Introduction

The cultural-normative perspective (Ekman 1973; Gordon, 1981, 1990; Hochschild 1975; [1983] 2012) and the perceptual control emphasis in identity theory (Burke 1991, 1996; Burke and Stets 2009; Stryker 2004; Stets 2014) offer a rich foundation for theoretical and empirical exploration. Cultural-normative perspectives within the sociology of emotions emphasize the centrality of culture in specifying emotion ideologies, emotional standards (Stearns and Stearns 1985), vocabularies of emotions, feeling rules and display rules (Hochschild 1979). These aspects of culture function as standards of behavior, they guide how emotions ought to be expressed within a situation and how or what one should feel and when (Wilkins and Pace 2014). They provide best practices and repertoires to be called upon in adjusting emotional responses to social conditions. Emotions are behavioral displays managed in social situations within identity. Emotion research within identity theory has largely been within the perceptual control emphasis (Burke and Stets 2009, p. 155; Stets and Serpe 2016). For those working in this program, emotion refers to the feelings people experience within situations (Burke and Stets 2009). Emotions serve as a signal of the degree of correspondence between perceptions of the self within the situation and the identity-standard meaning (Burke 1991, 1996; Burke and Stets 2009). Correspondence, or identity verification produce positive emotions whereas noncorrespondence, or identity-nonverification produce negative emotions (Burke and Stets 2009). Displays of emotions within identity are evaluated based upon our perceptions, compared to the perception of how significant others’ evaluate our emotional display. This comparison of our own and other perceptions provide feedback that is related to how we come to see our self (i.e., self-esteem) and psycholgical distress (i.e., depression and anxiety). In this study, I suggest integration of these approaches to emotion be used to better understand variation in college students’ self-esteem and psychological distress.

College students are often faced with a variety of challenges. College students’ mental health is a growing public health concern (Eisenberg et al. 2007). In a survey of 380 college counseling departments from across the United States by the Association for University and College Counseling Center Directors, anxiety was found to be the most common problem faced by students followed closely by depression (Reetz, Barr & Krylowicz 2013). In a similar survey of college students, over half of students reported feeling anxious and over a third reported feeling intense depression during the previous year (American College Health Association Executive Summary 2013). Most students enter college with a range of strategies to manage their emotions (Goldman and Goodboy 2014; Misra and McKean 2000). However, managing expectations related to the student identity (earning good grades, being studious, thriving socially) is complicated when coupled with the additional task of managing expectations regarding emotional displays within identity.

While emotions are profoundly personal, they are remarkably social. What are emotions? As Thoits (1989, p.318) highlights, “There are almost as many definitions of emotion as there are authors…” Thoits (1989, p. 318) defines emotion:

“(a) appraisals of situational stimulus or context, (b) changes in physiological or bodily sensations, (c) the free or inhibited display of expressive gestures and (d) cultural label applied to a specific constellation of one or more of the first three components.”

I explore how our self-evaluation of our emotional displays are compared to our perceptions of significant others’ judgments. If a discrepancy emerges in these comparisons, feedback to how we come to see our self. As a student, professor, mother or close friend - our ability to manage our emotional performance within identity signifies how successful (or unsuccessful) we are at enacting that identity. Through interaction with others, we learn what, how and under what conditions we are to display emotions. When we do not display emotions appropriately, we are sanctioned in some way to remind us what is, or is not appropriate for the situation. Emotional scripts create identity relevant “templates” people use to evaluate their own emotional responses (Wilkins and Pace 2014). This process leads us not just to ask, “Is this how a person should feel?” rather to ask, “Is this how a person like me should feel?” (Wilkins and Pace 2014).

Emotion norms are social conventions that dictate what emotional response is, or is not, appropriate for a given situation. They are often understood as statements that include “should,” “ought,” “must,” or “have a right to” (Thoits 1990, p. 181). Emotion norms give way to display rules. Display rules specify the emotion(s) we should display in a particular situation (Hochschild 1979; [1983] 2012; Ekman 1973 2006). They provide structure for emotional exchanges (Hochschild, [1983] 2012) and govern social behaviors. Display rules prescribe an emotional script to the social world. They become the reference by which we assess our own emotional response that guides how we manage emotions and establish whether the particular emotional response is obligated or called upon within an encounter.

Work within identity theory and emotion has flourished in recent years within the perceptual control emphasis in identity theory (Stets 2003, 2004, 2005; Stets and Asencio 2008; Stets and Osborn 2008; Stets and Carter 2011, 2012; Stets and Trettevik 2016; Stets and Tsushima 2001). Work within this tradition examines emotion as a product, or output, of the control system feedback loop. In this model, emotion is an outcome of success or failure in the act of verification of one’s identity. Failure to verify one’s identity has been shown to impact psychological distress (Linville 1987; Longmore and Demaris 1997; Marcussen and Large 2003; Marcussen 2006; Large and Marcussen 2000). Psychological distress manifests as symptoms of anxiety, depression and often other forms of somatic discomfort (i.e., sleeplessness, etc.). Psychological distress is a general state of emotional arousal or upset; it is associated with the onset and recurrence of clinical disorder such as clinical anxiety or depression (Thoits 2010; Payton 2009). The emphasis within this tradition has been to examine emotion as a response or an output. As part of a cybernetic control process however, emotions are also input.

Emotions are not only a response (as output) of identity verification (or nonverification), they are attached to identity relevant behavior and operate as input within the cybernetic system. Displays of emotion are behavioral. Our evaluation of the behavior is a perceptual process that requires we evaluate our own performance and the reflected appraisals of others related to the performance. As a behavioral response within the control system, identity relevant emotional displays are managed, moderated and processed as identity relevant behavior. Our perceptions of our emotional response compared to our perceptions of others judgment holds consequence to how we see our self. Identities prescribe how to perform emotion appropriately and successful emotional displays signal verification of the identity. The verficiation signal generates feedback to the self that we are acting in accord with our called upon identity. Any discrepancy between self and other perceptions of the emotional response will signal varying degrees of nonverification of the identity. In this study, I shift the focus of emotion in this model from output to input. Emotions, in this case emotional displays, are input that is managed in the same way other identity enactments are.

Given the importance of emotions within sociological theorizing, I explore how the evaluation of our emotional displays operate in the same way we evaluate other identity relevant behaviors, insofar as they provide meanings related to identity. Displaying emotion is an act and it is active. Many sociologists examine emotions as a signal function that alerts the individual to relevant information related to their immediate physical or social environment (Hochschild, [1983] 2012; Kemper 1978; Smith-Lovin and Heise 1988; Thoits 1990). I suggest that we “do” emotion much like we “do” other aspects of our identity all of which has consequences to the self. Previous work has established that nonverification of an identity and discrepancy within identity meanings result in implications such as the decrease in self-esteem and increase in psychological distress (Burke 1991, 1996, Cast and Burke 2002; Stets and Cast 2007; Stets and Burke 2000; Linville 1987; Longmore and Demaris 1997; Higgins 1987, 1989; Marcussen and Large 2003; Marcussen 2006, Large and Marcussen 2000).

Theoretical Background and Literature Review

Cultural-Normative Approaches to Emotion

Hochschild’s (1975) early work is representative of the cultural-normative perspective. She built on Ekman’s ([1973] 2006) conceptualization of display rules, which specify the emotion(s) we should display in a particular situation. This perspective highlights within each domain, or social setting there may be a unique emotion culture (Gordon, 1981, 1990). These culture(s) dictate what emotion is appropriate to feel or display and under what conditions. This approach also highlights the difference in cultural norms related to emotional displays resulting from individual-level characteristics (Hochschild 1975; Simon et al. 1992). How do we evaluate our emotions in these situations? According to Hochschild ([1983] 2012), we do so through emotion management.

Emotion management is the act of controlling one’s feeling(s) to create publicly observable facial and bodily displays (Hochschild, [1983] 2012:7). Emotion management is the act of changing in degree or quality an emotion or feeling within a given situation (Hochschild, [1983], 2012). We engage in emotion management to impact the emotions of other or to achieve goals. The process of evaluating one’s emotional response occurs secondary from the initial physiological arousal of emotion. The sociology of emotions is largely integrative of the behavioral and cognitive aspects of the emotional experience. Within this cultural-normative tradition, behavioral responses are processed through cognitive assessment. It is the secondary cognitive assessment that is critical; it shapes how we see ourselves, evaluate ourselves and helps us moderate how we express our emotions in the future. Emotion management occurs in the “pinch,”,” between “what I expressed” and “what I should have expressed” (Hochschild, [1983] 2012). Individuals evaluate their emotional performance through cognitive appraisals. Through the internalization of emotion and expression norms, one comes to understand the appropriate means to express their feeling within a given situation. Emotion norms prescribe what one “should,” “ought,” “must,” or “have a right to” (Thoits 1990, p.181) and operate as an internal response. Display rules or expression norms refer to expectations governing the public performance of emotion. Expression norms operate in much the same manner as emotion norms but relate to how we manage our displays of emotion rather than internal feelings. What occurs when we violate or deviate from these norms? Thoits (1990) points to a number of theoretical issues associated with emotional deviance. Emotional deviance refers to the experiences in displays of affect that differ in quality or degree from what is expected in the given situation (Thoits 1989, 1990). Social actors are motivated to obtain and maintain social approval, so we can assume that they work towards aligning their emotional displays to the culturally appropriate emotion norms.

In addition to emotion management, another process that shapes emotional displays is identity work (Wilkins and Pace 2014). People dictate to their self and others ideas regarding how members of that group should behave. Identity work refers to individual and collective efforts people engage to give meaning to their social categories (Schwalbe & Schrock 1996). It is through processes of emotional socialization that we can link emotion and identity (Wilkins and Pace 2014). Much of the theoretical and empirical growth merging emotion and identity in this perspective has been conducted examining emotion associated in some way to marginalized identities such as race, class and gender (Anderson 2011; Beeman 2007; Bonilla-Silva 2001; Erickson and Ritter 2001; Froyum 2013; Simon 2007; Stearns 1994; Smith-Lovin 1990; Snow and Anderson 1987; Wilkins and Pace 2014). While research connecting emotion and identity within the sociological literature often focus on social identities, another program of research examines identity as a social psychological process.

Identity Theory

Identity theory develops out of Structural Symbolic Interaction (Stryker, 1980) and shares assumptions of the frame. Identity theory seeks to explain why, given a choice, one invokes a particular identity over another (Serpe 1987; Serpe and Stryker 1987, 1993, 2011). Identity theory simultaneously recognizes that constructing the new and reproducing the old can fit within a single framework. It distinguishes the self (as the whole) from identities (parts of the whole) (Stryker 1980 [2002]). This distinction is critical. Roles carry expectations attached to positions in network relationships and identity is the internalization of those role expectations. A role is a set of appropriate behaviors attached to a social position. A role identity is the personal attachment or intrinsic acceptance of the identity associated with a specific role. For example, I can see myself as a mother, student, and friend. Each of these role identities comes with a social script that I have learned through interactions with others. Along the way, I have learned what it means to be a student, mother, and friend and thus I can negotiate between role identities. Stryker (1980 [2002]) conceptualized persons’ role identities to be organized in a salience hierarchy whereby a salient identity (near the top of the hierarchy) is one that has a greater likelihood of being activated across a variety of settings. Salient identities are more likely to call upon behavior that is associated with the identity and as such identifies how social actors will behave in a situation (Stryker 1980 [2002].

Perceptual Control Emphasis in Identity Theory

According to Burke (1991, 1996) and Burke and Stets (2009), the identity process operates to contol perceptions within interaction to maintain meanings associated with identity enactment. There are four key components to control system model. The four key components of the control system model include the identity standard, inputs, comparator and behavioral output (Burke and Stets 2009). The identity standard carries a set of meanings, which may be viewed as defining the character of the identity they serve. They act as a point of reference in the identity process. The meanings that define the identity standard are stored in memory and are accessible in the comparator (described below). Inputs and perceptions are central to the identity process. Perceptions are compared with the identity standard. The goal of is to align, or match perceptions to the standard (Burke and Stets 2009). Perceptions are the inputs to identities; in particular, perceptions are meanings in the situation that are relevant to the identity. The comparator compares the input perceptions of meanings relevant to the identity with the memory meanings of the identity standard. If there is a mismatch between input and identity standard a discrepancy, or “error signal,” is generated. In this model, output is the behavior in the situation. Any discrepancy on the output side of the identity model occurs when the meanings that are given off by one’s behavior in a situation is inconsistent with the meanings held in one’s identity standard (Stets & Carter 2006; Stets, Carter, & Fletcher 2008).

Research and theoretical development within the perceptual control emphasis in identity theory presents some propositions and empirical support to how emotion and identity theory can be integrated (Stets 2003, 2004, 2005; Stets and Asencio 2008; Stets and Osborn 2008; Stets and Carter 2011, 2012; Stets et al. 2008; Stets and Trettevik 2016; Stets and Tsushima 2001). According to Burke (1991), frequent interruptions in the identity-verification process produce intense negative feelings. Frequent interruption of normal action and thought processes, including those related to the process of identity-verification, generates stress (Mandler 1982). Burke (1991, 1996) emphasizes that stress results from the difficulty in, or failing to, achieve congruity between perceptions and standards. Production of negative feelings will occur regardless of whether the nonverification is in a positive (self-perceptions in the situation do not meet the standard) or negative direction (self-perceptions in the situation do not meet the standard) (Burke & Harrod 2005). For Burke (1991) the source of the disruption in the perceptual control system is of importance. Interruptions from significant others, such as friends or family, will generate more consequential negative emotions. Identity meanings among significant others are more likely to be tightly organized and produce more consequential negative effect (Burke and Stets 2009).

Perceptual Control System and Self-Esteem

Self-esteem is a positive or negative orientation toward oneself; it is the overall evaluation of one’s worth or value. For Rosenberg (1995, pp. 142) it is the “totality of the individual’s thoughts and feelings with reference to himself as an object.” For some, it is the evaluative and affective aspects of the self-concept indicating how “good” or “bad” we feel related to our sense of self (Gecas 1982, 1989, 2003; Owens and Serpe 2003) and others define it the self-evaluative part of the self-concept (Owens and Serpe 2003). Holding high self-esteem is the perception that one is a good, valued and competent (Thoits 2010). Self-esteem theory holds that individuals are highly motivated to protect and enhance their self-esteem, and according to Cast and Burke (2002), self-esteem is enhanced by successful self-verification.

Perceptual Control System and Psychological Distress

Within the perceptual control system, the goal to verify identity is continuous. Any interruption (or stress) activates a signaling system that requires attention, prompting the individual to cognitively manage their behavior to bring the system back into congruence. A discrepancy indicates an interruption in the control model process that halts the normal procedure of continuous congruence between reflected appraisals and the identity standard (Stotland and Pendleton 1989). Consequences of interruptions, incongruence, and identity non-verification have been found to impact psychological distress (Stotland and Pendleton 1989). Psychological distress is the co-occurring symptoms of anxiety, depression and somatic discomfort (e.g., sleeplessness, etc.) indicating a general state of emotional arousal or upset. Psychological distress is associated with the onset and recurrence of clinical disorder (Thoits 2010; Payton 2009).

Specific Emotions in the Sociology of Emotions

Most theories within the sociological study of emotions have been primarily focused on the positive-negative valence of emotion and ignore the complexity of human emotion (Leveto 2012, 2016; Turner and Stets 2005). Within identity theory, most research related to emotion has only looked at the valence of emotion with few exceptions (Stets and Trettevik 2016; Stets and Tsushima 2001; Stets & Burke 2005). Emotion scholars advocate the distinction of specific emotions as opposed to the general valence of emotion (Ekman 1992a, b; Kemper 1987). There is a general consensus among emotions scholars that there are four primary emotions; happiness, fear, anger, and sadness (Kemper 1987). Burke and Stets (2009) along with Trettevik (2016) call upon identity theorists to examine factors that interact with the verification as they relate to specific emotions such as those primary emotions of happiness, fear, anger and sadness.

Hypotheses

I explore the relationships among identity discrepancy, self-esteem, and psychological distress by focusing on the perception of individual’s assessment of their emotional displays and the identity standard meanings guiding the emotional displays within the student identity. I begin with the assumptions of the cultural-normative perspective, that emotional displays are managed within social encounters through the internalization of emotional scripts that are known as identity relevant. For example, I may see myself as a professor and along with that identity comes a certain prescription to what it means to display emotion within my identity. The emotion management process suggests that I cognitively assess my emotional response. In doing so, I not only evaluate my response as a professor; I compare my evaluation to that of significant others.

Discrepancy in the evaluation of our emotional response compared to our perception of the evaluation by others, signals nonverification of the identity. As previous research has shown a failure to verify the identity based upon the discrepancy is tied to lower self-esteem (Cast and Burke 2002; Stets and Cast 2007; Stets and Burke 2014a, b) therefore I expect:

  • Hypothesis 1: The greater the discrepancy in the evaluation of the emotional response of happiness related to the student identity the lower self-esteem.

  • Hypothesis 2: The greater the discrepancy in the evaluation of the emotional response of sadness related to the student identity the lower self-esteem.

The successful management of emotional display provides a signal function to the self that one is competent in enacting and occupying that identity. A display rule acts as the identity standard in the feedback loop against which emotional displays are compared. The input is the perception of the emotional display. The perception is compared to the display rule (standard) in the comparator. When congruence is obtained, the automatic process continues without disruption. However, when a discrepancy is felt between the input (perception of the emotional display) and the standard (perception of display rule) one will experience distress. The magnitude of the disruption or the greater the size of the discrepancy will result in greater depression and increased anxiety (Linville 1987; Longmore and Demaris 1997; Burke 1991; 1996; Marcussen and Large 2003; 2006, Large and Marcussen 2000). Specifically, I examine the following hypotheses;

  • Hypothesis 3: The greater the discrepancy in the evaluation of the emotional response of happiness related to the student identity the greater the depression.

  • Hypothesis 4: The greater the discrepancy in the evaluation of the emotional response of sadness related to the student identity the greater the depression.

  • Hypothesis 5: The greater the discrepancy in the evaluation of the emotional response of happiness related to the student identity the greater the anxiety.

  • Hypothesis 6: The greater the discrepancy in the evaluation of the emotional response of sadness related to the student identity the greater the anxiety.

Data and Methods

Participants

I rely on data from a convenience sample of 1100 undergraduate students at a large Midwestern University. Respondents were recruited from college courses in a variety of disciplines and asked to complete a self-administered pencil and paper questionnaire. All procedures performed in this study involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants included in the study. Of the 1122 students recruited to participate, 1100 students completed questionnaires resulting in a 98% response rate.

Table 1 presents the means, standard deviations, and ranges for all variables within the models. The average age was 20.4 years and ranged from 18 to 54 years. The sample was predominately white, with only 15% of the sample reporting as non-white. The sample was 70% female, with an average GPA of 3.1. Of the 974 students that reported feeling happy in the previous week, the average discrepancy related to their emotional performance was low at .89 (ranging from 0 to 10). Of the 762 students that reported feeling sad related to the student identity in the previous week, the average discrepancy was only slightly higher at 1.18 (ranging from 0 to 10). Self-esteem scores range from 0 (low) to 4 (hight), with an average of 2.1. Depression scores range from 0 (low) to 30 (high) with an average of 9.6. Anxiety scores range from 0 (low) to 12 (high) with an average at 4.1.

Table 1 Means, standard deviations and ranges for all variables
Full size table

Measures

Identity

Following previous research in identity theory using a college student population (Serpe 1987; Serpe and Stryker 1987, 1993), college students are most likely to claim identities as students that reflect the nature of being tied to social networks within a university. To activite the student identity, respondents were asked to think about themselves as students.Footnote 1 Specifically, before each set of questions asking about their emotional responses within their identity, they were instructed to: Think about your role as a student; specifically think of the activities that you engage in as a student (class, student organizations, study groups, etc.). For the following section think about your role as a student and answer the following questions. Another mechanism that I utilized to remind respondents to respond to the questions with respect to their identity was to embed prompts such as, “In thinking of yourself as a student….” followed by the emotion specific question.

Happiness and Sadness

Following the Emotions Module of the General Social Survey (1996), respondents were asked to evaluate the performance of emotion by thinking about the last time they felt the emotion in the previous week. Emotional responses are assessed upon self-reports of the emotion. “Have you felt sad (happy) related to being a student in the previous week?” If respondents reported the emotion, a series of follow up questions were asked. If not, they were prompted to skip to the next emotion.

Identity Discrepancy: Discrepancy in Emotional Performance

Identity discrepancy is operationalized as a discrepancy in the perception of the emotional performance. To measure discrepancy, I examine the difference between the reflected appraisals (perception of others evaluation of emotional response) and self-evaluation (self-evaluation of emotional response). The self-evaluation item asks, “In thinking of yourself as a student, think of the last time you felt sad (happy). Do you think your emotional response was appropriate?” Response categories ranged from 0 (not appropriate at all) to 10 (completely appropriate). To assess the reflected appraisal of the emotional response, I ask, “In thinking of yourself as a student, think of the last time you felt sad (happy). Do you think that others (your family, friends, etc.) would have thought your emotional response was appropriate?” Response categories ranged from 0 not appropriate at all to 10 completely appropriate. A new item is created that calculates the difference between the perceived reflected appraisals (others-evaluation of emotional response) from the self-evaluation (self-evaluation of emotional response). Following previous work (Stets and Harrod 2004) in identity-verification/discrepancy research, I took the absolute value of the difference. Discrepancy scores range from 0 (no discrepancy) to 10 (greatest discrepancy).

Self-Esteem

Self-esteem was assessed using the Rosenberg Self-Esteem Scale (RSE) (Rosenberg 1965), the most frequently and well-validated measure of self-esteem (Robins et al., 2001). The Rosenberg Self-Esteem Scale (RSE) consists of 10 items assessing global self-esteem (e.g. “On the whole, I am satisfied with my self”). Negative items from the Rosenberg scale were reverse coded. Self-esteem was captured by the creation of a 10– item scale. Respondents were asked rate their general feelings related to their self from one (strongly disagree) to four (Strongly agree) for each of the items the following items; “I feel that I am a person of worth, at least on an equal basis with others.”, “I feel that I have a number of good qualities.”, “All in all, I am inclined to feel that I am a failure.” (reverse coded), “I am able to do things as well as most other people.”, “I feel I do not have much to be proud of.” (reverse coded), “I take a positive attitude toward myself.”, “On the whole, I am satisfied with myself.”. “I wish I could have more respect for myself.” (reverse coded), “I certainly feel useless at times.” (reverse coded), “At times I think that I am not good at all. (reverse coded). The self-esteem scale has an alpha reliability of .89.

Depression

Depression is conceptualized using a short form version of the Center for Epidemiological Studies Depression (CES-D) screening instrument the CESD-10 (Andresen et al. 1994). The CES-D10 has been shown to be a reliable measure for assessing the number, types, and duration of depressive symptoms across racial, gender and age groups (Knight et al. 1997). Respondents were asked to evaluate a list of the ways they might have felt or behaved during the previous week. Items asked, during the past week; “I was bothered by things that usually don’t bother me.”, “I had trouble keeping my mind on what I was doing.”, “I felt depressed.”, “I felt that everything I did was an effort.”, “I felt hopeful about the future” (reverse coded). “I felt fearful.”, “My sleep was restless.”, “I was happy” (reverse coded), “I felt lonely.”, “I could not get “going.” Response categories for each item range from one (Rarely or None of the time, less than 1 day) to four (Most or All of the time, 5–7 days). The total score is calculated by finding the sum of the 10 items, any score above 10 is considered depressed (Björgvinsson et al. 2013). Alpha reliability for depression with this sample is .79.

Anxiety

Anxiety is measured by replicating the General Social Survey (1996) conceptualization of anxiety. Respondents were asked to evaluate a list of the ways they might have felt or behaved during the previous week. Items asked during the past week: “I felt so restless that I couldn’t sit long in a chair.”, “I felt worried a lot about little things.”, “I felt anxious and tense.”, “I had trouble keeping my mind on what I was doing.” Response categories ranged from one (rarely or none of the time, less than 1 day) to four (most or all of the time, 5–7 days). Cronbach alpha reliability for the four-item scale is .79.

Controls

Many factors can impact the self-esteem, depression, and anxiety of college students (Andrews and Wilding 2004; Dyrbye et al. 2006; Eisenberg et al. 2007). To attempt to control for the impact of these factors, I statistically control for age, race, sex, grade point average (GPA) and relationship status. Age is measured in years. Race is dummy coded: 0 represents white respondents, and 1 represents nonwhite respondents. Sex is dummy coded: 0 represents men, and 1 represents women. GPA is the grade point average students reported on ranging from 0 to 4. Relationship status is dummy coded: 0 represents not in a relationship and 1 represents in a relationship.

Analysis

I investigate the relationship age, race, sex, grade point average (GPA) and relationship status on the emotional response discrepancies related to the expression of happiness and sadness within the student identity on self-esteem, depression, and anxiety using structural equation modeling. Specifically, I estimate structural models that examine the direct and indirect effects (through emotional response discrepancies) related to both happiness and sadness within the student identity on self-esteem and psychological distress (depression and anxiety). Independent variables were allowed to correlate in both models (happiness discrepancy and sadness discrepancy). I select structural (SEM) over separate aggregate ordinary least square regression (OLS) analysis because SEM can account for measurement error and manage multiple endogenous variables simultaneously. SEM is also a linear analysis method that enables simultaneous testing of all relationships within one model. The structural model is depicted in Fig. 1.

Fig. 1
figure 1

Structural model

Full size image

Results

Zero-order correlations are reported in Table 2, examining the relationship between variables across both models (sadness and happiness). These correlations help determine the relationships between the theoretical concepts in the model and the data. As anticipated, the emotional performance discrepancy items (of happiness and sadness) are significantly correlated with the endogenous variables - self-esteem, depression, and anxiety. Other correlations worth noting are the relationships between GPA and both self-esteem and depression. The greater the GPA, the higher the self-esteem and lower the depression. Being in a relationship also seems to be correlated with higher reports of self-esteem. Among the endogenous variables, self-esteem, depression, and anxiety are all significantly correlated with each other.

Table 2 Correlation matrix
Full size table

Table 3 presents the standardized estimates of the emotional discrepancy related to happiness within the student identity. I find support for hypotheses one through three related to happiness. The greater the discrepancy in the evaluation of the emotional response of happiness within the student identity the less self-esteem (H1), the greater the depression (H2) and the more anxiety (H3). Model fit statistics demonstrate excellent fit (n = 744, chi-square/degrees of freedom = 4/9.785 with RMSEA = 0.040, CFI = .994). The greater the discrepancy in the display of happiness within the student identity the less self-esteem (β = −.09), the greater the depression (β = .42) and the more anxiety (β = .25).

Table 3 Standardized estimates
Full size table

Table 4 presents the standardized estimates of the emotional discrepancy related to sadness within the student identity. I find support for hypotheses four through six related to sadness. The greater the discrepancy in the evaluation of the emotional response of sadness within the student identity the less self-esteem (H4), the greater the depression (H5) and the more anxiety (H6). Model fit statistics demonstrate excellent fit (n = 572, chi-square/degrees of freedom = 4/5.130 with RMSEA = 0.022, CFI = .998). The greater the discrepancy in the display of sadness within the student identity the less self-esteem (β = −.06), the greater the depression (β = .53) and the more anxiety (β = .33).

Table 4 Standardized estimates
Full size table

Discussion & Conclusion

My work begins to explore the benefits of weaving cultural-normative approaches to emotion with identity theory approaches to emotion. Combining principles from the cultural-normative perspective such as emotion norms, display rules and emotion management to our understanding of the perceptual control emphasis of identity theory enables us to understand the process by which our emotional displays are closely aligned with identity meanings. I attempt to shift the focus of emotion within the perceptual control emphasis in identity theory beyond an outcome of identity verification. Emotional displays are behavioral manifestations of internal feelings they adhere to display rules and are structured as meanings attached to identity standards. Emotion is not only an outcome of identity verification or nonverification it is embedded in our perceptions of our emotional display and operates also as input into the control system. Emotional displays and the emotion relevant meanings associated operate on a continuous loop, they are motivated to align with expectations and to work toward identity verification. Discrepancies in the perceptions of emotional displays signal a deviation from the emotion script and the called upon display rules. My results press the question, at what point does an emotional discrepancy become an emotional interruption?

This data demonstrates an interesting distribution of the discrepancy measures related to happiness and sadness within the student identity. Figure 2 and Fig. 3 are histograms of the distribution of discrepancies for both happiness and sadness. As you can see, the reverse-J distribution for both indicates that the impact of these discrepancies is not necessarily from large discrepancies, but rather small emotional discrepancies. These small discrepancies show to have a significant impact on self-esteem, depression and anxiety. While my results do find that the greater the discrepancy the greater the impact, even small discrepancies carry feedback to the self that one is not performing to standards and as such, contributes to lower self esteem, greater depression and anxiety.

Fig. 2
figure 2

Histogram happiness discrepancies

Full size image
Fig. 3
figure 3

Histogram sadness discrepancies

Full size image

I find that regardless of the direction of the discrepancy, the greater the discrepancy the more pronounced the impact on self-esteem, depression and anxiety. There is an ongoing discussion within identity theory related to self-enhancement (people seek positive evaluations and avoid negative evaluations) versus cognitive consistency processes (people seek evaluations that match their self-views and avoid evaluations that do not match their self views) (Stets and Burke, 2014a, b Trettevik 2016). If students evaluated their emotional response appropriate and yet they perceived others to deem it inappropriate, the discrepancy is related to less self-esteem, greater depression and more anxiety. As identity theory would predict, if students evaluated their response as inappropriate but felt significant others would think their response was appropriate, this discrepancy also negatively impacted self-esteem and increased depression and anxiety. This supports identity theory and consistency principles that suggest that a discrepancy between the reflected appraisals and identity standard (nonverification of an identity) generate negative outcomes (including negative emotion (Stets and Burke, 2014a, b).

Display rules within emotion cultures associated with happiness and sadness are important as they relate to student’s self-esteem, depression and anxiety. By examining specific emotions, as opposed to the valence (positive/negative) we can see how these emotions in particular operate. Happiness and sadness are primary emotions that carry fairly well established and internalized social scripts with shared cultural meanings associated with the displays of the emotion. I would expect that more complex emotions or emotions that have less clear expression norms within the identity, present challenges for future work.

There are a number of ways in which this work can extend to have applied impact among college students. This work can inform approaches that clinicians and professionals working with students could potentially take to minimize depression and anxiety symptoms. Realistic emotion management coping responses associated with the student identity could be integrated into orientations to college life and embedded in the cultural landscape of the university setting. For instance, many institutions have incorporated a first year experience course that could integrate curriculum around common emotions experienced during their college experience and how to manage those emotions as well as how to handle discrepant feedback from others. Additionally, recognizing that depressive symptoms and anxiety are related to these discrepant perceptions of their emotional responses, coping strategies with the symptoms of both depression and anxiety would be useful. Depression and anxiety screenings as part of the outreach on college campuses would enhance the ability for colleges to buffer the impact.

Limitations and Future Research

Discrepancies in the verification process have implications for self-esteem and psychological distress. The evaluation of emotional display operates within the perceptual control emphasis in identity theory and identity standards include emotional display rules. Internalized emotion scripts and perceived judgments from significant others impact college students. Student self-esteem and psychological distress face consequence by the lack of verification of the identity related to the emotional display within the identity. Future work would be best served to examine the process by which we internalize emotion scripts, negotiate the interpersonal interactions that provide feedback related to our emotional display within identity enactment and explore implications beyond self-esteem and psychological distress.

There are a number of limitations to consider that provide pathways for future research to examine. The current data only provides a cross sectional view of the relationships. To truly examine the process; including the direction of these relationships, it is essential to have longitudinal data. Cross sectional data captured by a self-administered questionnaire can only give a “sneak peak” into the complex processes under exploration. Future research would benefit greatly by using innovative instruments to collect emotion relevant data in “real” time through the use of electronic journals that randomly prompts respondents to reflect on their emotions and emotional displays. Using electronic journals over time will enable us to investigate the process and direction of these relationships.

Another limitation is the measurement of the emotional “response.” The current work utilized data that assumed to be the evaluation of the emotional display however the respondent may have interpreted the question to be asking for an evaluation of their feeling (internal) and not display (external). Future research should ensure that clarity in the measurement item establishes a clear difference between the “feeling” and the “display” of the emotional response. Both the authentic feeling and the displays are important to this research, and future work can examine how inauthenticity contributes this model. Another limitation is the use of the “last” time respondents felt the emotion without context to the stimulus. Experimental research may be one way to elicit an emotional response that is specific and connected to only the identity. Respondents’ culture of orientation was not taken into consideration (i.e., students not from the United States). Future research will benefit to explore the variation by culture related to these processes.

This work only focuses on displays of happiness and sadness within one identity. While this is theoretically useful, emotions often manifest as constellations and across identities. To tease out these effects, future research would be well served to examine multiple emotions in a variety of identities to determine if the display rules for happiness and sadness differ from anger and fear and if these vary by identity. While this work focuses on the student identity individually, future work would benefit from looking at the salience of identity. It is possible that the greater the salience of the identity the greater the impact nonverification would have related to self-esteem and psychological distress. Sociological approaches to emotion have the potential to unlock the power of our empirical and theoretical work while simultaneously working toward applied applications. Research in emotions will continue to accelerate as sociologists find new ways to examine emotion as artifacts of structural and cultural patterns in society.