively in identity theory (Burke 1991; Burke et al. 2007; Carter and Mireles 2016b; Stets and Asencio 2008), but there is no research that compares outcomes of nonverifying feedback related to an identity in face-to-face versus digitally mediated environments. Citing this lack of research, Jenny L. Davis (2016) recently provided a set of research questions regarding the manner in which digital media and digitally mediated space influences identity processes. Davis has encouraged identity theorists to conduct studies aimed at discovering the types of digitally mediated interactions where identity verification is more or less likely, how identity disconfirming information impacts the identity salience hierarchy, and the strategies social media users employ to avoid identity disconfirming inputs.

The second part of our study attempts to answer Davis’s call for more research that compares how outcomes of identity processes may differ in face-to-face and digitally mediated settings by focusing on how identity prominence influences negative emotions that emerge when a significant other provides nonverifying feedback related to an identity. In identity theory, negative emotions that emerge from nonverifying feedback are conceived to vary in intensity depending on the relative prominence, salience, and commitment of the identity; the more prominent, salient, or committed the identity, the greater the experience of negative emotions when the identity is not verified (Burke and Stets 2009:164). This occurs because the interruption of more important and salient processes within the self leads to greater autonomic activity, which is experienced as distress (Burke 1991; Mandler 1982). In addition, identity theory posits that interruption of the identity process (e.g., due to nonverifying feedback from others regarding one’s identity performance) causes greater distress when the source of the interruption is highly significant to the individual. Based on these theoretical predictions, we offer the following hypothesis regarding the relationship between identity prominence and negative emotions in face-to-face and digitally mediated social settings:

Hypothesis 2: Nonverifying feedback for more prominent identities will produce greater negative emotions than nonverifying feedback for less prominent identities in both interaction settings under study.

We have stated that identity theory posits that individuals will feel negative emotions when they experience identity nonverification. Nothing in identity theory posits that an individual should feel varying degrees of negative emotions when they experience identity nonverification in different social settings. However, relatively few studies have actually compared the degree of emotional intensity that emerges from identity nonverification in different social settings. Studies that have examined similar processes have examined how differences in the source of identity nonverification relates to emotions (Stets and Asencio 2008) and how the intensity of emotional response depends on the relevance of identity meanings in situations (Stets and Burke 2014). Our data allow us to compare emotional reactions in (two) different social settings, specifically if a component of the identity verification processes operates similarly or differently in face-to-face and digitally mediated environments. Based on the tenets of identity theory, we expect that a similar degree of emotional intensity will emerge in face-to-face and digitally mediated settings when an individual experiences similar nonverifying feedback related to an identity in each environment. We do not offer a formal hypothesis that predicts this, as this would in effect be predicting that we would find no difference between groups (which predicts the null hypothesis). Instead, we ask the following general research question regarding negative emotions that emerge from identity nonverification in different social settings:

Research Question 1 (RQ1): Do negative emotions that emerge from identity nonverification vary in intensity in different social settings?

Method

Procedure

A survey questionnaire was created and administered to study participants to measure the study variables. Data were gathered in 2016–2017 using Survey Monkey, an online survey website. The study was introduced to students in their respective college courses. Students were informed that the purpose of the study was to learn about how individuals see themselves, how they use technology and social media, and how they communicate with others. Students were told that they needed to be at least 18 years of age and that they would receive extra credit in their class for participating. The survey took approximately 20 min to complete. Proper guidelines for the treatment of human subjects were followed in administering the study.

Participants

Study participants were recruited from sociology courses at two universities (one in Northern California and one in Southern California). One thousand one hundred eighty individuals participated in the study (out of 1,411 who were contacted and invited to participate, for an 84 percent participation rate). One thousand one hundred fifteen (1,115) individuals completed all survey items, for a 94 percent completion rate. The average age of participants was 22 years (SD = 5.15). More females (70 percent) than males (30 percent) participated in the study. Participants identified their race as Hispanic/Latino (60 percent), white (19 percent), Asian (8 percent), and black/African American (6 percent). Seven percent identified as some “other” race. On average, participants reported working between 11 and 20 hr per week, and their average income ranged from $25,000 to $45,000 per year.

Measures

Identity prominence

Following others who have measured identity prominence in prior research (Burke and Reitzes 1991; Carter and Mireles 2016a; Stets and Biga 2003), we measured study participants’ degree of identity prominence for five role identities: the student, partner, son/daughter, friend, and worker identities. Study participants were asked, “How important is it to you to think of yourself as a student, partner, son/daughter, friend, and worker?” Each variable was measured on a scale from 1 (not at all) to 7 (very much).

Communication method with significant others

Five items measured whether study participants were more likely to communicate with significant others face-to-face or in digitally mediated space. Participants were asked to think about how they communicate with the following people in their life: their professors, their partner, their parents/legal guardians, their closest friends, and their boss. Participants were then asked to report whether they were more likely to communicate with each of these people in person (coded 0) or by using an electronic device (to text message, e-mail, etc.; coded 1).¹

Negative emotions that emerge from nonverifying feedback

To measure the degree of negative emotions that emerge in face-to-face versus digitally mediated environments when significant others who are associated with an identity provide nonverifying feedback, study participants were asked to think about the last time they experienced five different situations, with each situation being relevant to the five identities measured in the study (the student, partner, son/daughter, friend, and worker identities). Situations included a professor informing participants that they received a lower grade in a class than what was expected (corresponding to the student identity), a partner telling participants that they want to break up (corresponding to the boyfriend/girlfriend/partner identity), parents telling participants that they are disappointed with a choice they made (corresponding to the son/daughter identity), a friend not communicating with the participant as much as they would like (corresponding to the friend identity), and a boss telling participants that their coworkers are doing a better job than they are and that they need to work harder (corresponding to the worker identity). Participants were asked whether they had experienced (coded 2) or had not experienced (coded 1) each situation², and how distressed, sad, and angry they felt during each situation. Each emotion was measured from 1 (not at all distressed, sad, or angry) to 7 (very distressed, sad, or angry). Each situation’s three emotion variables were then added together to create five composite emotion variables, each representing the degree of negative emotions experienced in each situation, ranging from 3 (lowest degree of negative emotions) to 21 (highest degree of negative emotions). Each scenario was worded in two discrete ways: one way to indicate that the situation involved an experience in a face-to-face environment, and a different way to indicate that the situation involved an experience in digitally mediated space. All 10 scenarios and their wording for face-to-face versus digitally mediated contexts are presented in Table 1.

Table 1.

Face-to-Face and Digitally Mediated Scenarios.

Face-to-face scenarios

1. Your professor informs you in person that you received a lower grade in a class than what you expected.

2. Your boyfriend/girlfriend/partner tells you in person that they want to break up.

3. Your parents tell you in person that they are disappointed with a choice you made.

4. A friend you like a lot can’t seem to make time to spend with you.

5. Your boss at work tells you in person that your coworkers are doing a better job than you and that you need to work harder.

Digitally mediated scenarios

1. You check Moodle [an online course management system] and see that you received a lower grade in a class than what you expected.

2. Your boyfriend/girlfriend/partner texts you that they want to break up.

3. Your parents text you telling you they are disappointed with a choice you made.

4. A friend you like a lot doesn’t text you as much as you want them to.

5. Your boss contacts you by text or e-mail and tells you that your coworkers are doing a better job than you and that you need to work harder.

Control variables

Because social structural characteristics are important for determining the experience of identities (Stryker 1980; Stryker et al. 2005), we employ five social structural indicators as control variables in the analysis. These variables include age (measured in years), sex (0 = male, 1 = female), income (measured by asking, “Which of the following categories best describes your income last year, from all sources: 1 = less than $15,000; 2 = $15,001–$25,000; 3 = $25,001–$35,000; 4 = $35,001–$45,000; 5 = $45,001–$60,000; 6 = $60,001–$75,000; 7 = $75,001–$100,000; and 8 = more than $100,000), and hours working per week (measured on a 6-point scale where 1 = unemployed, 2 = 1–10 hr, 3 = 11–20 hr, 4 = 21–30 hr, 5 = 31–40 hr, and 6 = 41 or more hours). Race is measured using dummy variables, with “white” as the referent category (i.e., Asian = 1, white = 0; black = 1, white = 0; Latinx = 1, white = 0).

Results

Study data were analyzed using the Stata statistical software program. Table 2 presents the means and standard deviations for the study variables. In examining Table 2, we see that generally the identities examined in the study tend to be highly prominent identities, with the mean of each being higher than the midpoint of the scale. The most prominent identities are the student, son/daughter, and friend identities. The partner and worker identities are less prominent comparatively, but even these two identities are higher than the scale midpoint. We ran a series of t tests to determine whether the means for the identity prominence variables were significantly different from one another. The results showed that the means for the partner and worker identities were significantly different compared with all other identities (and each other). The mean for the partner identity (M = 5.35) was significantly lower compared with the student identity, M = 6.22, t(1,170) = −14.53, p ⩽ .001; son/daughter identity, M = 6.30, t(1,170) = −16.34, p ⩽ .001; friend identity, M = 6.26, t(1,170) = −15.92, p ⩽ .001; and worker identity, M = 5.78, t(1,170) = −7.31, p ⩽ .001. The mean for the worker identity (M = 5.78) was significantly lower compared with the student identity, M = 6.22, t(1,170) = −10.30, p ⩽ .001; son/daughter identity, M = 6.30, t(1,170) = −11.61, p ⩽ .001; and friend identity, M = 6.26, t(1,170) = −11.71, p ⩽ .001, and significantly higher compared with the partner identity, M = 5.35, t(1,170) = 7.31, p ⩽ .001. There were no significant differences among the means for the student, friend, and son/daughter identities (besides those mentioned above).

Table 2.

Means and Standard Deviations of Study Variables.

Variables	N	M	SD	Minimum	Maximum
Identity prominence variables
Student identity	1,171	6.22	1.18	1	7
Partner identity	1,171	5.35	1.92	1	7
Son/daughter identity	1,171	6.30	1.25	1	7
Friend identity	1,171	6.26	1.22	1	7
Worker identity	1,171	5.78	1.43	1	7
Communication method variables
Professor	1,108	0.47	0.50	0	1
Partner	820	0.24	0.43	0	1
Parents	1,105	0.26	0.44	0	1
Friends	1,092	0.49	0.50	0	1
Boss	969	0.22	0.41	0	1
Negative emotions in face-to-face scenarios
Low grade	361	15.22	4.62	3	21
Breakup	493	15.65	5.08	3	21
Disappointed parent	854	14.84	4.81	3	21
Friend communication	779	11.55	4.70	3	21
Boss reprimand	217	12.56	5.18	3	21
Negative emotions in digitally mediated scenarios
Low grade	918	15.46	4.62	3	21
Breakup	479	16.31	5.17	3	21
Disappointed parent	495	13.83	4.99	3	21
Friend communication	750	11.11	4.64	3	21
Boss reprimand	95	12.29	5.13	3	21
Income	1,115	4.14	2.28	1	8
Age	1,115	21.62	5.15	18	62
Sex	1,115	0.70	0.46	0	1
Hours working	1,115	2.89	1.52	1	6

Regarding the communication method variables, the means in Table 2 indicate that most participants communicate with a partner, parent, or boss in person (though approximately one fourth of the sample uses an electronic device when communicating with these three entities). On average, participants were divided equally when it comes to communicating with a professor or friend: Approximately half the sample communicates in person and the other half uses an electronic device.

Table 2 also presents descriptive statistics for the face-to-face scenario and digitally mediated scenario emotional reaction variables. In examining the means for these variables, we see that most participants reported feeling some degree of negative emotions when they experienced each situation, both in face-to-face and digitally mediated environments. To understand these results better, we ran a series of t tests to see whether the mean emotional reactions for each situation were significantly different in the face-to-face versus the digitally mediated scenarios. The results of the t tests show that, on average, participants experienced a greater degree of negative emotions in face-to-face settings when experiencing the disappointed parent, t(463) = 6.04, p ⩽ .001, and friend communication, t(643) = 4.53, p ⩽ .001, scenarios. Participants experienced a greater degree of negative emotions in the digitally mediated breakup scenario compared with the face-to-face breakup scenario, but this was statistically significant at the p ⩽ .10 level, t(310) = −1.94, p = .054. No significant differences in the degree of negative emotions felt were found between face-to-face and digitally mediated space regarding the low grade and boss reprimand scenarios.

Table 3 presents the results of five logistic regression equations that measure the effect of identity prominence on study participants’ communication method. We ran these equations to determine whether the prominence of study participants’ identities predicts how they communicate with significant others who are linked to them through an identity. The results in Table 3 are presented as odds ratios, where a departure from 1 in a positive direction equals an increase of the odds in communicating using an electronic device for a one-unit change in the independent variable. In examining Table 3, we see that generally, the more prominent one’s student, partner, son/daughter, or friend identity, the less likely they are to use an electronic device (and the more likely they are to communicate in person) when communicating with a professor (OR = .91, p ⩽ .10), partner (OR = .76, p ⩽ .001), parent (OR = .87, p ⩽ .05), or friend (OR = .85, p ⩽ .01), respectively. These results support Hypothesis 1. The degree of prominence for the worker identity had no effect on the communication method used when contacting a boss.

Table 3.

Logistic Regressions: Effect of Identity Prominence on Communication Method.

Independent variables	Dependent variables: Communication method
	Professor	Partner	Parents	Friends	Boss
	OR	OR	OR	OR	OR
Identity prominence
Student identity	0.91^†	—	—	—	—
Partner identity	—	0.76***	—	—	—
Son/daughter identity	—	—	0.87*	—	—
Friend identity	—	—	—	0.85**	—
Worker identity	—	—	—	—	0.93
Income	1.03	0.97	1.06^†	1.03	0.96
Age	1.00	0.97	1.01	1.03^†	1.00
Sex	1.36*	1.50*	1.45*	1.31^†	1.34
Hours working	1.03	0.93	1.07	0.98	0.99
Race: Asian	1.79*	1.23	1.02	1.45	1.50
Race: black	0.99	0.51	2.20**	1.14	1.36
Race: Latinx	1.53*	1.27	0.48***	1.74**	0.85
χ²	15.79*	39.39***	62.34***	26.32***	9.92
n	1,026	756	1,023	1,011	894

Note. OR = odds ratio.

†

p ⩽ .10. *p ⩽ .05. **p ⩽ .01. ***p ⩽ .001.

Table 3 also shows that sex is a predictor of communication method, with females more likely to use an electronic device to communicate with a professor (OR = 1.36, p ⩽ .05), partner (OR = 1.50, p ⩽ .05), parent (OR = 1.45, p ⩽ .05), or friend (OR = 1.31, p ⩽ .10). As with worker identity prominence, sex had no effect on communication method with a boss. Race was a significant predictor of communication method as well, with Asian (OR = 1.79, p ⩽ .05) and Latinx (OR = 1.53, p ⩽ .05) participants being more likely to communicate with a professor using an electronic device, black (OR = 2.20, p ⩽ .01) participants being more likely to communicate with a parent using an electronic device, and Latinx participants less likely to communicate with a parent using an electronic device (OR = .48, p ⩽ .001). Latinx participants were more likely to communicate with friends using an electronic device (OR = 1.74, p ⩽ .001).

Tables 4 and 5 present ordinary least squares regression analyses (standardized coefficients) for the effect of identity prominence on emotional reactions during situations where participants receive nonverifying feedback from significant others associated with an identity (Table 4 presents results for face-to-face scenarios, Table 5 presents results for digitally mediated scenarios). In first examining Table 4, we see that student, partner, and son/daughter identity prominence all predict a higher degree of negative emotions in the low grade (β = .14, p ⩽ .05), breakup (β = .16, p ⩽ .001), and disappointed parent (β = .13, p ⩽ .001) face-to-face scenarios, respectively. These results support Hypothesis 2. Table 4 also shows that females feel a higher degree of negative emotions in all five face-to-face scenarios compared with males (low grade: β = .13, p ⩽ .05; breakup: β = .25, p ⩽ .001; disappointed parent: β = .20, p ⩽ .001; friend communication: β = .13, p ⩽ .001; boss reprimand: β = .15, p ⩽ .05). Age relates to negative emotions in the boss reprimand scenario, with older participants experiencing more severe emotions (β = .22, p ⩽ .01). And, being Latinx (β = .10, p ⩽ .05) predicted a higher degree of negative emotions in the disappointed parent face-to-face scenario.

Table 4.

OLS Regressions for the Effect of Identity Prominence on Negative Emotions in Face-to-Face Environments.

Independent variables	Dependent variables: Degree of negative emotions
	Low grade scenario	Breakup scenario	Disappointed parent scenario	Friend communication scenario	Boss reprimand scenario
	β	β	β	β	β
Identity prominence
Student identity	.14*	—	—	—	—
Partner identity	—	.16***	—	—	—
Son/daughter identity	—	—	.13***	—	—
Friend identity	—	—	—	.05	—
Worker identity	—	—	—	—	.09
Income	.01	.03	.03	−.07	−.07
Age	.04	.01	.00	.01	.22**
Sex	.13*	.25***	.20***	.13***	.15*
Hours working	.02	−.08^†	−.06	.00	−.09
Race: Asian	.02	−.06	.05	−.02	.00
Race: black	.04	.00	−.03	−.03	−.01
Race: Latinx	.03	−.05	.10*	−.07	−.02
R ²	.05*	.11***	.08***	.03*	.09*
n	329	461	792	723	200

Note. OLS = ordinary least squares.

†

p ⩽ .10. *p ⩽ .05. **p ⩽ .01. ***p ⩽ .001.

Table 5.

OLS Regressions for the Effect of Identity Prominence on Negative Emotions in Digitally Mediated Environments.

Independent variables	Dependent variables: Degree of negative emotions
	Low-grade scenario	Breakup scenario	Disappointed parent scenario	Friend communication scenario	Boss reprimand scenario
	β	β	β	β	β
Identity prominence
Student identity	.10**	—	—	—	—
Partner identity	—	.15**	—	—	—
Son/daughter identity	—	—	.13**	—	—
Friend identity	—	—	—	.11**	—
Worker identity	—	—	—	—	.00
Income	−.05	.08	−.01	−.02	.09
Age	−.01	.03	.03	.03	.08
Sex	.16***	.19***	.19***	.06	.34**
Hours working	−.02	−.06	−.11*	−.08	.01
Race: Asian	−.02	−.03	.02	.02	.19
Race: black	.02	−.03	−.10*	−.03	.11
Race: Latinx	.02	−.04	−.02	.01	.30^†
R ²	.05***	.07***	.09***	.03*	.16
n	855	446	457	704	86

Note. OLS = ordinary least squares.

†

p ⩽ .10. *p ⩽ .05. **p ⩽ .01. ***p ⩽ .001.

In examining Table 5, we see that student, partner, son/daughter, and friend identity prominence all predict a higher degree of negative emotions in the low grade (β = .10, p ⩽ .01), breakup (β = .15, p ⩽ .01), disappointed parent (β = .13, p ⩽ .01), and friend communication (β = .11, p ⩽ .01) digitally mediated scenarios, respectively. These results also support Hypothesis 2. Table 5 also shows that females feel a higher degree of negative emotions in four of the five digitally mediated scenarios compared with males (low grade: β = .16, p ⩽ .001; breakup: β = .19, p ⩽ .001; disappointed parent: β = .19, p ⩽ .001; boss reprimand: β = .34, p ⩽ .01). Working more hours (β = −.11, p ⩽ .05) and being black (β = −.10, p ⩽ .05) predicted a lower degree of negative emotions in the disappointed parent digitally mediated scenario.

Our results reveal that, generally, identity prominence relates to the experience of negative emotions when individuals receive nonverifying feedback from significant others associated with an identity in both face-to-face and digitally mediated environments. However, the results presented in Tables 4 and 5 do not answer our research question (RQ1) because on their own they do not test for significant differences in the degree of negative emotions experienced in face-to-face versus digitally mediated environments. To test this and answer RQ1, we used the suest command in Stata, which allows the comparison of coefficients across the face-to-face and digitally mediated scenario regression models (i.e., the low-grade identity prominence coefficient in the face-to-face setting was compared with the low-grade identity prominence coefficient in digitally mediated space, the breakup identity prominence coefficient in the face-to-face setting was compared with the breakup identity prominence coefficient in digitally mediated space, and so on). The results showed no significant differences between the identity prominence and emotional reaction coefficients in any of the face-to-face and digitally mediated scenarios. This finding answers RQ1 and indicates that individuals feel a similar emotional reaction when they receive nonverifying feedback from significant others associated with an identity, regardless of whether it occurs in a face-to-face or digitally mediated environment.

Discussion

Our results indicate that importance of an identity may predict choices about communication context. For example, we find that participants in our study choose face-to-face communication over electronic communication more often when interacting with partners, parents, and bosses, than with professors and friends. The observation of differences in chosen communication method generally confirms our hypotheses, which suggest that the experiences of certain identities may make one method of communication preferable to others. Specifically, we find support for our first hypothesis, that identity prominence is a predictor of communication method. In our study, the more prominent an identity, the more likely a person chose face-to-face communication with another person relevant to that particular identity over electronic communication for the partner (to partner), son/daughter (to parent), and friend (to friend) identities.

Our findings with respect to identity prominence and interaction context offer support for identity theory’s predictions that individuals will choose to enact important identities in settings that offer the greatest opportunity structures for verification. However, these results should be interpreted with caution as our data do not allow for an empirical assessment of whether people consider the opportunity structure for identity verification in face-to-face interaction differently than that of electronic communication. Future research should assess the value people place on these different interaction settings. For example, face-to-face settings may allow for the exchange of more types of signs and symbols in the same physical space, but digitally mediated environments may be preferred because they offer more instantaneous, urgent communication with others.

Our results also show that prominence of the student and worker identities do not influence choices about communication method with professors and bosses. This could be due to the differences in identity standard meanings related to boundaries in communication between students and professors, and employees and bosses. Generally, there are more formal norms with respect to when and how individuals contact superiors than there are regarding when and how individuals contact more intimate interaction partners such as partners, parents, and friends. It could also be explained by the notion that the significant others we address in this study differ in terms of whether they are part of study participants’ primary or secondary groups. Primary groups are defined by strong emotional ties among members. Secondary groups are often task-oriented groups that are defined by loose emotional ties among members. It is likely that study participants feel more emotionally attached to their partner, parents, and friends (common primary group entities) than their professors or boss (common secondary group entities). So our findings suggest that identity prominence predicts communication method when one communicates with significant others who are part of one’s primary groups. Identity prominence seems less related to communication choice when one contacts secondary group members. In focusing on only five identities and their relationship with five discrete significant others, our study design is not robust enough to conclude with certainty that identity prominence influences behavior (and emotions) differently among others associated with primary versus secondary groups. However, our findings are compelling moving forward. Future work is needed to further address the role of substantive differences in the meanings of identities in determining choices about interaction settings and the relationship between identity processes and primary/secondary groups.

We also see interesting results regarding Hypothesis 1 with respect to sex. Past research has shown that females tend to use digital communication technology to connect with significant others while males use such technology to obtain information (Forgays, Hyman, and Schreiber 2014; Wei and Lo 2006). Our findings seem in line with these prior studies, with females being more likely than males to use an electronic device to communicate with significant others. These prior studies also suggest that females use communication technology to develop extensive social networks and enrichen intimate relationships. This also may explain the nonsignificant finding between sex and communication method regarding the boss scenario. Most people (including females) are probably not likely to seek to develop an intimate relationship with their boss; thus, we see no differences between males and females regarding one’s choice of method for communicating with a boss.

We also find support for our hypothesis regarding how identity prominence influences negative emotions when a significant other provides nonverifying feedback related to an identity (Hypothesis 2). We find that, for most of the identities in our study, nonverifying feedback from significant others produces a greater experience of negative emotions when identities are more prominent, in both face-to-face and digitally mediated environments. For example, identity prominence of the student, partner, and son/daughter identities increased negative emotional responses to nonverifying feedback in face-to-face environments, but identity prominence of the friend and worker identities did not demonstrate an influence. In the digitally mediated environment scenarios, identity prominence of the student, partner, son/daughter, and friend identities increased negative emotional reactions when experiencing nonverifying feedback from significant others, but prominence of the worker identity did not.

The results for emotional reactions to identity nonverification are also interesting regarding sex. Almost without exception (i.e., in all five face-to-face scenarios and in four of five digitally mediated scenarios), females reported feeling a greater degree of negative emotions across the regression models compared with males. This is not surprising, as prior research has shown that expressions of emotions often vary by status characteristics such as sex. For example. Jan E. Stets (2004) showed that, compared with males, females are more likely to report a greater degree of negative emotions when the source of identity nonverification is from a significant (familiar) other. Our findings show this as well.

Our study results also help answer our research question (RQ1), which addresses whether negative emotions that emerge from identity nonverification vary in intensity across social settings. In finding that individuals feel a similar degree of emotional intensity across social settings when their identities are not verified, we further identity theory. While a complete understanding of how individuals feel negative emotions when they experience identity nonverification requires additional research, our findings reveal a bit about how identity processes operate in different environments. It may be that where people experience identity nonverification is less important than simply experiencing identity nonverification. This is what our data suggest.

Study Limitations and Recommendations for Future Research

While the results presented here are useful in helping to understand how identities operate in face-to-face versus digitally mediated settings, there are limitations worth noting. First, the data used in this study are cross-sectional rather than longitudinal, so the relationships reported here are more correlational than causal. We have relied on identity theory to interpret the relationships found in our regression models, but any causality must ultimately be inferred.

An additional limitation regards the specific identities chosen for the study. While great care was taken to select identities that are relevant to the sample population, the results should only be generalized to those identities. However, the identities examined here have been shown to be highly committed, prominent, and salient in student populations in previous research (Burke and Stets 2009). Another limitation is that our study sampled college students whose ages, on average, were relatively young. So, the results should be generalized to the age range of the sample used in this study.

Our findings that identity prominence influences negative emotions similarly when a significant other provides nonverifying feedback related to an identity in face-to-face versus digitally mediated environments provide support for identity theory’s prediction that identity processes operate in the same manner across social contexts. However, the different findings for the different identities under study suggest the individual experience of different kinds of identities may produce variation in emotional outcomes of identity processes for different people. In particular, for this study, we see that the worker identity consistently produces a different outcome from the other identities. This could be reflective of the difference in substantive meaning of the worker identity compared with other identities under study, again calling for future research to attend to the importance of assessing the implications of meaning difference among identities. We chose to measure the worker identity because the universities sampled in the study enroll a large number of students who work. A recently conducted internal study on one of the universities in the sample revealed that 86 percent of students work while concurrently being enrolled, with 49 percent working 20 hours or more. For this particular study, the mean hours working for our sample is relatively low (between 11 and 20 hr/week), which may have affected the results with respect to the worker identity. And, as determined by the t tests reported previously, compared with all other identities (except the partner identity), the worker identity is significantly lower in prominence. Future research should sample a nonstudent population in an effort to generate more balance in the time spent enacting each identity under study.

Another consideration and limitation regards the direction of identity nonverification measured in this study. Here, we only address situations where individuals experience negative nonverifying feedback (e.g., a professor informing one of a low grade, parents telling a son/daughter they are disappointed with them, etc.). However, other possible scenarios exist where individuals could experience nonverifying information in a positive direction (e.g., a professor informing one that they received a higher grade than what was expected). Because we only address negative identity nonverification, we should acknowledge that a self-enhancement process may be behind individuals’ reports of distress (i.e., negative emotions) rather than a self-verification process. This is important to consider as past work in identity theory has found that—while counter to what is predicted in identity theory—in some situations nonverification in a positive direction can actually cause positive emotions (Stets 2005). This can occur for various reasons, including the nature of one’s relationship with the nonverifying source. When individuals interact with those they do not know well and those with whom they do not expect to have a long-term relationship (e.g., one’s boss or professor), they tend to seek enhancing feedback. With others who are well known (e.g., family members and friends), individuals tend to seek verifying feedback. So it is possible that the findings we report here might be different if the significant others we address in this study provided nonverifying feedback in a positive direction. Future research is needed to understand potential differences in positive versus negative identity nonverification as it pertains to this study, and whether the emotions reported in the scenarios addressed here are the result of an enhancement or verification motive.

An final caveat to our results is that we do not assess the type of electronic communication used to contact these particular interaction partners, so we do not know if there is a difference in the likelihood that a person would contact a professor, partner, parent, friend, or boss via e-mail, text, skype, or other electronic method over face-to-face communication. As each different mode of electronic communication method provides different opportunity structures in terms of identity processes, it is important for future studies to include information about which electronic method is employed to connect with different interaction partners to get a clearer understanding of how identity processes operate in digitally mediated space.

Conclusion

Our study contributes to the research in identity theory by confirming consistency in identity processes across social contexts. Specifically, we find that while variation in identity experience and meanings (prominence) influences choices about interaction contexts, identity processes generally operate consistently across contexts to the extent that nonverifying feedback about performance of certain identities produces negative emotions in both face-to-face and digitally mediated settings.

Our research indicates that differences in experiences or meanings of an identity are important when deciding whether to interact with others in a face-to-face or digitally mediated context. However, as we were not able to empirically assess whether there are qualitative differences in how people regard face-to-face and digitally mediated interaction environments with respect to opportunity structures, future research should include measures of participants’ perception of opportunities to enact identities in face-to-face and digitally mediated environments to better understand why people choose to engage some identities in context or the other.

We also did not measure identity standard meanings for the identities under study in our research. Knowing participants’ identity meanings for the identities examined here would assist in examining how different meanings associated with particular identities influence choice of interaction environments. Future work should address the potential for understanding how meaning differences of identities are important for choosing face-to-face or digitally mediated communication.

Our findings regarding the emotional outcomes of nonverifying feedback suggest the need for researchers to examine how differences in meanings affect outcomes of identity processes across social contexts. For example, identity prominence intensified negative emotional reactions to nonverifying feedback for some, but not all of the identities we examined. To conclude why the emotional reactions differ by identity, it is necessary to understand the different definitions associated with each identity.

Overall, the results of our study reveal that individuals prefer some contexts over others when playing out prominent identities, but that once in a specific context, the response to nonverifying feedback is the same as it would be in other contexts. Thus, our results confirm the assumption of identity theory that identity processes operate consistently across social situations. Of course, only two discrete contexts were examined here, and future research is needed to discover if this is true for all types of social contexts. This is needed to determine how meanings and experiences of identities affect outcomes of identity processes in different interaction environments. Specifically, our study suggests the need to more closely examine face-to-face versus digitally mediated environments to understand the implications of these different interaction contexts with respect to identity development and maintenance.

The results presented here advance recent theoretical contributions by Davis (2016) and others who have noted that identity scholars should begin to address how identities operate in the digital age. By comparing how identities operate in face-to-face versus digitally mediated environments in an empirical study, we answer this call, while also expanding on recent work that has addressed the general relationship between identities and social context. Of course, research in this vein will need to continue as technology develops and more sophisticated methods of digitally mediated communication emerge.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for research, authorship, and/or publication of this article: The research for this study was supported by a grant from the College of Social and Behavioral Sciences at California State University, Northridge.

Notes

Author Biographies

Michael J. Carter is an associate professor of Sociology at California State University, Northridge. His main research interests are in social psychology and microsociological theory, specifically the areas of self and identity. His research has examined moral identity processses, Deaf identity processes, identity change, and how identities operate in digital media environments.

Emily K. Asencio is an associate professor of Criminology and Criminal Justice Studies at Sonoma State University. Her main research interests are social psychology and criminology. Her primary research agenda focuses on the relationship between identity development and criminal behavior.

References

Arazy

Gellatly

I. R.

2012. “Corporate Wikis: The Effects of Owners’ Motivation and Behavior on Group Members’ Engagement.” Journal of Management Information Systems 29:87–116.

Asencio

Emily K.

2011. “Familiarity, Legitimation, and Frequency: The Influence of Others on the Criminal Self-View.” Sociological Inquiry 81:34–52.

boyd

danah

. 2010. “Social Network Sites as Networked Publics: Affordances, Dynamics, and Implications.” Pp. 39–58 in A Networked Self: Identity, Community, and Culture on Social Network Sites, edited by Papacharissi

New York: Routledge.

Brenner

Philip S.

Serpe

Richard T.

Stryker

Sheldon

. 2014. “The Causal Ordering of Prominence and Salience in Identity Theory: An Empirical Examination.” Social Psychology Quarterly 77:231–52.

Brockner

Higgins

E. Tory

. 2001. “Regulatory Focus Theory: Implications for the Study of Emotions at Work.” Organizational Behavior and Human Decision Processes 86:35–66.

Burke

Peter J.

1991. “Identity Processes and Social Stress.” American Sociological Review 56:836–49.

Burke

Peter J.

Franzoi

Stephen L.

1988. “Studying Situations and Identities Using Experiential Sampling Methodology.” American Sociological Review 53:559–68.

Burke

Peter J.

Reitzes

Donald C.

1991. “An Identity Theory Approach to Commitment.” Social Psychology Quarterly 54:239–51.

Burke

Peter J.

Stets

Jan E.

1999. “Trust and Commitment through Self-verification.” Social Psychology Quarterly 62:347–66.

10.

Burke

Peter J.

Stets

Jan E.

2009. Identity Theory. New York: Oxford University Press.

11.

Burke

Peter J.

Stets

Jan E.

Cerven

Christine

. 2007. “Gender, Legitimation, and Identity Verification in Groups.” Social Psychology Quarterly 70:27–42.

12.

Carter

Michael J.

2013. “Advancing Identity Theory: Examining the Relationship between Activated Identities and Behavior in Different Social Contexts.” Social Psychology Quarterly 76:203–23.

13.

Carter

Michael J.

Mireles

Danielle C.

2016a. “Deaf Identity and Depression.” Pp. 509–38 in New Directions in Identity Theory and Research, edited by Stets

J. E.

Serpe

R. T.

New York: Oxford University Press.

14.

Carter

Michael J.

Mireles

Danielle C.

2016b. “Examining the Relationship between Deaf Identity Verification Processes and Self-Esteem.” Identity: An International Journal of Theory and Research 16:102–14.

15.

Davis

Jenny L.

2016. “Identity Theory in a Digital Age.” Pp. 137–64 in New Directions in Identity Theory and Research, edited by Stets

J. E.

Serpe

R. T.

New York: Oxford University Press.

16.

Deaux

Kay

Martin

Daniela

. 2003. “Interpersonal Networks and Social Categories: Specifying Levels of Context in Identity Processes.” Social Psychology Quarterly 66:101–17.

17.

Diani

Mario

. 2000. “Social Movement Networks Virtual and Real.” Information, Communication and Society 3:386–401.

18.

Ervin

Laurie

Stryker

Sheldon

. 2001. “Theorizing the Relationship between Self-esteem and Identity.” Pp. 29–55 in Extending Self-esteem Theory and Research: Sociological and Psychological Currents, edited by Owens

T. J.

Stryker

Goodman

New York: Cambridge University Press.

19.

Foote

Nelson N.

1951. “Identification as the Basis for a Theory of Motivation.” American Sociological Review 26:14–21.

20.

Forgays

Deborah Kirby

Hyman

Ira

Schreiber

Jessie

. 2014. “Texting Everywhere for Everything: Gender and Age Differences in Cell Phone Etiquette and Use.” Computers in Human Behavior 31:314–21.

21.

Goffman

Erving

. 1959. The Presentation of Self in Everyday Life. New York: Doubleday.

22.

Higgins

E. Tory

. 1987. “Self-discrepancy: A Theory Relating Self and Affect.” Psychological Review 94:319–40.

23.

Higgins

E. Tory

. 1989. “Self-discrepancy Theory: What Patterns of Self-Beliefs Cause People to Suffer?” Advances in Experimental Social Psychology 22:93–136.

24.

Chuan

Zhao

Huang

Jiao

. 2015. “Achieving Self-congruency? Examining Why Individuals Reconstruct their Virtual Identity in Communities of Interest Established within Social Network Platforms.” Computers in Human Behavior 50:465–75.

25.

Jaret

Charles

Reitzes

Donald C.

1999. “The Importance of Racial-ethnic Identity and Social Setting for Blacks, Whites, and Multiracials.” Sociological Perspectives 42:711–37.

26.

Kim

H. S.

Sherman

D. K.

2007. “Express Yourself: Culture and the Effect of Self-expression on Choice.” Journal of Personality and Social Psychology 92:1–11.

27.

Mandler

George

. 1982. “Stress and Thought Processes.” Pp. 88–104 in Handbook of Stress: Theoretical and Clinical Aspects, edited by Goldberger

Breznitz

New York: The Free Press.

28.

McCall

George J.

Simmons

J. L.

1978. Identities and Interactions: An Examination of Human Associations in Everyday Life. New York: The Free Press.

29.

McKenna

Katelyn Y. A.

Bargh

John A.

1998. “Coming Out in the Age of the Internet: Identity ‘Demarginalization’ through Virtual Group Participation.” Journal of Personality and Social Psychology 75:681–94.

30.

Mehrabian

1972. Nonverbal Communication. New Brunswick, NJ: Aldine Transaction.

31.

Merolla

David M.

Serpe

Richard T.

Stryker

Sheldon

Schultz

P. Wesley

. 2012. “Structural Precursors to Identity Processess: The Role of Proximate Social Structures.” Social Psychology Quarterly 75:149–72.

32.

Rainie

Lee

Wellman

Barry

. 2012. Networked: The New Social Operating System. Cambridge, MA: MIT Press.

33.

Riley

Anna

Burke

Peter J.

1995. “Identities and Self-verification in the Small Group.” Social Psychology Quarterly 58:61–73.

34.

Rosenberg

Morris

. 1979. Conceiving the Self. New York: Basic Books.

35.

Ruyter

D. D.

Conroy

2002. “The Formation of Identity: The Importance of Ideals.” Oxford Review of Education 28:509–22.

36.

Schau

H. J.

Gilly

M. C.

2003. “We Are What We Post? Self-presentation in Personal Web Space.” Journal of Consumer Research 30:358–404.

37.

Scholer

A. A.

Higgins

E. Tory

. 2008. “Distinguishing Levels of Approach and Avoidance: An Analysis using Regulatory Focus Theory.” Pp. 489–504 in Handbook of Approach and Avoidance Motivation, edited by Elliott

A. J.

New York: Psychology Press.

38.

Stets

Jan E.

2004. “Emotions in Identity Theory: The Effect of Status.” Pp. 51–76 in Advances in Group Processes, vol. 21. Greenwich, CT: JAI Press.

39.

Stets

Jan E.

2005. “Examining Emotions in Identity Theory.” Social Psychology Quarterly 68:39–56.

40.

Stets

Jan E.

Asencio

Emily K.

2008. “Consistency and Enhancement Processes in Understanding Emotions.” Social Forces 86:1055–78.

41.

Stets

Jan E.

Biga

Chris F.

2003. “Bringing Identity Theory into Environmental Sociology.” Sociological Theory 21:398–423.

42.

Stets

Jan E.

Burke

Peter J.

2014. “Emotions and Identity Nonverification.” Social Psychology Quarterly 77:387–410.

43.

Stets

Jan E.

Cast

Alicia D.

2007. “Resources and Identity Verification from an Identity Theory Perspective.” Sociological Perspectives 50:517–43.

44.

Stets

Jan E.

Serpe

Richard T.

2013. “Identity Theory.” Pp. 31–60 in Handbook of Social Psychology, edited by DeLamater

J. D.

Ward

New York: Springer.

45.

Stone

Gregory P.

1962. “Appearance and the Self.” Pp. 86–118 in Human Behavior and Social Processes, edited by Rose

Boston: Houghton Mifflin.

46.

Stryker

Sheldon

. 1968. “Identity Salience and Role Performance.” Journal of Marriage and the Family 4:558–64.

47.

Stryker

Sheldon

. 1980. Symbolic Interactionism: A Social Structural Version. Menlo Park, CA: Benjamin Cummings.

48.

Stryker

Sheldon

Serpe

Richard T.

1982. “Commitment, Identity Salience, and Role Behavior: A Theory and Research Example.” Pp. 199–218 in Personality, Roles, and Social Behavior, edited by Ickes

Knowles

E. S.

New York: Springer-Verlag.

49.

Stryker

Sheldon

Serpe

Richard T.

1994. “Identity Salience and Psychological Centrality: Equivalent, Overlapping, or Complementary Concepts?” Social Psychology Quarterly 57:16–35.

50.

Stryker

Sheldon

Serpe

Richard T.

Hunt

Matthew O.

2005. “Making Good on a Promise: The Impact of Larger Social Structures on Commitments.” Advances in Group Processes 22:93–123.

51.

Suh

Ayoung

. 2013. “The Influence of Self-discrepancy between the Virtual and Real Selves in Virtual Communities.” Computers in Human Behavior 29:246–56.

52.

Turkle

Shirley

. 1995. Life on the Screen: Identity in the Age of the Internet. New York: Simon and Schuster.

53.

Wei

Ran

Ven-Hwei

. 2006. “Staying Connected While on the Move: Cell Phone Use and Social Connectedness.” New Media and Society 8:53–72.

54.

Wenjing

Xie

. 2005. “Virtual Space, Real Identity: Exploring Cultural Identity of Chinese Diaspora in Virtual Community.” Telematics and Informatics 22:395–404.

55.

Yarrison

Fritz W.

2016. “Contextualizing Proximate Social Structure in Identity Theory.” Pp. 343–65 in New Directions in Identity Theory and Research, edited by Stets

J. E.

Serpe

R. T.

New York: Oxford University Press.