Over time, research has provided an increasingly clear picture of the influence of threat on ascribed or involuntary identities, such as gender identity (Cadaret, Hartung, Subich, & Weigold, Reference Cadaret, Hartung, Subich and Weigold2017), national identity (Fischer, Haslam, & Smith, Reference Fischer, Haslam and Smith2010), and ethnic identity (Hanselman, Bruch, Gamoran, & Borman, Reference Hanselman, Bruch, Gamoran and Borman2014). Studies also suggest that negative consequences are the major focuses of identity threat, such as decreasing performance (Baysu & Phalet, Reference Baysu and Phalet2019), undermining leadership aspirations (Davies, Spencer, & Steele, Reference Davies, Spencer and Steele2005), burnout (Hall, Schmader, & Croft, Reference Hall, Schmader and Croft2015), counterproductive workplace behavior (Aquino & Douglas, Reference Aquino and Douglas2003), and unethical behavior (Zhou, Dou, & Wang, Reference Zhou, Dou and Wang2019). This stream of research provides a comprehensive explanation of the negative consequences caused by ascribed or involuntary identity threats and how organizations can help their employees cope with these types of threats.
Despite the considerable empirical studies on the consequences of ascribed or involuntary identity threats, they have provided little guidance on how individuals actually respond to occupational identity threat. As an important component of self-conception, occupational identity reflects the meanings and value associated with an occupational or professional group (Dutton, Roberts, & Bednar, Reference Dutton, Roberts and Bednar2010) that help define ‘who we are’ by ‘what we do’ (Nelson & Irwin, Reference Nelson and Irwin2014). If the meanings and value represented by an occupation are not acknowledged by others, individuals will experience occupational identity threat. Given that occupational identity threat is an extension of identity threat, we follow the definition of identity threat (Petriglieri, Reference Petriglieri2011) and define occupational identity threat as experiences appraised as being of potential harm to the meanings, value, or enactment of occupational identity. In the face of occupational identity threat, individuals are likely to engage in some positive coping strategies. However, given the lacuna of research on both occupational identity threat and positive consequences of identity threat, understanding of whether encountering occupational identity threat triggers individuals' responses that promote the functioning of their organizations is lacking.
In this research, we draw on identity theory to argue that individuals find the lack of verification of their occupational identity when encountering occupational identity threat, and will experience negative emotion (Burke, Reference Burke1991, Reference Burke and Kaplan1996). This negative emotion then serves as motivation to drive individuals to engage in feedback-seeking behavior as a coping strategy. As an important proactive behavior, feedback-seeking behavior helps individuals evaluate whether their work has met performance standards and their behavior is considered appropriate, and therefore is beneficial to performance and organizational functioning (Anseel, Beatty, Shen, Lievens, & Sackett, Reference Anseel, Beatty, Shen, Lievens and Sackett2015; Lam, Peng, Wong, & Lau, Reference Lam, Peng, Wong and Lau2017). Moreover, we argue that the magnitude of the effects of occupational identity threat on negative emotion and the downstream feedback-seeking behavior are contingent on individuals' performance self-esteem. Capturing the overall evaluation of their performance and competence (Heatherton & Polivy, Reference Heatherton and Polivy1991), performance self-esteem affects individuals' identity standards and, thus, their responses to occupational identity threat. Individuals high in performance self-esteem hold higher or more positive occupational identity standards and therefore should respond more strongly to occupational identity threat. We, thus, posit that individuals high (vs. low) in performance self-esteem are more likely to experience negative emotion and engage in feedback-seeking behavior.
Our research makes three contributions to the literature. First, by exploring the emotional and behavioral consequences of occupational identity threat, we broaden and enrich the research on identity threat. Second, our study offers a new theoretical lens to identity threat research by drawing on identity theory to examine how individuals respond to occupational identity threat. Identity theory offers a new perspective on occupational identity threat and helps us identify negative emotion and feedback-seeking behavior as important consequences. Doing so strengthens the application of identity theory in identity threat literature. Third, we adopt a contingency view and introduce performance self-esteem as a critical boundary condition, thus deepening the understanding of the effects of occupational identity threat and expanding the research on identity threat from the perspective of individual differences. Figure 1 depicts our theoretical framework.
Figure 1. Conceptual model.
Theory and hypotheses development
Identity theory and interpretations of occupational identity threat
Identity theory (Burke, Reference Burke1991, Reference Burke and Kaplan1996) posits that the identity process is a self-regulating control system that involves continuous interplay between individuals and situations. This identity process presents a feedback loop that consists of four components (Burke, Reference Burke1991). The first component is identity standard, which refers to the set of self-meanings that define the character of the identity (i.e., what it means to be who one is). The second is input, which captures reflected appraisals or perceptions of self-meanings in the situation (i.e., how people think others see them). The third is comparator, which is a process that compares the input with the standard and verifies the identity. When the input meanings are incongruent with the meanings of identity standard in the comparator (i.e., identity nonverification), negative emotional reactions occur and result in output. The fourth component is output, which manifests as meaningful actions and behaviors in the environment that change the input function and complete the feedback loop.
According to identity theory, occupational identity threat is essentially identity nonverification in a negative direction; that is, the input meanings in the situation are negative compared with the identity standard (Burke, Reference Burke1991; Meister, Jehn, & Thatcher, Reference Meister, Jehn and Thatcher2014; Petriglieri, Reference Petriglieri2011). By definition, occupational identity threat refers to perceived others' negative views on the meanings, value, or enactment of an individual's occupational identity (Petriglieri, Reference Petriglieri2011). Perceived others' negative views correspond to the input in the situation, whereas the meanings, value, and enactment represent the identity standard. When individuals believe that others' views about their occupational identity are negative compared with the meanings they hold, they would experience occupational identity threat. For example, a person may associate the meaning of ‘competence’ with his or her occupational identity and rate ‘competence’ in identity standard highly (e.g., a 9 on a 10-point scale). If this person believes that others perceive his or her ‘competence’ of occupation as a 3, the input meaning falls short of the identity standard meaning. This identity nonverification is in a negative direction and presents a threat to the person's occupational identity. Given that identity theory provides a comprehensive perspective on identity threat, we use it as our theoretical lens to understand occupational identity threat and emotional and behavioral consequences.
Occupational identity threat and negative emotion
Identity theory suggests that emotions play an important part in identity verification (Stets, Reference Stets2005, Reference Stets2015; Stryker, Reference Stryker2004). According to identity theory, individuals are motivated to maintain consistency between meanings in the identity standard and meanings in the social situation (input), so as to attain identity verification within and across situations (Stets, Reference Stets2005, Reference Stets2015). If individuals fail to verify their identity (i.e., the meanings in the social situation do not match the meanings contained in their identity standard), they will experience threats to the maintenance of identity meanings and to their ability to control the situation, which in turn will lead them to experience negative emotions (Stets, Reference Stets2015; Stets & Burke, Reference Stets and Burke2005). Conversely, when the input meanings in the situation successfully meet the identity standards, identity can be verified and thus evoke positive emotions (Burke, Reference Burke1991; Burke & Stets, Reference Burke and Stets1999; Cast & Burke, Reference Cast and Burke2002). Recent studies have provided empirical evidence for these predictions (e.g., Burke, Reference Burke, Serpe, Stryker and Powell2020; Stets & Burke, Reference Stets and Burke2014; Tripathi, Zhu, Jacob, Frese, & Gielnik, Reference Tripathi, Zhu, Jacob, Frese and Gielnik2020).
As such, it is reasonable to infer that occupational identity threat as a kind of identity nonverification will evoke negative emotions. Occupational identity plays an especially important role in giving meaning and direction to individuals' career, thus leading them to seek jobs that match their personal strengths, interests, preferences, and goals (Skorikov & Vondracek, Reference Skorikov, Vondracek, Schwartz, Luyckx and Vignoles2011). Thus, given the importance of occupational identity, threats to occupational identity can cause great harm to individuals' identity meanings and, in turn, evoke negative emotion (Burke, Reference Burke1991, Reference Burke and Kaplan1996). Thus:
Hypothesis 1: Occupational identity threat is positively related to negative emotion.
Occupational identity threat and feedback-seeking behavior
According to identity theory, in the output phase, individuals engage in meaningful actions and behaviors to cope with occupational identity threat (Burke, Reference Burke1991, Reference Burke and Kaplan1996). Given the effectiveness against occupational identity threat, we identify feedback-seeking behavior as a straightforward coping strategy. Feedback-seeking behavior refers to individuals' proactive search for evaluative information about their job-related behavior and performance (Ashford & Tsui, Reference Ashford and Tsui1991; De Stobbeleir, Ashford, & Buyens, Reference De Stobbeleir, Ashford and Buyens2011). By seeking feedback, individuals can reappraise the occupational identity threat and strengthen self-verification (Anseel et al., Reference Anseel, Beatty, Shen, Lievens and Sackett2015; Ashford & Cummings, Reference Ashford and Cummings1983; Morrison & Bies, Reference Morrison and Bies1991), which is effective at coping with occupational identity threat.
On the one hand, feedbacks offer information to help individuals appraise their experience about occupational identity threat more comprehensively and reduce their uncertainty. According to identity theory, the input meanings in the social situation are individuals' subjective perceptions, so occupational identity threat arising from their subjective appraisals (Burke, Reference Burke1991; Petriglieri, Reference Petriglieri2011). In this sense, when encountering occupational identity threat, individuals will try to obtain more overall information to verify their subjective judgment of the threat, so that they can take further coping strategies. Engaging in feedback-seeking behavior enables individuals to obtain technical, role, and evaluative information about the correctness and adequacy of their behaviors and performance (Ashford, Reference Ashford1986; Lam et al., Reference Lam, Peng, Wong and Lau2017; Vandenberghe et al., Reference Vandenberghe, Landry, Bentein, Anseel, Mignonac and Roussel2021). With this information, individuals can reappraise their occupational identity standard, the input in the situation, and thus the perception of occupation identity threat.
On the other hand, although accurate self-relevant information is most instrumental for appraising occupational identity threat, individuals tend to prefer favorable information that helps them maintain a positive self-view and promotes self-verification (Ashford, Blatt, & VandeWalle, Reference Ashford, Blatt and VandeWalle2003). Research shows that seeking feedback consistent with individuals' self-concept is beneficial to the process of self-verification (Swann & Read, Reference Swann and Read1981). As people in an occupational group have some similarities, they are more likely to form a consistent positive view of their shared occupational identity. As such, it is easier for individuals to obtain positive views about their occupational identity from their colleagues or supervisors. These positive views strengthen the meanings of individuals' occupational identity standard and offset the negative meanings from other groups, thus reducing the accuracy and severity of threats (Petriglieri, Reference Petriglieri2011).
Taken together, we argue that individuals are likely to engage in feedback-seeking behavior as a coping strategy for the purpose of reappraising occupational identity threat or strengthening self-verification after encountering occupational identity threat. Thus:
Hypothesis 2: Occupational identity threat is positively related to feedback-seeking behaviors.
The mediating effect of negative emotion
Identity theory assumes that negative emotion creates a force, pressure, or motivation to reduce the inconsistency between the input meanings and identity standard (Stets, Reference Stets2015). In other words, negative emotions motivate individuals to take actions to solve occupational identity threat. Given that feedback-seeking behavior is an effective way to cope with occupational identity threat, we argue that individuals who experience negative emotions under occupational identity threat will engage in feedback-seeking behavior. Taken together, we expect that negative emotion serves as an important emotional mechanism accounting for the occupational identity threat on consequent feedback-seeking behavior. Thus:
Hypothesis 3: Negative emotion mediates the positive relationship between occupational identity threat and feedback-seeking behaviors.
The moderating effects of performance self-esteem
Occupational identity threat has potential positive implications on negative emotion and feedback-seeking behavior. However, we expect this positive influence to vary across individuals depending on their performance self-esteem. As an important dimension of self-esteem, performance self-esteem refers to individuals' overall evaluation of their performance and competence (Heatherton & Polivy, Reference Heatherton and Polivy1991). Given that competence is the main component of occupational identity standard and, therefore, the main target of occupational identity threat (Aquino & Douglas, Reference Aquino and Douglas2003), we argue that individuals with different levels of performance self-esteem will experience different levels of negative emotion and further engage in different levels of feedback-seeking behavior after encountering occupational identity threat.
According to identity theory, when the inconsistency between the input meanings and identity standard (i.e., identity nonverification) is unanticipated or difficult to correct, stronger negative emotions will come into play (Burke, Reference Burke1991, Reference Burke and Kaplan1996; Cast & Burke, Reference Cast and Burke2002). By integrating performance self-esteem into identity theory, we argue that compared with individuals low in performance self-esteem, individuals high in performance self-esteem will have a higher overall evaluation of their competence and, as such, will hold higher or more positive occupational identity standards (Ferris, Lian, Brown, & Morrison, Reference Ferris, Lian, Brown and Morrison2015). These individuals firmly believe that they perform well in their occupation, so they will be less prepared to handle the occupational identity threat and thus experience stronger negative emotion. In addition, given the belief in their competence, individuals high in performance self-esteem are less likely to make concessions to threats by decreasing their occupational identity standard to align with the negative meanings in the situation. Without adjusting the occupational identity standard, it is more difficult for individuals high in performance self-esteem to overcome threats in the feedback loop and thus will experience stronger negative emotions (Burke & Stets, Reference Burke and Stets1999; Cast & Burke, Reference Cast and Burke2002). Thus:
Hypothesis 4: Performance self-esteem moderates the positive effect of occupational identity threat on negative emotion, such that the positive relationship is stronger when performance self-esteem is high (vs. low).
Moreover, given that individuals high in performance self-esteem will feel stronger negative emotions than those low in performance self-esteem when encountering occupational identity threats, it is more urgent for them to deal with the dilemma of occupational identity by seeking feedback. However, as seeking feedback from others can be beneficial or ‘ego-deflating,’ feedback seekers may face a conflict between the desire to obtain accurate feedback information and the need to protect self-image (Ashford & Cummings, Reference Ashford and Cummings1983). Thus, we argue that individuals high in performance self-esteem are more confident about their competence and show less fear of negative feedback, so they will be more willing to seek feedback to cope with threats. By contrast, individuals low in performance self-esteem may pay more attention to protecting their weak egos and thus try to avoid negative feedback (Emery, Gardner, Carswell, & Finkel, Reference Emery, Gardner, Carswell and Finkel2018). As such, they are less likely to engage in feedback-seeking behaviors. Taken together, we propose the following:
Hypothesis 5: Performance self-esteem moderates the positive indirect effect of occupational identity threat on feedback-seeking behavior through negative emotion, such that the positive indirect effect is stronger when performance self-esteem is high (vs. low).
Study overview
We conducted three studies to test the hypotheses. Study 1 is an experiment to verify the validity of the manipulation of occupational identity threat and the causal relationship between occupational identity threat and feedback-seeking behavior. Study 2 is also an experiment to replicate the results of study 1 and to further verify the causal relationship between occupational identity threat and negative emotion, as well as the mediating effect of negative emotion. Study 3 is a field study to replicate the results of studies 1 and 2, further verify the moderating effects of performance self-esteem, and provide external validity for our theoretical model in an organizational setting.
Occupational identity threat mainly comes from the overt actions of other parties (Aquino & Douglas, Reference Aquino and Douglas2003), so it is pervasive in occupations that frequently interact with stakeholders outside the organization. In this sense, we choose people in medical occupation as our sample because of those people need to interact with patients in their daily work. Frequent interaction with patients providing more opportunities for those people to perceive patients' views about their occupation identity, and thus may evoke occupational identity threat. Given the convenience of experimental manipulation, we recruited medical students who had clinical experience as participants in studies 1 and 2. To reduce the potential bias from student samples, we invited clinicians as participants in study 3.
Study 1
Participants and design
In total, 64 (mean age = 24.45 years; 38% male) senior undergraduate and graduate students majoring in medicine with clinical experience from a university in Northwest China participated in our study in exchange for a gift (worth US$1). We randomly assigned participants to one of two between-subjects conditions (occupational identity threat: high vs. low).
Experimental procedure and materials
Participants first read the study instructions, indicated their consent, and then completed the reading task, which we adapted from Fischer et al. (Reference Fischer, Haslam and Smith2010) as the manipulation of occupational identity threat. Following this manipulation, participants responded to manipulation check questions, completed measures assessing feedback-seeking behavior, and reported their demographics. All measures were translated from English to Chinese following standard translation and back-translation procedures (Brislin, Reference Brislin, Triandis and Berry1980).
Occupational identity threat manipulation and manipulation check
The occupational identity threat manipulation task included a reading task about news reports and pictures (Fischer et al., Reference Fischer, Haslam and Smith2010). For the reading task, participants were asked to imagine themselves as the doctors in the reading material as vividly as possible. In the high occupational identity threat condition, participants were asked to read a news report and to look for a picture about patients questioning the medical skills and ethics of doctors. In the low occupational identity threat condition, participants were asked to read a news report and to look for a picture about patients defending doctors against questions about doctors' medical skills and ethics. Details of these manipulations were translated from Chinese to English and presented in the Appendix. After the reading task, participants were asked to answer three items adapted from Jetten, Postmes, and McAuliffe (Reference Jetten, Postmes and McAuliffe2002) on whether they experienced occupational identity threat (e.g., ‘I feel threatened when I read the information about doctors’; 1 = strongly disagree, 6 = strongly agree; Cronbach's α = .74).
Feedback-seeking behavior measure
We measured participants' intention to seek feedback from peers and supervisors after the reading task with 11 items adapted from Callister, Kramer, and Turban (Reference Callister, Kramer and Turban1999). Then, they indicated the extent to which they agreed with each statement after imaging that they were the doctors in the reading material (e.g., ‘I would ask my coworkers if I am meeting my job requirements’; 1 = strongly disagree, 6 = strongly agree; Cronbach's α = .71).
Results and discussion
Manipulation check
We conducted a one-way between-subjects analysis of variance (ANOVA) to check the manipulation of occupational identity threat. The results reveal that participants in the high occupational identity threat condition (M = 4.40, sd = .97) generated responses that reflected higher levels of occupational identity threat than participants in the low occupational identity threat condition (M = 2.48, sd = 1.05; F(1, 62) = 57.26, η2 = .48, p < .01). Therefore, the manipulation of occupational identity threat was effective.
Effect on feedback-seeking behavior
To test the main effect of occupational identity threat on feedback-seeking behavior, we conducted a one-way between-subjects ANOVA. The results show that participants in the high occupational identity threat condition (M = 4.42, sd = .53) reported higher levels of feedback-seeking intention than participants in the low occupational identity threat condition (M = 4.09, sd = .70; F(1, 62) = 4.29, η2 = .06, p < .05). These results provide support for hypothesis 2.
In study 1, we confirmed the validity of the manipulation of occupational identity threat and the causal relationship between occupational identity threat and feedback-seeking behavior. However, we could not verify individuals' emotional responses to occupational identity threat (hypothesis 1), the mediating effect of negative emotion (hypothesis 3), or the moderating effect of performance self-esteem (hypotheses 4 and 5). To further verify these hypotheses, we conducted studies 2 and 3.
Study 2
Participants and design
In total, 136 (mean age = 24.94 years; 60% male) senior undergraduate and graduate students majoring in medicine with clinical experience from the same university as in study 1 participated in study 2 in exchange for a gift (worth US$1). Similar to study 1, we randomly assigned participants to one of two between-subjects conditions (occupational identity threat: high vs. low).
Experimental procedure and materials
Participants first read the study instructions, indicated their consent, and then completed the reading task adapted from Fischer et al. (Reference Fischer, Haslam and Smith2010), which served as the manipulation of occupational identity threat (same manipulation as in study 1). Following this manipulation, participants responded to manipulation check questions, completed measures assessing negative emotion and feedback-seeking behavior, and reported their demographics. All measures were translated from English to Chinese following standard translation and back-translation procedures (Brislin, Reference Brislin, Triandis and Berry1980).
Occupational identity threat manipulation and manipulation check
We used the same reading task as in study 1 to manipulate participants' occupational identity threat. Following this manipulation, participants responded to the same three items (Jetten et al., Reference Jetten, Postmes and McAuliffe2002) as in study 1 to check whether they experienced occupational identity threat (Cronbach's α = .75).
Negative emotion measure
We measured negative emotion using the 10-item PANAS scale from Watson, Clark, and Tellegen (Reference Watson, Clark and Tellegen1988). Participants were asked to indicate the extent to which they agreed with each emotion they experienced after imaging that they were the doctors in the reading material (e.g., ‘afraid,’ ‘upset’; 1 = strongly disagree, 6 = strongly agree; Cronbach's α = .91).
Feedback-seeking behavior measure
We measured the intention to seek feedback using the same scale (Callister et al., Reference Callister, Kramer and Turban1999) as in study 1. Participants were asked to indicate the extent to which they agreed with each statement after imaging that they were the doctors in the reading material (e.g., ‘I would ask my coworkers if I am meeting my job requirements’; 1 = strongly disagree, 6 = strongly agree; Cronbach's α = .88).
Results and discussion
Manipulation check
Similar to study 1, we conducted a one-way between-subjects ANOVA to check the manipulation of occupational identity threat. The results revealed that participants in the high occupational identity threat condition (M = 4.55, sd = 1.09) reported higher levels of occupational identity threat than participants in the low occupational identity threat condition (M = 3.03, sd = 1.36; F(1, 134) = 51.72, η2 = .28, p < .01). Therefore, the manipulation of occupational identity threat was effective.
Effect on negative emotion
To test the main effect of occupational identity threat on negative emotion, we conducted a one-way between-subjects ANOVA. The results show that participants in the high occupational identity threat condition (M = 2.62, sd = .96) reported higher levels of negative emotion than participants in the low occupational identity threat condition (M = 1.68, sd = .87; F(1,134) = 35.69, η2 = .21, p < .01), in support of hypothesis 1.
Effect on feedback-seeking behavior
To test hypothesis 2, we also conducted a one-way between-subjects ANOVA. As hypothesized, we found that participants in the high occupational identity threat condition (M = 4.31, sd = .87) reported higher levels of feedback-seeking intention than those in the low occupational identity threat condition (M = 3.97, sd = .93; F(1, 134) = 4.78, η2 = .03, p < .05). These results provide support for hypothesis 2.
The mediating effect of negative emotion
To test the mediating effect of negative emotion, we adopted the bootstrapping analytic method (Preacher & Hayes, Reference Preacher and Hayes2008). As hypothesized, the indirect effect of occupational identity threat on feedback-seeking behavior through negative emotion was significantly positive (b = .17, p < .05, 95% confidence interval (CI) [.02, .36]). In addition, the direct effect of occupational identity threat on feedback-seeking behavior was not significant when we added negative emotion as a mediator (b = .17, ns, 95% CI [–.17, .51]). Negative emotion fully mediated the relationship between occupational identity threat and feedback-seeking behavior, in support of hypothesis 3.
In study 2, we replicated our findings in study 1 and further verified the causal relationship between occupational identity threat and negative emotion, as well as the mediating effect of negative emotion. However, studies 1 and 2 have three limitations. First, we did not test the moderating effects of performance self-esteem (hypotheses 4 and 5). Second, the samples from both studies 1 and 2 were students majoring in medicine, who may be slightly different from real medicine practitioners, causing potential bias. Third, although high in internal validity, the findings in both studies 1 and 2 might not be externally valid. To resolve these limitations, we conducted study 3.
Study 3
Procedure and participants
Participants for this study came from a hospital located in Northwest China. With the approval and assistance of senior managers and the human resources department, we invited 160 clinicians to take part in the survey. Of these, 140 clinicians completed the survey, with a response rate of 87.5%. The average age of participates was 34.14 years (sd = 9.78), 46% were male (sd = .50), 82.1% had a bachelor's degree or higher (sd = .81), 53.6% had a title of resident doctor or higher (sd = .88), and the average tenure was 9.0 years (sd = 9.50).
Measures
We followed standard translation and back-translation procedures from English to Chinese (Brislin, Reference Brislin, Triandis and Berry1980) for all measures.
Occupational identity threat
We used the 9-item scale adapted from Aquino and Douglas (Reference Aquino and Douglas2003) to assess occupational identity threat. Given that our samples are clinicians, we changed the items to reflect ‘patients’ instead of ‘coworkers.’ An adapted sample item is ‘Your patients questioned your abilities or judgments.’ All items were assessed with a 6-point scale (1 = never, 6 = always; Cronbach's α = .94).
Negative emotion
We used the 10-item PANAS scale adapted from Watson et al. (Reference Watson, Clark and Tellegen1988), which asked participants to recall their negative emotions in the past week (e.g., ‘afraid,’ ‘upset’). All items were assessed with a 6-point scale (1 = never, 6 = always; Cronbach's α = .94).
Feedback-seeking behavior
Similar to studies 1 and 2, we measured feedback-seeking behavior using 11 items adapted from Callister et al. (Reference Callister, Kramer and Turban1999). A sample item is ‘I ask my coworkers if I am meeting my job requirements.’ All items were assessed with a 6-point scale (1 = never, 6 = always; Cronbach's α = .93).
Performance self-esteem
We used the 4-item scale adapted from Heatherton and Polivy (Reference Heatherton and Polivy1991). A sample item is ‘I feel confident about my abilities.’ All items were assessed with a 6-point scale (1 = strongly disagree, 6 = strongly agree; Cronbach's α = .88).
Control variables
We controlled for participates' demographic variables, including gender, age, education, position, and tenure. Prior research has shown that these demographics are related to identity threat (Aquino & Douglas, Reference Aquino and Douglas2003).
Analytic strategy
To test our hypotheses, we used path analytic method with maximum likelihood estimation using Mplus 7.4 software (Muthén & Muthén, Reference Muthén and Muthén2015). To test the mediation effect (hypothesis 3), we used the Monte Carlo bootstrapping method with 20,000 replications to create bias-corrected CIs (Selig & Preacher, Reference Selig and Preacher2008). We then calculated the magnitude of the indirect effect at high (+1sd) and low (−1sd) levels of the moderator to verify the moderated mediation effect (hypothesis 5).
Results and discussion
Confirmatory factor analyses
We used confirmatory factor analyses (CFAs) to verify the discriminant validity of our focal variables. Given our relatively small sample size, we constructed item parcels following Brooke, Russell, and Price's (Reference Brooke, Russell and Price1988) approach. For example, we created three parcels of occupational identity threat, negative emotion, and feedback-seeking behavior. Specifically, we reduced the number of measurement items for the latent variable by averaging items with the highest and lowest loadings and then repeating the procedure until all items were assigned to one of the three parcels. The results showed that the hypothesized four-factor measurement model fit the data better (χ2(59) = 113.45, CFI = .97, TLI = .96, RMSEA = .08) than other alternative models: a three-factor model collapsing negative emotion and occupational identity threat (χ2(62) = 372.97, CFI = .82, TLI = .77, RMSEA = .19), a two-factor model further collapsing feedback-seeking behavior and performance self-esteem (χ2(64) = 830.89, CFI = .55, TLI = .45, RMSEA = .29), and a one-factor model collapsing all variables to a single factor (χ2(65) = 1,129.44, CFI = .37, TLI = .24, RMSEA = .34). These results provide support for our hypothesized measurement model. Table 1 shows the means, standard deviations, and correlations among our study variables.
Table 1. Descriptive statistics and correlations among variables
Note: N = 140. *p < .05, **p < .01, ***p < .001.
Common method variance
Considering the cross-sectional data used in this study, we adopted procedural and statistical methods to mitigate and evaluate the potential influence of common method bias (Podsakoff, MacKenzie, Lee, & Podsakoff, Reference Podsakoff, MacKenzie, Lee and Podsakoff2003). In the data collection phase, we adopted procedural remedies. Specifically, we provided assurances of anonymity and confidentiality, which is useful to reduce participants' evaluation apprehension and make them more likely to provide truthful responses (Podsakoff et al., Reference Podsakoff, MacKenzie, Lee and Podsakoff2003). In addition, we performed two statistical remedies to test the common method bias. First, we performed a single-factor procedure based on CFAs (Podsakoff et al., Reference Podsakoff, MacKenzie, Lee and Podsakoff2003). We examined the fit of the single-factor model in which all items loaded onto one factor to address the problem of common method variance. The results showed that the single-factor model was highly nonsignificant and thus can clearly be rejected (CFI = .37, TLI = .24, RMSEA = .34). Second, another well-documented set of statistical remedies for common method variance is classified as partial correlation techniques (Podsakoff et al., Reference Podsakoff, MacKenzie, Lee and Podsakoff2003). We identified a general methods factor using factor analysis; none of the significant correlations among the variables turned nonsignificant after we controlled for this factor. These results indicate that the relationships between variables are not significantly biased by common method variance (Podsakoff et al., Reference Podsakoff, MacKenzie, Lee and Podsakoff2003).
Hypotheses testing
Table 2 presents the unstandardized path coefficients. Consistent with hypothesis 1, occupational identity threat was positively related to negative emotion (b = .66, p < .001). Also consistent with hypothesis 2, occupational identity threat was positively related to feedback-seeking behavior (b = .22, p < .05). In support of hypothesis 3, the indirect effect of occupational identity threat on feedback-seeking behavior through negative emotion was significantly positive (b = .18, p < .05, 95% CI [.02, .33]). In addition, the direct effect of occupational identity threat on feedback-seeking behavior was not significant when we added negative emotion as a mediator (b = .04, ns, 95% CI [–.19, .28]). The relationship between occupational identity threat and feedback-seeking behavior was fully mediated by negative emotion. In support of hypothesis 4, we found significant interactions between occupational identity threat and performance self-esteem in predicting negative emotion (b = .15, p < .05). Figure 2 depicts the interaction pattern based on the tests of simple slopes. The effect of occupational identity threat on negative emotion was stronger when performance self-esteem was high (+1sd, b = .79, p < .01) rather than low (−1sd, b = .48, p < .01). To test hypothesis 5, we used conditional indirect effect models. As Table 3 shows, performance self-esteem moderated the indirect effect of occupational identity threat on feedback-seeking behavior through negative emotion, such that the indirect effect was stronger when performance self-esteem was high (+1sd, b = .22, p < .01, 95% CI [.07, .37]) rather than low (−1sd, b = .13, p < .05, 95% CI [.02, .25]). The difference between these indirect effects was significant (Δb = .09, p < .05, 95% CI [.01, .18]). Therefore, hypothesis 5 was supported.
Figure 2. Interactive effect of occupational identity threat and performance self-esteem on negative emotion.
Table 2. Summary of path-analytic results
Note: N = 140. *p < .05, **p < .01, ***p < .001.
Table 3. Summary of indirect effects of occupational identity threat on feedback-seeking behavior through negative emotion at low and high performance self-esteem
Study 3 extends the findings of studies 1 and 2 in three ways. First, we examined the moderating effect of performance self-esteem (hypothesis 4) and thus provided a comprehensive test of our theorized moderated mediation model (hypothesis 5). Second, using a clinician sample, we replicated our findings in studies 1 and 2 and decreased the potential bias from student samples. Third, this study was a field study, which helped strengthen the external validity of our research.
General discussion
Drawing on identity theory (Burke, Reference Burke1991, Reference Burke and Kaplan1996), we examine the emotional and behavioral consequences of occupational identity threat. The results from two experimental studies and a field study show that occupational identity threat has positive effects on negative emotion and feedback-seeking behavior. And negative emotion fully mediates the relationship between occupational identity threat and feedback-seeking behavior. Moreover, we identify performance self-esteem as a boundary condition that strengthens the positive effects of occupational identity threat on negative emotion and, thus, on feedback-seeking behavior.
Theoretical contributions
Our study makes three theoretical contributions to the literature. First, we extend identity threat literature by investigating individuals' emotional and behavioral responses to occupational identity threat. To date, studies have largely overlooked occupational identity threat, leaving the consequence of this kind of threat underexplored. In our studies, we found that individuals experience negative emotion after encountering occupational identity threat and thus are more likely to engage in feedback-seeking behaviors in the workplace. This finding suggests that identity threat does not always lead to negative consequences; instead, it promotes feedback-seeking behavior, which can be beneficial to organizational functioning (Anseel et al., Reference Anseel, Beatty, Shen, Lievens and Sackett2015).
Second, our research also contributes to identity theory. Our findings verify the function of identity theory in interpreting the essence of occupational identity threat, thus broadening the application of identity theory in identity threat literature. Our research also verifies negative emotion as an important mechanism in identity theory (Burke, Reference Burke1991, Reference Burke and Kaplan1996) by linking occupational identity threat to feedback-seeking behavior, thus responding to the call to explore the positive effects of negative emotion (Lindebaum & Jordan, Reference Lindebaum and Jordan2012). In addition, our research enriches the identity process of identity theory by capturing feedback-seeking behavior as an important output of the feedback loop (Burke, Reference Burke1991, Reference Burke and Kaplan1996).
Third, we identify performance self-esteem as a boundary condition under which occupational identity threat has different effects on individuals' emotional and behavioral responses, which contributes to both the identity threat and self-esteem literature streams. On the one hand, these findings underscore the role of performance self-esteem in influencing the extent of emotional and behavioral responses and thus encourage researchers to pay more attention to the impact of individual difference on identity threat. On the other hand, our results highlight a positive side of performance self-esteem in strengthening individuals' emotional and behavioral responses to occupational identity threat, thus lending positive empirical support to the nascent contingent self-esteem perspective in self-esteem literature (Ferris et al., Reference Ferris, Lian, Brown and Morrison2015).
Practical implications
Our research also has important implications for managerial practice. First, our findings identify negative emotion as an important mechanism linking occupational identity threat to feedback-seeking behavior. This finding suggests that managers should be alert to employees' negative emotions and try to help them translate their negative emotion into feedback-seeking behavior. To do so, we encourage managers to cultivate employees' emotion management skills, such as by providing emotional management courses or emotional guidance rooms to help employees translate negative emotions into positive behaviors. Organizations and managers should also create a positive feedback-seeking atmosphere and encourage employees to seek feedback from their colleagues and supervisors.
Second, considering the moderating effects found in our study, the positive effects of occupational identity threat on individuals' negative emotion and feedback-seeking behaviors are likely to vary by their performance self-esteem. We show that the positive effects are weaker for employees with low performance self-esteem. As such, organizations should identify employees low in performance self-esteem and take action to protect their weak egos and encourage them to seek feedback more freely.
Limitations and future directions
Despite our contributions, our research has some limitations that can be redressed by future research. First, although we identify emotional and behavioral consequences of occupational identity threat, future research still needs to explore more specific consequences. Compared with general negative emotions, we suggest focusing on specific negative emotions to better understand the emotional consequences of occupational identity threat. For example, an internal attribution of the threat may lead to feelings of disappointment and sadness, whereas external attribution may lead to feelings of annoyance or hostility (Francis & Adams, Reference Francis and Adams2019; Stets & Burke, Reference Stets and Burke2005). In addition, future studies should account for other positive behavioral consequences to help individuals cope with occupational identity threat, such as voice, creativity, and job crafting. Beyond these emotional and behavioral consequences, cognitive consequences (e.g., cognitive decline; Barber, Reference Barber2017) may also explain the effect of occupational identity threat. In this regard, we encourage future research to explore other potential consequences from other theoretical perspectives to shed greater light on occupational identity threat.
Second, research could consider the role of other potential moderators to explore the consequences of occupational identity threat. In our research, we neglected other individual factors that likely exert an influence on the moderating effects, such as attribution style (Francis & Adams, Reference Francis and Adams2019; Stets & Burke, Reference Stets and Burke2005) and identification of the threatened identity (Morris, Mok, & Mor, Reference Morris, Mok and Mor2011). In addition, some contextual factors may provoke individuals to be more proactive to cope with occupational identity threat, such as transformational leadership (Schmitt, Den Hartog, & Belschak, Reference Schmitt, Den Hartog and Belschak2016) and authentic leadership (Qian, Lin, & Chen, Reference Qian, Lin and Chen2012). Overall, we encourage future research to explore other potential moderators on the effects of occupational identity threat.
Third, we focused on people in medical occupations in terms of how they respond to occupational identity threat. Although we found similar results across people in the same occupation, we are optimistic about their generalizability. Thus, we also call for future research to examine the external validity of our findings, especially the effectiveness of occupational identity threat, in multiple occupational settings.
Finally, although we conducted two experimental studies and a field study to test the between-person effects of occupational identity threat on subsequent emotional and behavioral consequences, these studies have some limitations. Specifically, the two experimental studies adopted the same manipulation method to induce occupational identity threat. Thus, future studies could explore different manipulation procedures to further strengthen the ecological and external validity of our findings (Kouchaki & Desai, Reference Kouchaki and Desai2015). In addition, all the variables in the field study were self-reported by clinicians at the same time, which may raise concerns about common method variance and perceptual biases (Podsakoff et al., Reference Podsakoff, MacKenzie, Lee and Podsakoff2003). Although we adopted some procedural and statistical methods to mitigate these biases, future studies could conduct multi-source and time-lagged research designs to avoid them.
Conclusion
Drawing on identity theory, we examined how individuals respond to occupational identity threat. Our research reveals that individuals experience negative emotion following occupational identity threat, which in turn prompts feedback-seeking behavior. We also found that the positive effects on negative emotion and feedback-seeking behavior become stronger when individuals with a higher performance self-esteem. As such, our results extend knowledge on identity threat by capturing the emotional and behavioral consequences of occupational identity threat through the lens of identity theory (Burke, Reference Burke1991, Reference Burke and Kaplan1996). Overall, our research provides pertinent implications for research and practice.
Acknowledgements
The authors appreciate and thank Dr. Andrei Lux (Associate Editor, JMO) and five anonymous reviewers for their helpful comments and guidance to improve this manuscript. This research was supported by the grants funded by MOE (Ministry of Education in China) Project of Humanities and Social Sciences (Grant No. 16YJC630168), and by the Fundamental Research Funds for the Central Universities (2019jbkyzy024).
Appendix: Experimental materials in studies 1 and 2
High occupational identity threat condition
Shenzhen News (reporter: Haiting Wang). In recent years, the relationship between doctors and patients is becoming increasingly tight, and the incidence of medical troubles has increased. In order to explore the reasons, our reporter interviewed some patients randomly in January 2019, the following are some of their views:
Patient A: Doctors are impatient and indifferent to patients.
Every person who goes to a doctor is afraid, hoping that the doctor will treat him or her carefully. But, how many doctors can comfort their patients and treat them carefully? Many doctors don't wait for patients to finish their sentences before prescribing drugs or medical tests. I think there is a big gap between the current doctor and the old ‘angel in white’ image.
Patient B: No matter what the disease is, all kinds of medical tests will be performed.
No matter how serious or minor of your disease, doctors always ask you to pay for tests. Once my daughter had a stomachache, she was asked by her doctor to run a series of medical tests. Finally, her doctor said she was ascariasis and prescribed some medicine for her. In my opinion, ascariasis is a minor disease that can be judged by clinical experiences. But, why doctors like asking their patients to pay a lot of money for a series of medical tests? Is it a decline in doctors' medical ethics, or a lack of confidence in their medical skill?
Patient C: Doctors like using expensive drugs, no matter how effective they are.
Doctors now prescribe a lot of expensive imported drugs for minor ailments. I got a sty once, and my doctor gave me some Belgian and French medicines. I remembered that a bottle of eye drops costed 200 yuan, it was pricy for me. Some people may say that the efficacy of imported drugs would be better, I think the efficacy gap between imported drugs and domestic drugs for some minor diseases may be very small, but the price gap is too large. Therefore, I think doctors are just looking for ways to make money from their patients.
Low occupational identity threat condition
Shenzhen News (reporter: Haitao Sui). In recent years, the relationship between doctors and patients is getting better, and the incidence of medical troubles has decreased. In order to find out the reasons, our reporter interviewed some patients randomly in March 2019, the following are some of their views:
Patient A: We should be considerate of the doctor's hard work.
I was touched by the recent report of ‘The most Beautiful Doctor,’ in which a female doctor fell asleep in the closet after an operation because she was too tired. From this report, I also know that doctors sometimes have to spend more than 10 hrs for one operation, which is really hard to imagine. I think patients should understand that it is not easy for doctors to keep smiling all the time under such a high work pressure.
Patient B: On the one hand, patients blame their doctors for prescribing lots of tests; on the other hand, patients wouldn't trust their doctors without test results.
I've heard people around me complained about being asked by their doctors to do a lot of medical tests no matter what kind of disease. Some people think these doctors are getting kickbacks, whereas others think the doctors are incompetent. In fact, I don't think that's the case. I am not sure about doctors taking kickbacks, but I don't think it makes sense to doubt doctor's competence. The former doctor does not ask you to do some tests, that is because there wasn't such a medical condition in the past. Now, we have this condition, using medical tests can help doctors diagnose the diseases, which can also reassure patients.
Patient C: Costliness in seeking medical treatment cannot be blamed on doctors alone.
Some patients think doctors prescribe expensive drugs and tests deliberately. In fact, I think, no matter how expensive the drugs and tests are, it must be right. If the expensive drugs and tests have nothing to do with your disease, the doctor will not dare to prescribe them to you. The problem of expensive medical treatment mainly depends on the government and social security. Doctors are only one part of the medical system, and it's not appropriate to blame them for anything.
(Note: Parts of these sentences for the experimental materials of occupational identity threat were taken from several real newspaper articles and online sources from China. To make above materials more convinced, you can also attach one related picture with news reports in both two conditions.)
Na Yang (PhD), School of Business, Renmin University of China. Her research focuses on identity, ethics, and organizational behavior.
Ruoyong Zhang (PhD) is a professor in the School of Management, Lanzhou University. His research focuses on proactive behavior, identity, emotion in workplace, and leadership.
