Professional and religious identity conflict: individual and organizational dynamics in ethically-charged circumstances

ABSTRACT In this article, we investigated individual-level mechanisms and organizational-level conditions to explain the emergence and unfolding of professionals’ identity conflict in ethically-charged situations. Specifically, we examined identity conflict in doctors as triggered by discrepant work and non-work identity values of moral nature, namely their professional and religious identity values. We conducted a multilevel analysis in a two-time-lagged, questionnaire-based study (N = 120 doctors) in 22 National Healthcare Service Trusts in England. We found that professional-religious identity conflict has a negative influence on psychological well-being, but does not mediate the relationship between the interplay of religious and professional identity values and psychological well-being. Furthermore, whilst hospital peer social support positively buffered the negative relationship between identity conflict and psychological well-being, hospital ethical climate negatively moderated this relationship. By incorporating both intrapersonal processes and socio-ethical boundary conditions surrounding identity conflict, we extended research on identity dynamics within organizational contexts shedding light on the implications of such dynamics for the psychological well-being of professionals.


Introduction
Increasing uncertainty and complexity in today's healthcare organizations (Borgstrom & Walter, 2015) have challenged doctors' clinical practice and decision making, especially in ethically-charged circumstances (Curlin et al., 2007).Legal changes in medical regulations and advancements in medical technology have blurred the right or wrong perception of previously straightforward courses of actions (Karnik & Kanekar, 2016) creating room for subjective interpretations of situations of ethical collision (Birchley, 2012).Consequently, doctors can experience identity conflict between their work and other non-work identities, such as professional and religious identities (Lawrence & Curlin, 2009).
As other social identities, professional and religious identities help individuals to feel a sense of belonging to a group, perform societal roles and act according to their community's expectations (Borgstrom & Walter, 2015;Tracey, 2012).They can thus significantly influence people's behavior (Watson, 2008).However, compared to other identities, these identities are characterized by a well-defined and strict ethical code of conduct (Wright et al., 2017;Ysseldyk et al., 2010).As a result, whilst when in harmony identities may enhance each other, when incongruent but still salient and central, they are more likely to trigger identity conflict (Héliot et al., 2020;Ramarajan et al., 2017).
In the medical context, this conflict can be associated with ethical dilemmas (Birchley, 2012;Carminati & Héliot, 2021;Kälvemark et al., 2004) and profoundly impact doctors' psychological outcomes, increasing their psychological burn-out rates as well as lowering their well-being and performance (Patel et al., 2018;Scheepers et al., 2020).Alarmingly, psychologically-unwell doctors are more likely to exhibit low levels of professionalism and provide suboptimal patient care, such as ordering unnecessary care or committing medical errors (Brady et al., 2018;Hayes et al., 2017).This can ultimately lead to an increased risk of mortality among patients which can profoundly affect the overall quality of the healthcare service (Halbesleben & Rathert, 2008).
And yet, how such identity conflict in doctors arises, unfolds, and influences their psychological well-being remains unexplored (Carminati & Héliot, 2021).More specifically, from identity theory approaches, we know little of the role of identity values and contextual factors in affecting identity conflict dynamics (Alvesson et al., 2008;Islam, 2020).On the one hand, values are constituting parts of identities and provide principles and imperatives that help individuals to navigate life (Schwartz, 2015).On the other hand, contextual factors, such as organizational ethical climate and peer social support, can influence the salience and centrality of these values and, seen through the lens of interunit conflict (Horton et al., 2014), serve the purpose of reducing identity conflict by suggesting an alignment with one of the conflicting identities.
Since how these individual-level mechanisms and organizational-level conditions can explain identity conflict dynamics is still under-researched (Carminati & Héliot, 2021), in this article we contribute to the literature on identity dynamics within organizations in two ways.Firstly, we extend current knowledge of identity conflict dynamics by investigating at the individual-level how professionals' work and non-work identity values can influence individuals' psychological well-being.Secondly, through Social Information Processing Theory (Salancik & Pfeffer, 1978), we incorporate the crucial role of organizational-level forces into individual-level identity dynamics by accounting for the socialcontextual embeddedness of individual processes (Islam, 2020).Hence, we investigate how "extra-individual forces" (Alvesson et al., 2008, p. 18) can interact with "intra-psychic" processes (Islam, 2020, p. 2) by evaluating the interactive role of organizational ethical climate and peer social support across organizations as socio-ethical boundary conditions in identity conflict dynamics.In so doing, we address the need for multilevel approaches in identity studies to bridge intrapersonal and organizational levels (Carminati & Héliot, 2021;Horton et al., 2014;Islam;2020).
In the remaining part of this article, firstly, we provide definitions of the key concepts and explanations of their relationships.Secondly, we discuss the methodology and the results.Lastly, a discussion of the findings of this study as well as its limitations will close the paper pointing to future research and overall conclusions.

Intrapersonal processes of identity conflict
According to identity theory approaches, over the course of their lives, individuals take on multiple identities reflecting who they are and how they should act within a community (Stryker & Serpe, 1982;Tajfel & Turner, 1986).Usually only one identity is activated at a time by superseding the others via its salience and centrality 1 to the contextual environment and the individual.However, recent strands in identity research have underlined that in some circumstances more than one identity can be simultaneously triggered (Ramarajan et al., 2017).If salient and central, but incongruent identities are activated at the same time, individuals may experience identity conflict.Although identity conflict is defined as discrepancy "between values, beliefs, norms and demands inherent in individual and group identities" (Horton et al., 2014, p. 6) and arises when individuals feel they must satisfy identity-based expectations (Ramarajan, 2014), the specific role played by values in such identity conflicts is often ignored.Still, in the workplace, values are vital and particularly core to professionals (Wright et al., 2017) since they may drive individuals toward the right action and direction in situations of identity conflicts or ethical dilemmas (Kälvemark et al., 2004;Huhtala et al., 2020).In this sense, values are morally-charged.
As members of a prototypical profession (Pratt et al., 2006), doctors may face identity conflict involving values of moral nature on a daily basis, given the frequency they encounter ethically-charged circumstances in their clinical and medical practice (Hurst et al., 2007).These challenging circumstances tend to question individuals' values and actions (Birchley, 2012;Genuis & Lipp, 2013), leading doctors to experience conflicts between, for instance, duties toward their professional values and duties toward other non-work identity beliefs, among which religious values (Giubilini, 2016;Carminati & Héliot, 2021).Religious values constitute individuals' religious identity and, by informing strong ethical principles, can directly or indirectly influence doctors' decision making and practice (Gebert et al., 2014).Consequently, the application of doctors' code of medical practice is not straightforward and doctors may find themselves under enormous cognitive, emotional and moral pressures to maintain a fine balance between high standards of professionalism and respect toward their own conscience (Borgstrom & Walter, 2015;DuVal et al., 2004;Giubilini, 2016).
Albeit not always clashing, 2 medical studies have explored the potential conflict arising in doctors between religious and professional values in ethically-charged situations (Curlin et al., 2007) and its negative repercussions on doctors' psychological well-being (Genuis & Lipp, 2013;Hayes et al., 2017).Whilst often neglecting the professional-religious interplay in work contexts (Héliot et al., 2020;Tracey, 2012), Management-Organizational Studies research has similarly investigated psychological outcomes of identity conflict mostly in terms of burnout, stress, exhaustion and job satisfaction (Brook et al., 2008;Demerouti et al., 2016;Karelaia & Guillén, 2014;Rabinovich & Morton, 2016;Settles, 2004).Less attention has been paid to psychological well-being, even though it represents a key psychological response of individuals to complex situations and challenging social-cultural behaviors (Ryff & Singer, 2008), and its negligence can significantly endanger patient care (Brady et al., 2018;Halbesleben & Rathert, 2008;Scheepers et al., 2020).
Building on the above, we suggest that the severity of the identity conflict and its influence on individuals' psychological well-being can vary depending on how salient and central, but incongruent, religious and professional identity values are.In their influential work about women leaders' identity conflict, Karelaia and Guillén (2014) showed that identity conflict can function as a mediator between positive social identities and stress, life satisfaction and affective and socio-normative motivation.However, no interaction between identities or identity values was considered as affecting identity conflict, and psychological well-being was not explored as a specific outcome.Similarly, Settles (2004) investigated the role of the interaction of central identities, i.e., woman and scientist identities, as a predictor of identity interference, and as moderator of the relationship between interference and outcomes such as performance, self-esteem, depression and life satisfaction.Notably, identity values and a mediation path were not assessed and psychological well-being was not included as a defined criterion variable.
Yet, in ethically-charged situations, doctors' salient and central but incongruent professional and religious identity values could lead to identity conflict (Curlin et al., 2007, Genuis & Lipp, 2013).This conflict, in turn, by triggering internal dissonance and inconsistency (Horton et al., 2014), could lower doctors' levels of psychological well-being.Hence, we put forward the following hypothesis: Hypothesis 1: Doctors' identity conflict, triggered by the interaction of incongruent professional and religious identity values, mediates and negatively affects their psychological well-being.

Contextual boundary conditions: ethical climate and peer social support
Since human interactions do not happen in a vacuum (Treviño et al., 2008), the social and organizational contexts in which these interactions take place can considerably shape individuals' attitude and behavior (Héliot et al., 2020;Johns, 2017).Such contextual, extraindividual organizational forces could be seen through the lens of inter-unit conflicts, defined as "conflicts in the values, beliefs norms, and expectations held by different individuals or groups within a collective identity" (Horton et al., 2014, p. 10).Consequently, these contextual forces can affect identity conflict dynamics and help people to understand which identity and values should be either put on hold or prioritized (Alvesson et al., 2008;Huhtala et al., 2020) ultimately determining the direction toward which people's moral compass should turn (Weaver, 2006).
However, despite its importance, what comprises organizational context and its influence on the subjective experiences of identity conflict has been rather neglected by identity research in Management-Organizational Studies (Alvesson et al., 2008;Johns, 2017).Thus, we suggest that Social Information Processing Theory (Salancik & Pfeffer, 1978) can help bridging this void offering theoretical explanation on how extra-personal forces impact decision making in situation of moral conflicts.Furthermore, medical studies on ethical dilemmas have pointed out some extra-individual forces that could impact doctors' identity conflict and decision making in ethically-charged situations (Hurst et al., 2007;Woo et al., 2006).Indeed, even if the medical profession is often considered a "solo-person" job, hospital ethical climate and peer social support stood out as vital factors in these situations (Carminati & Héliot, 2022;Deshpande & Joseph, 2009;Hart, 2005;Hurst, 2005;Woo et al., 2006).Hence, we see ethical climate and peer social support as potential organizational-level boundary conditions in the relationship between doctors' identity conflict and their psychological well-being.Figure 1 depicts the overall model tested in this study.

Hospital ethical climate level and strength
Ethical climate in organizations has been conceptualized as employees' shared perceptions of the policies, practices and procedures that an organization rewards and supports regarding ethics (Arnaud & Schminke, 2012;Martin & Cullen, 2006).The emergence of organizational ethical climate can exist as an aggregated measure only when perceptual agreement among individuals also exists (González-Romá et al., 2002;Otaye-Ebede et al. 2020).We thus conceptualized perceptual agreement following both a "direct consensus composition model" and a "dispersion model" (Chan, 1998).
In a direct consensus composition model, employees' average score on organizational climate represents organizational ethical climate level, or quality (Schminke et al., 2005).According to Social Information Processing Theory (Salancik & Pfeffer, 1978), by allowing individuals to collect identity-relevant signals about what is ethically treasured and more desirable within a hospital, organizational ethical climate can help doctors to identify what values should be prioritized) thus influencing their decision making and behavior (Schneider et al., 2013).
On the contrary, in a dispersion model the degree of variability among individuals in climate perceptions depicts organizational ethical climate strength (González-Romá et al., 2002).Climate strength has an incremental role on climate level and such incremental role has been justified by Situational Strength Theory (Mischel, 1973).In line with this theory, when individuals perceive a strong ethical climate within their organization, they will tend to respond to moral issues in a unanimous way due to the high certainty and consistency of the contextual cues and vice versa (Shin, 2012).Management-Organizational Studies and medical research has considerably explored the direct, indirect and moderating effects of psychological and organizational ethical climate on numerous outcomes, such as job satisfaction, performance, ethical behavior and moral obligations (Deshpande & Joseph, 2009;González-Romá et al., 2002;Otaye-Ebede et al. 2020;Wang and Hsieh, 2013), turnover intentions, stress and burnout (Hart, 2005), as well as organizational commitment, citizen behavior and helping behavior (Cullen et al., 2003;Shin, 2012;Simha & Cullen, 2012).
However, despite numerous calls to solve inconsistency, the joined influence of both organizational ethical climate level and strength has scarcely been explored as a moderating factor (Schneider et al., 2013;Shin, 2012).Therefore, we argue that when doctors perceive high degrees of ethical climate level and strength in their hospital, they will tackle identity conflict more efficiently, since the combined effect of high quality and agreement of organizational ethical climate can help them to prioritize the identity that they should deem more in line with organization's rules and practices (Victor & Cullen, 1988).Organizational ethical climate thus works as a boundary condition that can help doctors to ameliorate the negative relationship between their identity conflict stemming from professional and religious values and psychological well-being.Hence, we suggest the following hypothesis: Hypothesis 2: The interaction between hospital ethical climate level and strength positively moderates the relationship between doctors' professional/religious identity conflict and their psychological well-being.

Hospital peers social support level and strength
Peer social support constitutes an interactive process based on the concept of reciprocity between individuals and is considered to be an essential resource for employees' wellbeing (Zeijen et al., 2020).It is grounded in individuals' expectation of receiving support in moments of need and uncertainty (Schwarzer & Knoll, 2007), of which identity conflict can be considered an example (Horton et al., 2014).
As for hospital ethical climate, peer social support can be conceptualized at the organizational level in terms of peer social support level and strength.These concepts indicate the average score and the variability, respectively, of the shared perception of help that individuals can receive from their peers across the whole organization (Kozlowski & Klein, 2000).Indeed, providing individuals with cues and directions to understand those values that are treasured by the peer community in the workplace, peer social support can guide individuals' decisions toward the more appropriate behavior (Salancik & Pfeffer, 1978) and help them to solve their inner struggle.
Management-Organizational Studies research has reported that workplace social support can influence individuals' appraisal of stressful situations and prevent maladaptive outcomes (Chiaburu & Harrison, 2008), providing resources to cope with depression and stress (Wang & Hsieh, 2013).Studies have particularly showed a positive association between colleagues' social support and employee commitment and development, job satisfaction, as well as productivity, job performance and job involvement (Chiaburu & Harrison, 2008;Zeijen et al., 2020).
Medical research has similarly shown that interactions with peers can significantly influence doctors' cognitions, emotions and behavior (Holman et al., 2018, LeBlanc et al., 2007).Doctors are increasingly involved in multidisciplinary group practices that imply medical collaboration across the hospital (Hewett et al., 2009), especially when facing personal struggles in ethically-charged situations (DuVal et al., 2004).Indeed, doctors do rely on peer social support for help when experiencing ethical dilemmas (Kälvemark et al., 2004), difficult emotions (Woo et al., 2006) and seeking validation, advice or fresh perspectives in critical situations (Hurst et al., 2007).
Therefore, we argue that the joined effect of high degrees of hospital peer social support level and strength may help doctors to overcome their ethical conflict more promptly, since it would allow them to gather high quality and agreement information on what should be deemed more important within the organization (Salancik & Pfeffer, 1978).By signaling those values, attitudes and behaviors, the combination of peer social support level and strength can buffer the negative effects of doctors' identity conflict on their psychological well-being in challenging situations (Salancik & Pfeffer, 1978).Hence, we suggest the following hypothesis: Hypothesis 3: The interaction between hospital peer social support level and strength will moderate the relationship between doctors' identity conflict and psychological wellbeing.

Procedure and sample
On the basis of the services provided (e.g., cancer care, pain management, palliative care), we invited 61 Research and Development departments of hospitals of National Healthcare Service (NHS) Foundation Trusts (thereafter referred as hospitals) across England.A total of 38 hospitals agreed to take part in our research.To participate in the study doctors had to have experienced identity conflicts or dilemmas in an ethically-charged situation and define themselves as having a certain faith (e.g., Christian, Muslim, Jewish).Doctors' anonymous and voluntary participation was assured, and ethical approvals were obtained both from the University of Surrey Ethics Committee and the Health Research Authority.
To reduce common method biases (Podsakoff et al., 2012), we conducted a timelagged, questionnaire-based study in which a two-part online survey was administered to participants, disseminating part 1 (T1) and part 2 (T2) of the survey roughly six weeks apart.We measured control variables (e.g., age, gender, religiosity and medical specialty), the independent variables (religious identity and professional identity) and one of the moderators (peer social support) from the first point of measurement (part 1, T1), and the dependent variable (psychological well-being), the other moderator (organizational ethical climate) and identity conflict from the second point of measurement (part 2, T2).
The first round of survey (part 1) provided an initial sample of 236 doctors in 38 hospitals.Hospitals with a minimum of three doctors were included in the analysis (Kreft & de Leeuw, 1998).We excluded 5 hospitals (corresponding to 8 unusable surveys) that did not reach this threshold and other 23 questionnaires which were left incomplete.Our final part 1 sample at T1 included 205 doctors and 33 hospitals.The second round of data collection (part 2) yielded an initial sample of 138 questionnaires collected in 28 hospitals.We eliminated any questionnaires that were incomplete or could not be matched with a doctor's previous questionnaire, as well as responses from hospitals with less than three doctors.Our final sample was comprised of 120 doctors nested in 22 hospitals, with an average cluster size of 5.455 doctors per hospital (SD = 2.385, ranging from 3 to 12 doctors).
Participants were between 26 and 62 years old (M age = 43.48,SD age = 9.623), 45% were male and 55% were female.Participants had an average of 19.2 years of working experience (SD time into practice = 9.761) and 70.83% defined themselves as consultants, 20% as junior doctors and 9.17% as senior doctors.Overall, 77% described themselves as Christians, 7.5% of doctors were Muslims, 3.33% were Buddhists, 4.17% were Jewish and 8% answered "other religions".The sample was diverse in terms of the medical specialty they practiced in as 15.83% worked in Palliative Care, 12.5% in Anesthesia, 4.17% in cardiology, 9.17% in Geriatrics, 8.33% in Pediatrics, 2.5% in Gynecology, 3.33% in Hematology, 1.67% in Neurology, 8.33% in Oncology and 34.17% in specialties such as Internal Medicine, Emergency, Critical/Intense Care and Surgery.

Measures
Since the goal of our study was to evaluate intrapersonal dynamics and doctors' perceptions, self-report measures were deemed the most appropriate to address our goal (Podsakoff et al., 2012).We employed well-established scales in Management-Organizational Studies research, and all items used a 7-point Likert-type scale anchored at 1 = strongly disagree and 7 = strongly agree, unless otherwise pointed out (see Supplementary Appendix A for full scales).

Salience and centrality of religious and professional identity values
We measured the salience and centrality of doctors' religious (α = 0.844) and professional (α = 0.884) identity values by adapting the 6 items of the Identity Salience and Centrality Scale (Cameron, 2004).Sample items include: (1) "The values related to being a religious person/a doctor are an important reflection of who I am" and (2) "In my everyday life, I often think about what it means to have religious/medical values".

Identity conflict
Identity conflict was measured using a 6-item scale (α = 0.798), adapted from the Woman/ Leader Identity Conflict Scale (Karelaia & Guillén, 2014).The items reflect our operational definition of identity conflict (see above) and included: (1) "I feel that other doctors do not take me seriously because of my religious values" and (2) "I run into obstacles in my role as a doctor because of my religious values".

Psychological well-being
We assessed psychological well-being using the short version (12 items) (α = 0.739) from the 6-point Psychological Well-being Scale (Ryff, 2014), covering 4 of the 6 subscales: (1) Autonomy (e.g., "I judge myself by what I think is important, not by the values of what others think is important"); (2) Environmental Mastery (e.g., "I am quite good at managing the many responsibilities of my daily life"); (3) Personal Growth (e.g., "I think it is important to have new experiences that challenge how you think about yourself and the world"); (4) Self-Acceptance (e.g., "When I look at the story of my life, I am pleased with how things have turned out").

Hospital ethical climate
We measured ethical climate using the 6 items (α = 0.792) of the Rules and Law and Code aspects of the 6-point Ethical Climate Scale (ECQ, Victor & Cullen's, 1988) since they best represent the essence of ethical climate as suggested by Shin (2012).Sample items are: (1) "In this hospital, people are expected to strictly follow legal or professional standards"; (2) "In this hospital, everyone is expected to stick by hospitals rules and procedures".For hospitals' shared perceptions of ethical climate, individual-level ratings were aggregated according to both consensus and dispersion models (Chan, 1998;LeBreton & Senter, 2008).

Hospital peer social support
Peer social support was assessed with the 5-item (α = 0.866) 5-point Social Support Scale (Frese, 1999).Sample items are: (1) "I can rely on my peers when things get tough at work"; and (2) "My peers are helpful to me in getting my job done".For hospitals' shared perceptions of peer social support, individual-level ratings were aggregated according to both consensus and dispersion models (Chen, 1998;LeBreton & Senter, 2008).

Control variables
We controlled for the religion participants identified themselves with since we were interested in maintaining a broad perspective on religious identity and values (Ysseldyk et al., 2010), rather than investigating specific religions and their individual association with identity conflict.Since the majority of participants (77%) defined them-selves as Christians, we dichotomized the variable to be Christian Religions = 1, Other Religions = 2. Additionally, we filtered out gender and age because previous research has documented moderate to strong relationships between these variables, identity conflicts and well-being (Deshpande et al., 2006).Lastly, we controlled for doctors' medical specialties since medical studies have underlined differences in doctors' perception and response to identity conflict depending on their medical expertise (Hurst, 2005).The means, standard deviations and correlations of measures at individual and team levels are reported in Table 1.

Levels of analysis and analytic strategy
The focal units of our study were the individuals and the hospitals they worked in.At the micro, individual level, we were interested in doctors' perceptions of identity conflict as triggered by the salience and centrality of professional and religious identity values and leading to low levels of psychological well-being.At the macro, organizational level, we looked at the influence of doctors' shared perception of hospital ethical climate and peer social support on the relationship between identity conflict and psychological well-being.ͨ Group Size was used as a control variable for our nested data.Variables at the team level were assigned to individuals and correlated at the individual level (rows 1-8).The magnitude of these correlations accurately reflects the relationships at their respective level of analysis.However, due to the nested nature of our data, standards errors might be biased, and significance level should be interpreted cautiously.
We thus implemented a multilevel analysis using Structure Equation Modeling (SEM).Even if the ICC(1) of our dependent variable (psychological well-being) was very small, ICC(1) = 0.003 (James et al., 1984, LeBreton & Senter, 2008), studies have argued that strong group-level relationships, not evident in the lower level data, can still be found (Bliese & Halverson, 1998), and that researchers should go beyond typical cutoff points when random effects are hypothesized (LeBreton & Senter, 2008).More importantly, we could not ignore the nested nature of our data which is accurately accounted for by hierarchical linear modeling (Kozlowski et al., 2016).Hence, we proceeded with our multilevel analysis.
We divided our analysis in two main parts.First, we tested Hypothesis 1 and the mediating role of identity conflict between the interaction of doctors' professional and religious identity salience and centrality and psychological well-being.Second, we assessed Hypothesis 2 and Hypothesis 3, hence the combined moderating effect of organizational ethical climate level and strength, and peer social support level and strength on the relationship between identity conflict and psychological well-being.We adopted a maximum likelihood approach and conducted the analyses using the statistical software package MPlus7.

Confirmatory factor analysis
We conducted a confirmatory factor analysis (CFA) to test the validity and distinctiveness of our study measures, for both the models with organizational ethical climate and peer social support as level-2 moderators.We reported exact model fit (χ 2 ), degrees of freedom and three approximate fit indices (CFI, RMSEA and SRMR) following well-established cutoff points, i.e., CFI > 0.95, RMSEA < 0.06 and SRMS < 0.08 (Hu & Bentler, 1999).

Measurement model
For both the models with organizational ethical climate and peer social support, all latent variables were modeled using the observed items at an individual level.In both models, we tested a 5-factor model that showed a good overall fit (χ 2 (395) Ethical Climate = 624.412,CFI = 0.95, RMSEA = 0.04, SRMR = 0.07; χ 2 (367) Peer Social Support = 584.688,CFI = 0.94, RMSEA = 0.04, SRMR = 0.07).As showed in Tables 2 and 3, we tested alternative models to consider the discriminant validity of the constructs.Given the correlation between professional and religious identity salience and centrality (r = -0.610,p < 0.01) and between identity conflict and psychological well-being (r = -0.278,p < 0.01), we assessed two different 4-factor models combining the correlated variables.We also tested a 1-factor model.For all the alternatives models, the chi-square difference test provided evidence that our hypothesized 5-factor model was the best fit for our data and demonstrated sufficient discriminant validity between the constructs.

Aggregation
To test our second and third hypotheses -regarding the moderating role of the interaction between organizational ethical climate level and strength (Hypothesis 2) and organizational peer social support level and strength (Hypothesis 3) in the relationship between identity conflict and psychological well-being-, we aggregated participants' scores on the organizational ethical climate and peer social support scale at individual level to the hospital-unit level following LeBreton and Senter's (2008) recommendations.We calculated rwg (J) (James et al., 1984), applying three different distributions (i.e., uniform, slightly skewed and triangular), and AD M(J) (Burke & Dunlap, 2002) as measures of agreement within group.We also calculated ICC(1) to evaluate the proportion of the variance in the study measures that could be explained by hospital membership (LeBreton & Senter, 2008).
For ethical climate, the application of a uniform distribution yielded no out-of-range values, while the slightly skewed outlined 2 out-of-range results and the triangular distribution revealed instead 9 out-of-range values.We used the uniform distribution with a minimum standard of 0.51 as the cutoff-value, indicating moderate agreement.The average rwg (J) for ethical climate within hospitals was 0.82, indicating a strong agreement (LeBreton & Senter, 2008), and ranged from 0.57 to 0.96.Regarding the AD M(J) , the average value of  the AD M(J) was 0.64 and ranged between 0.27 to 0.91.Thus, all the values were below the threshold of 1 suggested by Burke and Dunlap (2002) for a 6-point Likert-type scale.
For peer social support, the application of a uniform distribution led to none out-of-range values, while the slightly skewed and triangular distributions revealed 2 out-of-range values.We used the uniform distribution with a minimum standard of .51 as the cutoff-value, indicating moderate agreement.The average rwg (J) for social support was 0.87, indicating a strong agreement, and ranged between 0.69 and 0.93 across the 22 hospitals.The average AD M(J) was 0.49, ranging between 0.33 and 0.67, with all values significantly below the cutoff point of 0.8 suggested for a 5-point Likert-type scale (Burke & Dunlap, 2002).
Lastly, all our ICC(1) scores were bigger than 0 and significant at p < 0.05 (Bliese & Halverson, 1998).These results supported the aggregation of doctors' shared perception of their hospital ethical climate and peer social support (see Supplementary Appendix B, Table S2).

Hypotheses testing
In the first part of our analysis, we tested Hypothesis 1 and evaluated the mediating role of identity conflict between the interplay of doctors' professional and religious identity values and psychological well-being.Figure 2 illustrates our model and Table 4 the results of the mediated-moderation analysis.
After fitting a Null Model (Model 1, Table 4) and a model with control variables (Model 2, Table 4), we analyzed the relationship between salient and central professional and religious identity values and psychological well-being.The results show a negative relationship between the interaction of professional and religious identity values and doctors' psychological well-being (β = -0.059,p = 0.011) (Model 3, Table 4, see Figure 3).In particular, when religious and professional identity salience and centrality were both high, participants' psychological well-being was lower (see Supplementary Appendix B, Table 3, for a detailed breakdown of the slope analysis).
A plot of the interaction is illustrated in Figure 4.
We then tested the relationship between identity conflict and psychological well-being (Model 4).Our data showed that a negative and significant relationship existed between conflict and PWB (β = -0.202,p = 0.009).Lastly, we tested the full model and thus the mediating role of identity conflict between the interaction of doctors' salient and central professional and religious identity values and their psychological well-being.Contrary to    our hypothesis, we found that identity conflict was still negatively related to psychological well-being (β = -0.228,p = 0.003) (Figure 5), but it did not mediate the interaction between doctors' salient and central professional and religious values and psychological well-being (β = 0.051, p = 0.055) 3 (see Supplementary Appendix B, Table S4).Hence, our first hypothesis was not supported.
For Hypotheses 2 and Hypothesis 3 we tested whether the joined effect of organizational ethical climate level and strength, as well as the joined effect of organizational peer social support level and strength could moderate the negative relationship between doctors' identity conflict and their psychological well-being.We tested these hypotheses in two separate models that build on the model depicted for Hypothesis 1.For clarity and parsimony in Tables 5 and 6 we reported only the results relevant to Hypotheses 2 and Hypothesis 3.
To test Hypothesis 2, we followed three main steps.In the first step (Model 3, Table 5), we assessed ethical climate level and its cross-level interaction with identity conflict.Neither the direct effect of organizational ethical climate level on psychological wellbeing (β = -0.389,p = 0.488), nor the interaction between identity conflict and hospital ethical climate level on psychological well-being (β = -0.051,p = 0.589) were significant.In the second step (Model 4), we obtained a similar pattern of results showing no significant relationships for both the direct effect of organizational ethical climate strength on psychological well-being (β = -0.011,p = 0.985) and the interaction between identity conflict and hospital ethical climate strength on psychological well-being (β = 0.029, p = 0.786).Nonetheless, when we tested Hypothesis 2 and thus the joined moderating effect of organizational ethical climate level and strength (Model 5), we found that the 3-way interaction between identity conflict, hospital ethical climate level and hospital ethical climate strength was highly significant but negative (β = -0.397,p = 0.002).
Specifically, when employees experience identity conflict, the joined influence of the organizational ethical climate quality and agreement negatively influences psychological well-being rather than the predicted positive relationship.Therefore, Hypothesis 2 was only partially supported.Figure 6 illustrates the slope analysis of the 3-way cross-level interactions (please see Supplementary Appendix B, Table S5 for the breakdown of the slope analyses of the 3-way interaction for hospital ethical climate).To address Hypotheses 3, which examines the moderating role of organizational peer social support, level and strength, we followed the aforementioned three-step process (see Table 6).For the first two steps (Model 3 and 4, Table 6) testing the direct effect of organizational peer social support level on psychological well-being and its interaction with identity conflict, as well as the direct effect of organizational peer social support strength on psychological well-being and its interaction with identity conflict, respectively, we obtained no significant results.Nonetheless, the 3-way interaction (Model 5) between identity conflict, organizational peer social support and peer social support strength was significant and positive (β = 0.361, p = 0.029).Hypothesis 3 was thus fully supported, showing that when organizational peer social support is high in both its quality and agreement, this can considerably reduce the negative consequences of identity conflict on employees' psychological well-being.Figure 7 represents the plot of the 3-way interaction for organizational peer social support (please see Supplementary Appendix B, Table S6 for the breakdown of the slope analyses of the 3-way interaction for peer social support).
In addition, we conducted an analysis without control variables (Becker, 2005).We obtained similar findings (please see Supplementary Appendix B, Figures S2 and S3), giving us even more confidence in the robustness of the estimates yielded by our model with control variables (Podsakoff et al., 2012).

Discussion
The goal of this article was to extend research on identity dynamics within organizations by accounting for both intrapersonal mechanisms and extra-personal processes of identity conflict.More specifically, we investigated how doctors' identity conflict, triggered by the salience and centrality of professional-religious identity values, could affect their psychological well-being, as well as integrated extra-individual boundary conditions, namely hospital ethical climate and peer social support, that could ameliorate the relationship between identity conflict and psychological well-being.Firstly, our findings show that identity values are crucial in identity dynamics and that the professional-religious identity conflict experienced by doctors in ethically-charged situations can negatively influence their psychological well-being.Albeit apparently intuitive and in line with previous quantitative studies (e.g., Brook et al., 2008;Karelaia & Guillén, 2014;Settles, 2004), these results do not corroborate other more recent qualitative identity research pointing to positive psychological consequences of identity conflict, such as personal growth, self-esteem and learning (Carminati & Héliot, 2022).This is interesting since, echoing the "too much of a bad thing effect", this contradictory evidence seems to show that clashing identity values may have both negative and positive effects, depending on whether short-or long-run effects are considered.This might be also related to the fact that in the literature identity conflict is not only considered as an intrapersonal state of the mind (Alvesson et al., 2008), but also   a dynamic process that may take time and also evolve over time (Horton et al., 2014).In this regard, our mediated-moderation hypothesis about the role of identity conflict as the step through which important interacting values can influence individuals' well-being was not supported.Even though this would have added further evidence of the dynamics underpinning identity conflict (see: Brook et al., 2008;Settles, 2004), this result might be related to the limited power in our study.This means that our non-significant effects are not necessarily evidence against our hypothesis, and pave the way to further investigations into the actual role and process concerning identity conflict.Furthermore, we found that since such identity values differ in terms of their salience and centrality, they can influence how individuals experienced the severity of the identity clash and thus its consequences on professionals' psychological wellbeing.This is in line with previous research suggesting not only that identity (in) congruence varies on a continuum of different degrees and has significant impacts on a variety of different outcomes (Héliot et al., 2020); but also that scholars need to understand identity dynamics at a deeper level (i.e., the value level) and discern between latent conflicts (i.e., those that go unnoticed) and manifest conflicts (i.e., those that are recognized and prominent) (Horton et al., 2014) to be able to aid the understanding of their consequences at a mental, emotional and psychological level.Hence, our findings highlight the importance of unpacking individuals' professional and religious identities and accounting for their more inner and constitutive elements, namely values, beliefs and principles, as elements that can have a significant impact on individuals' psychological well-being, as well as potential triggers of identity conflict.By focusing on this specific combination of work and non-work identity values, we ultimately address a rather neglected but increasingly important topic in organizational settings, namely the influence of religious values on professional practice (Carminati & Héliot, 2021;Héliot et al., 2020;Tracey, 2012).Secondly, we found partial support for the cross-level, moderating effects of organizational ethical climate and peer social support as contextual factors ameliorating the relationship between identity conflict and psychological well-being.Hence, we were able to account for some potential contextual factors that could influence individuallevel identity dynamics and show the interconnection between the individual self and the broader social context (Carminati & Héliot, 2021;Islam, 2020).Indeed, although identity conflict can be viewed mostly as an intrapersonal process, people's experience of identity conflict can be shaped by the surrounding environment and by human interactions (Carminati & Héliot, 2022;Islam, 2020;Treviño et al., 2008).By integrating Social Information Processing Theory and identity theory approaches, we accounted for the influence of social-contextual factors on individual processes (Alvesson et al., 2008).
Our findings demonstrated the moderating effect of two particular extra-individual forces, namely organizational ethical climate and peer social support.The influence of these extra-personal forces on identity conflict underlines how the juxtaposition of individual and organizational levels can comprehensively and better explain the complexity of identity conflict, accounting for the social-contextual embeddedness of individual processes (Islam, 2020).By connecting these individual and organizational processes, we were thus able to offer a more complete picture of the multilevel mechanisms surrounding identity conflict dynamics.
More specifically, previous studies found a positive moderating effect of ethical climate level and strength toward individual outcomes (Schminke et al., 2005;González-Romá et al., 2002;Shin, 2012).However, research has also noted that these moderating results are often inconsistent (Schneider et al., 2013).The negative interaction we obtained could thus be explained by the fact that in conditions of identity conflict involving sensitive personal values of moral nature, a consistent and strong organizational ethical climate might have been perceived by doctors as something oppressive and suffocating their own moral code of conduct and freedom of conscience (Genuis & Lipp, 2013;Giubilini, 2016).Indeed, in line with the lens of inter-unit conflicts, organizational discrepancies can foster distinct expectations and ambitions, which can provide the basis for divergent identity claims and behaviors (Horton et al., 2014).Consequently, organizational ethical climate exacerbated the negative influence of identity conflict on psychological wellbeing, making the experience of the conflict even more difficult to tackle.
On the contrary, our results showed that organizational peer social support was perceived as a genuine aid in situation of moral impasse (Wallace & Lemaire, 2007).Whilst previous studies tended to underline the medical profession as a "solo-person" job (Deshpande & Joseph, 2009;Hurst, 2005;Woo et al., 2006), we illustrated the key importance of peer support in lessening the negative psychological consequences of identity conflict.This is in line with previous research reporting how social support is crucial in circumstances characterized by lack of cues (Wiesenfeld et al., 2001), in helping individuals to make peace with difficult emotions (Woo et al., 2006) and tackle ethical decision making (DuVal et al., 2004;Hurst et al., 2007).However, we also showed that these positive moderating effects between identity conflict and psychological well-being were possible only when level and strength of organizational peer social support were considered in conjunction.This seems to imply that in order to significantly ameliorate the negative impact of identity conflict on psychological wellbeing, the opinions of colleagues not only must be deemed valuable and trustworthy, but also show a certain degree of alignment and consistency.Having either one or the other would not be sufficient to solve ethical dilemmas.Only when both conditions are met, organizational peer social support can become one of the most effective factors impacting people's cognitive, emotional and behavioral outcomes (Holman et al., 2018).

Practical and research implications
By examining doctor's identity conflict in ethically-charged circumstances, our findings illustrate that doctors' professional-religious identity conflict can lead to lower levels of their psychological well-being.Therefore, firstly, healthcare organizations and hospitals are called to raise employees' awareness of ethical impasses due to religious or other strong moral values.Although in today's secularized society religion is sought in more privatized forms, religious identity matters for individuals within the work contexts and especially in ethically-charged circumstances (Carminati & Héliot, 2021;Tracey, 2012).Indeed, it can significantly affect doctor's decision making, job performance and wellbeing (Héliot et al., 2020).Hence, to safeguard doctors' well-being related to professional and religious conflicts, human resource departments should support diversity management to foster respect, tolerance and a climate of psychological safety (Héliot et al., 2020).In this way it would be possible to prevent or tackle the emergence of identity conflict, as well as reduce stress and improve individuals' psychological well-being (Hurst, 2005;Pelechova et al., 2012), thus aiding the management of occupational health.
Secondly, our findings have practical implications on cooperation between doctors within healthcare organizations.The joined positive effect of peer social support level and strength on the relationship between identity conflict and psychological well-being implies that hospitals should foster greater levels of cohesion across the organization.Too often such an organizational collaboration is unstructured, scattered or informal (Héliot et al., 2020).Since the medical profession is increasingly relying on multidisciplinary and interdependent group practices (Hewett et al., 2009), we call for organizations to endorse a strong and formal cooperation between healthcare professionals (e.g., through cross-department team-building activities, communication or celebration of organizational success), since this could improve the overall effectiveness of the service provided (Halbesleben & Rathert, 2008).Hence, promoting peer cohesiveness across the hospital could enhance a climate of support which doctors could trust and strongly rely on when facing identity conflict (Kälvemark et al., 2004;Karnik & Kanekar, 2016;LeBlanc et al., 2007).
Lastly, our findings provide important insights for research.By merging solid theoretical approaches from Management-Organizational Studies research with medical insights on moral dilemmas, well-being, and contextual extra-individual forces, we were able to address, and contribute to, the complexity of identity conflict dynamics within organizations.In particular, by juxtaposing and integrating individual and organizational factors, we investigated how, and the extent to which, intrapersonal-level processes of professional-religious identity conflict could interact with socio-ethical, contextual forces to further our understanding of identity conflict dynamics.In this way, we extended current knowledge of how individual processes are embedded in, and influenced by, socio-contextual factors.More specifically, at the individual level, we addressed concerns regarding the role played by identity values in triggering the conflict and their consequences on psychological outcomes (Horton et al., 2014).At the organizational level, answering important calls regarding the implementation of multilevel analyses in the identity realm, we have instead included extra-individual forces in the relationship between identity conflict and psychological well-being to account for contextual factors that could lessen such negative relationship (Alvesson et al., 2008;Héliot et al., 2020;Islam, 2020).

Limitations and future research
As for all research, our research is not without limitations.Firstly, although to the best of our knowledge our work represents one of the first studies to implement multilevel analysis with regards to identity dynamics to account for contextual, extra-individual forces, our sample size at both individual and hospital level was relatively small.Although at the individual level we had enough cases (Faul et al., 2009) and at the organizational level enough hospitals (Park & Yu, 2018;Snijders & Bosker, 2012), and we appreciate the difficulty of conducting research with busy healthcare professionals, we encourage future work to conduct similar research with a bigger sample size to avoid potential limitations in terms of statistical power (Sim, Kim & Suh, 2022).Furthermore, one of our key variables, namely "salience and centrality of religious identity values", was slightly negatively skewed.Even though we implemented bootstrap procedures (Russell & Dean, 2000), this might have impacted the interaction effect with "salience and centrality of professional identity values", and might thus explain why only professional identity was able to predict identity conflict.Hence, if aligned with their study objectives, future research encountering similar problems may consider other data transformations to improve the quality of their data (Russell & Dean, 2000).
Secondly, given that all our measurements relied upon self-reported questionnaires, the internal validity of our findings may have limits.We counteracted these issues by assessing our variables at two points of measurements (Podsakoff et al., 2012) and through a multilevel analysis, which give us more confidence in our results.However, to further strengthen the validity of the findings, we suggest a full longitudinal design in which all constructs are measured at multiple time points (Kline, 2011), as well as the implementation of alternative rating sources for identity conflict (e.g., colleagues, spouses and patients) (Podsakoff et al., 2012).
Thirdly, we acknowledge that our model captures only part of the variables that are likely to influence identity conflict.Supported by the medical literature, we identified religious identity values as carrying potential incongruent principles for medics and ethical climate and peer social support as buffering conditions of the relationship between identity conflict and psychological well-being.We recommend that future research should incorporate other individual-level factors, such as moral or family identity values and emotions, to extend the range of potential explanations of identity conflict and ethical dilemmas (Genuis & Lipp, 2013).Akin to this, we also encourage the integration of other contextual forces that could influence such relationship, namely patient and family perspectives and institutional pressures.

Conclusion
Our research found that identity conflict can negatively influence doctors' psychological well-being depending on the salience and centrality of the conflicting religious and professional identity values at stake.We also found that organizational, sociocontextual factors can impact the way individuals tackle the discomfort of identity conflict.Importantly, our findings shown that, whilst hospital ethical climate exacerbates the negative influence of the identity conflict on psychological well-being, organizational peer social support significantly ameliorates such relationship.We believe that these results are critical signals that even traditionally recognized intrapersonal processes can be profoundly influenced by interpersonal, extra-individual dynamics, and that "solo-person" professions (e.g., the medical profession, the legal profession, etc.) can benefit enormously from opening up to support from colleagues from the same organization.

Notes
1.The concept of "identity salience" represents the likelihood that a particular identity is cognitively accessible and enacted in accord with the expectations attached to that identity (Ashmore et al., 2004).It can be thought at as the relative importance of an identity.On the contrary, "identity centrality" represents the importance attributed to each identity by an individual and assumes level of self-awareness (Ashmore et al., 2004).
It can be viewed as the absolute importance of an identity.In this article, following Stryker and Serpe (1994, p. 34) suggestions that "in the absence of a reasonable understanding of the conditions under which identity salience and centrality are independent of one another, we recommend that both concepts be included in research designed," we use both the two concepts.2. For instance, religions such as Unitarianism would not cause an identity conflict since it is not against controversial and legally bounded actions (e.g., physician assisted suicide).On the contrary, Catholic physicians may refuse to perform abortion, physician-assisted suicide, etc., since these actions are against their religious values (Carminati & Héliot, 2021).3. The specific indirect effect from Doctors' Professional Identity Values toward Psychological Well-being was β= -0.086, p < 0.068.

Figure 1 .
Figure 1.Hypothesized multilevel model with both hospital ethical climate and hospital peer social support level and strength as organizational, level-2 moderators.

Figure 4 .
Figure 4. Interaction plot of the salience and centrality of doctors' professional and religious identity values toward psychological well-being (PWB).

Figure 6 .
Figure 6.Plot of the slope analysis of the 3-way cross-level interaction of hospital ethical climate level, strength and identity conflict.PWB = psychological well-being.

Figure 7 .
Figure7.Plot of the slope analysis of the 3-way cross-level interaction of hospital peer social support level, strength and identity conflict.PWB = psychological well-being.

Table 1 .
Means, standard deviations and correlations of the study variables.

Table 2 .
Comparison of alternative models for hospital ethical climate.

Table 3 .
Comparison of alternative models for hospital peer social support.
10 PWB = Psychological Well-Being.4-factor (PRO -REL) depicts a model in which both professional and religious identities are grouped together under the same latent factor.4-factor (CONF -PWB) comprises a model in which identity conflict and PWB are clustered under the same latent factor.1-factor is a model in which all the variables are modeled under the same latent factor.

Table 5 .
Unstandardized effects for moderating effects of hospital ethical climate.

Table 6 .
Unstandardized effects for moderating effects of hospital peer social support.