Confirmation of Military Identity after Exposure to Stress – Influence of Status Dimensions

The first aim of this study of Norwegian veterans was to examine how deployment to interna tional operations relates to a strengthening or weakening of military identities, and whether status dimensions are associated with a change in operational identity. The second aim of this study was to examine whether cohesion and employment status are associated with a change in operational identity. One hundred and ninety-nine Norwegian soldiers serving in Afghanistan completed an identity questionnaire before and after deployment, including various measure ments of stressor load, cohesion, personal characteristics and mission characteristics. Mixed model analysis revealed that, as predicted, a change in the dimension operational identity, but not individualism, was associated with the status dimension threat . The analysis revealed no association between operational identity and cohesion, but found differences between tem porarily and permanently employed personnel regarding level of operational identity. Possible implications of these findings are discussed, and further research is recommended.

During the last decades, many Western Armed Forces shifted their operational focus from defending their own territory to participating in international operations. This consecutive restructuring of the military organisation increased the number of professional soldiers within these Armed Forces (Eriksson, 2004;2007). In Norway, for example, thousands of soldiers have participated in international operations over the last decades. Being deployed to such operations probably represents a hallmark of their soldiering, and one could assume that such an experience further strengthens their personal skills, tactical skills and overall military identity. The military operational identity encompasses professionalism and warriorism (Johansen, Laberg and Martinussen, 2013a;2013b). Even though operational identity is regarded as valuable for a military organisation, it might also give cause for concern. One matter of concern is that strengthening an operational identity means weakening idealistic motives for serving (Eriksson, 2004). This might imply that soldiers become less motivated for peacetime service aimed at territorial defence. Additionally, a professionalisation of the Armed Forces might recruit self-centred individuals and thus promote individualistic values (Eriksson, 2004(Eriksson, , 2007Johansen et al., 2013a). To better understand important motivational characteristics of the soldiers, the current study examines whether a deployment to international operations strengthens or weakens military identities. In addition, the current study investigates possible predictors of operational identity such as status dimensions and cohesion and the possible effects of employment status.
In order to describe and differentiate between different military identities, researchers typically draw on concepts and theories by Huntington (1957), Janowitz (1960) and Moskos (1977). For instance, to investigate the development of US cadets, Franke (1997) distinguished between patriotism (attitudes toward serving and fighting for own nation), warriorism (attitudes toward the warfighting and peacekeeping role of the Armed Forces, own expectations and personal satisfaction) and globalism (attitudes toward global institutions). Johansen et al. (2013a) conceptualised Norwegian military identities through the constructs of operational identity, idealism and individualism. Operational identity is based on both professionalism and warriorism and characterised by ethos, including a willingness to deploy to international operations, sustaining high standards in conduct of operations, being skilled in warfare and being motivated to serve based on cohesion SCANDINAVIAN JOURNAL OF MILITARY STUDIES and comradeship rather than a superior cause (Johansen et al., 2013a). Deployments to international operations often include conducting live military operations and a chance to prove and strengthen your competence and skills (Godal et al., 2016). However, there are probably numerous soldiers of the Norwegian Armed Forces today who do not fully embrace the ethos and values of professionalism (Eriksson, 2007). This might especially relate to some of the soldiers who volunteer to join ad hoc units that make up heterogonous groups of soldiers with respect to military competence and expertise (Lien, Firing, Bendixen, & Kennair, 2016).
Idealism is characterised by known institutional military values such as collectivism, patriotism and altruistic values (Ulriksen, 2002). Idealism relates to whether soldiers believe in the importance of defending their own nations' territory and interests, but are reluctant to participate in international operations (Johansen et al., 2013a). Some suggest that professionalism makes idealism less relevant (Eriksson, 2004). Other studies conclude differently, such as that of Ben-Dor et al. (2008) which suggests that patriotism is a predictor of motivation to serve among Israeli reservists, and Griffith's research which shows the same results among US reservists (2009a). Even if a deployment to international operations will confirm and increase operational identity, it is not given that existing idealistic values of the soldiers will decrease. There are no obvious, contrasting values between professionalism and idealism, and a deployment might be considered an opportunity to help other people in need (Eriksson, 2004). In addition, some deployments, such as the engagement in Afghanistan, might be perceived as US and NATO support (Godal et al., 2016) which, again, can be perceived as valuable for the protection of national interests and territory. Furthermore, it is possible that many soldiers have a pragmatic view of a deployment to international operations and also consider increased military competence acquired from a deployment as beneficial to their regular service back home.
Individualism is characterised by egocentrism and selfishness and mainly concerns whether soldiers believe salary, working conditions and individual rights are important prerequisites to work in the Armed Forces. Individualism also involves a search for self-fulfilment, possibilities and challenges (Johansen et al., 2013a). Huntington (1957) called for a professional soldier who was selfless, responsible and served for the good of society. Moskos (1977;1986) has argued that a change in society has caused a tendency for military service to be viewed more as a regular occupation than a calling, and that the transition to an all-volunteer force enhances this tendency. Individualism is described as a threat to operational identity (Johansen et al., 2013b). Johansen et al. (2013a) found that a need for self-fulfilment, opportunities and challenges were associated with operational identity, but argued that it might be needed qualities in the complex and unpredictable military operations of today. However, most aspects of individualism seem to concern issues of payment and individual rights and that soldiers view their work as an ordinary job (Johansen et al., 2013a). Such values might influence the decision to volunteer for such service, but it is not given that these values will change significantly as a consequence of being deployed to international operations.
In general, soldiers are socialised into group norms, values, attitudes and skills when they enter the military organisation. Part of this socialisation process involves group bonding, teamwork and peer support (Winslow, 2004). This describes the essence of team cohesion which is highly valued in operational identity (Johansen et al., 2013a). However, cohesion is a contested term, and the division of task cohesion (shared unit goals that demand collective efforts) and social cohesion (relational and emotional bonds between soldiers) is discussed, particularly with regard to the effects they have on military performance (King, 2013;Kirke, 2009;Wong, Kolditz, Millen & Potter, 2003). One could claim that task cohesion rather than social cohesion is connected to professional military competence, for instance in the cooperation between soldiers in combat. However, there are also researchers who view social cohesion as a key motivation in combat (King, 2013). So far, research by Johansen et al. (2013a, b) and Franke (1997;2001) has not included cohesion as a predictor of military identity. Thus, an investigation into several aspects of cohesion, such as task cohesion and social cohesion, is called for in relation to military identity. This might also include horizontal cohesion (peer-peer) and vertical cohesion (leader-unit) (Kirke, 2009).
Social identity is described as the part of the self-concept that derives from knowledge of group memberships (Tajfel, 1981). A positive social identity is motivated by securing or by reducing uncertainty about the self-concept, and this is a process that is based on social comparison (Hogg, 2000). Several levels of comparisons are possible as soldiers identify themselves as one individual of several in a group, as a member of a unit compared to other units or with the organisation as a whole (e.g., soldier, officer), dependent on what is salient in the particular context ( van Dick, Wagner, Stellmacher, Christ & Tissington, 2005). The comparison is based on the characteristics of the group perceived as social status dimensions (Ellemers, Spears, & Doosje 2002). However, military identity is not always described as a fixed structure of personal or group characteristics. Woodward & Jenkings (2011) consider personal military identity as a construction that is based on practice and experience (what we do) rather than on certain characteristics (what we are). This corresponds with professionalism and warriorism which include a desire to act and conduct military operations based on high standards, as shown in descriptions of the British military in Afghanistan (King, 2010). It also corresponds to operational identity as described by Johansen et al. (2013a). Furthermore, a study by Lien, Firing, Bendixen and Kennair (2016) showed how veterans believed that having experienced action such as combat, taking risks in general and participation in numerous operations were ranked high on the status hierarchy. What is more, it showed that the veterans compared themselves to other individuals and units and upgraded and downgraded their own efforts accordingly, probably affecting both their personal and social identity (Lien et al., 2016). This is described as a meaning-making process (Lien et al., 2016;Park, 2010). Thus, experiencing risk and threat is not only consistent with existing meaning structures of being a soldier, but exposure to such stress probably represents experiences that help to manifest their identity as professional soldiers.
There might be several reasons why the strength of military identity may vary among groups of military personnel. Units that are selected, trained, equipped and dedicated to deployment to international operations are expected to identify with the military profession more than regular military units or ad hoc units. For instance, the Norwegian Army Rapid Reaction Forces score higher on professional identity than war college cadets or a sample of soldiers from mixed branches and services (Johansen et al., 2013a;2013b;Johansen, Martinussen, & Kvilvang, 2016). What is more, a person might have several social identities, some of which may influence their identity as a professional soldier. In Norway, it is common to deploy personnel who are not permanently employed in the Armed Forces to international operations. It is usually former military personnel engaged from pre-deployment preparations and training until post-deployment. They might have civilian jobs or be students. Alternatively, it might be personnel who have finished conscription and are temporarily engaged, pending a permanent contract with the Armed Forces. Of all Norwegian personnel being deployed to Afghanistan from 2001 to 2012, over fifty per cent were temporarily employed personnel (Forsvarets Sanitet, 2013). In Norway, seemingly, only a few studies investigate the effects of the use of such personnel in a military context, for instance whether they identify less with the operational identity than permanently employed soldiers do. This category of personnel might distance themselves from the military profession due to competing social roles, for instance by having a civilian profession, as found among reservists (Griffith, 2009b). Moreover, if temporarily employed soldiers have lower military skills than their peers, they might be less respected and integrated into their unit (Lien et al., 2016). For instance, temporarily employed firefighters who are alienated, are in danger of becoming less committed to their job, less integrated in their work group and identifying less with their profession (Halbesleben and Clark, 2010).

Aims and Predictions
The first aim of this study is to examine whether deployment to international operations strengthens or weakens military identities and whether status dimensions, operationalised as the stressor variables threat and action, are associated with change in military identity. The second aim of this study is to examine whether cohesion is associated with military identity, and whether there are any differences between groups of personnel on the basis of employment status. The participating soldiers in the current study have showed their willingness to deploy to international operations. They have also performed live military operations and, as such, probably proven their ability to conduct operations with high standards as they have been trained to do. Thus, we expect operational identity to be strengthened as a result of deployment to international operations. We also expect that individualism will not change after service in international operations as it mostly concerns payment and individual rights that are not expected to be influenced by experiences from a deployment. In addition, as most deployments abroad are not necessarily in conflict with idealistic values, we expect that idealism does not change after service in international operations. We also expect that the cohesion among most soldiers increases within their units as a consequence of external threat as well as having shared experiences (Bartone, Johnsen, Eid, Brun, & Laberg, 2002). Here, status dimensions refer to experiences during the deployment that reflect Action, such as having experienced contact with enemy forces, exposure to risk and participation in numerous missions as described in Lien et al. (2016). Thus, we expect cohesion and the experience of status dimensions to be associated with change in operational identity. This leaves us with the following predictions: 1. Operational identity strengthens, while individualism and idealism do not change following deployment to international operations. 2. Status dimensions and cohesion are predictors of a strengthened operational identity.
The current study also examines whether employment status is associated with operational identity. Studies of reservists (Griffith, 2009b) imply that membership in other important social groups might distance sol-diers from the military profession. In addition, civil, temporarily employed workers have experienced alienation (Halbesleben and Clark, 2010), which also speaks against increased cohesion during a deployment.

Methodology Design and Participants
The overall prospective study was designed and administered by the Norwegian Armed Forces Joint Medical Services, Institute of Military Psychiatry. The data used in the study are part of a larger study where the aim was to investigate stress, coping, health and growth of veterans. The entire sample included 549 soldiers, most of whom were stationed in Maymaneh in Afghanistan and were part of the Provincial Reconstruction Team (PRT). The PRT secured the Norwegian camp in Maymaneh and conducted security operations in the adjacent area. This unit changed their initial objectives during the deployment and was ordered to hand over the camp to the Afghans and relocate to Mazar-e-Sharif after four months of service out of the total of six months. 1 Rather than conducting security operations in the area for the entire six months, after four months of service they thus abandoned the camp and redeployed to Mazar-e-Sharif where they prepared for the upcoming departure back home. Minor units were stationed in Mazar-e-Sharif for their entire six-month deployment, primarily as a part of the National Contingent Commander (NCC) and National Support Element (NSE). 2 The data were collected at seven points of time (T1-T7). 3 The data subset consisted of repeated measures of military identity (T1 and T6), objective stressors (T4) and cohesion (T1, T2, T3 and T4). Also, demographic variables (sex, number of children, marital status) and mission variables (rank, previous number of months served, number of previous deployments) were made available for the current research. The sample consisted of civilians, enlisted personnel as well as non-commissioned and commissioned officers in the Norwegian Armed Forces.

Ethics
All participants were given detailed written information about the study as well as the means taken to ensure the participants' confidentiality. They were informed that the data would be stored in the Norwegian Armed Forces Health Registry, and that the results, in anonymous form, might be used for research purposes later. Participation was voluntary, and all participants provided written informed consent before taking part in the study. Study procedures, collection, storing and distribution of data were done in accordance with the existing legislation regulating the Norwegian Armed Forces Health Registry.

Military Identity
Military identity was measured by means of the Military Professional Identity Scale (NPIS, Johansen et al., 2013a). The 33-item questionnaire covers three subscales: Operational identity (12 items), idealism (11 items) and individualism (10 items). As regards operational identity, questions concerning the importance of military skills and war experience were asked. For individualism, questions concerning the importance of salary, working conditions, individual rights, opportunities and challenges were asked and, for idealism, questions concerning the importance of protection of Norwegian territory, values and interests. The participants responded to a 7-point Likert scale with anchors 1 (totally disagree) and 7 (totally agree). Since this measure has previously showed marginal reliability scores on two of the dimensions (idealism and individualism; Johansen et al., 2013a), we conducted a factor analysis to confirm the factor structure. In contrast to the original factoring (Johansen et al., 2013a), our exploratory factor analyses (extraction: Maximum likelihood, rotation: Oblimin) at T1 and T6 suggest two consistent subscales: Operational identity (10 items) and individualism (7 items). The original subscale idealism was not a consistent factor at T1 and at T6 and was therefore omitted from further analysis. The internal reliability (Cronbach's α) for professionalism was 0.83 at T1 and 0.85 at T4 and, for individualism, 0.72 at T1 and 0.70 at T6.

Status Dimensions
Measurement of status dimensions were based on 14-item measure related to combat exposure and a 7-item measure related to moral provocations (Hougnaes, Bøe, Dahl and Reichelt, 2017). A 4-point rating scale was applied with responses 0 (No), 1 (1-3 times), 2 (4-7 times) and 3 (more than 7 times). Using Lien et al. (2016) as a basis, status dimensions in the current research were defined as to the degree to which the soldiers had been experiencing the following different categories during their deployment: Combat, Improvised Explosive Device (IED) attack, risky operations and number of missions. However, since few of the veterans had experienced any of these categories, the categories were limited to action (combat and IED attacks) and threat (risky operations and number of missions), and the dichotomising of the variables was assessed. Action referred to eight items concerning being attacked, ambushed or firing towards the enemy, taking lives or being in a military column struck by an IED. Threat referred to three items concerning participating in operations outside camp, handling of IEDs and crossing of unsafe mine/IED areas. The other items of the measure were not used.

Cohesion
Cohesion was measured using a Norwegian adaptation of the Platoon Cohesion Index (PCI, Siebold & Kelly, 1988;Siebold, 1999). The measures consisted of a 29-item (T1), a 12-item (T2) and a 16-item (T3 and T4) version. A 5-point Likert scale was applied with anchors 1 (not at all true) and 5 (completely true). A similar Norwegian adaptation of the PCI has revealed a narrower factor structure than the original version (Bartone et al., 2002). We therefore conducted a principal component analysis (PCA) (extraction: Maximum likelihood, rotation: Oblimin). This revealed a 2-component structure of cohesion for the 16-item version, namely vertical cohesion (leader to soldier) and horizontal cohesion (soldier to soldier). The PCA analysis for the 12-item solution only revealed a 1-component structure. To be able to compare development of cohesion from T1 to T4, the same eight items were selected for horizontal cohesion and the same four items for the vertical cohesion for T3 and T4. 4 The internal reliabilities for cohesion were excellent (vertical cohesion at T1, 0.91, vertical cohesion at T2, 0.93 and vertical -and horizontal at T3 and T4, 0.94).

Statistical Analysis
The data set (n = 199) had the following missing values: T6 -the dependent variables (below 1%), T1employment (2.5 %), T1 -horizontal cohesion (4 %), T2 -horizontal cohesion (18 %), T3 -horizontal cohesion (17 %), T4 -horizontal cohesion (16 %), T3 -vertical cohesion (17 %), T4 -vertical cohesion (16 %), T4 -action (14 %) and T4 -Threat (14 %). Mixed model analysis does not demand the use of traditional imputation methods, but rather analysis of maximum likelihood. Since the data covered repeated measures of the same individuals at T1 and T4, had several missing values and continuous dependent variables, mixed regression models were chosen as analysis method (Brady, 2009). The analyses were conducted by means of two different subscales of military identity (change in Operational identity and Individualism) as the dependent variables. The last subscale of military identity, idealism, was not a consistent factor at T1 and T6. All analyses were performed using IBM SPSS Statistics version 24. This includes a missing value analysis.
Higher levels of operational identity were reported after the deployment (M T6 = 4.75) than before the deployment (M T1 = 4.35), with about the same levels of individualism (M T6 = 4.88) and (M T1 = 4.75). Of the variables defined as status dimension (action and threat), action was dichotomised based on a low exposure. In the sample of veterans (n = 199), the variable action consisted of 136 who had not experienced action, 35 who had experienced action, while 28 had missing values. The normality of distributions for the rest of the variables was checked, and transformations were deemed unnecessary (Dunlap, Chen, & Greer, 1994). A correlation matrix of the different variables is presented in Table 1.
The zero-order correlations show that operational identity had a low, but significant correlation with individualism. This is expected, given the relations between these two dimensions (Johansen et al., 2013a). Horizontal cohesion has significant and high correlations between the four time points (r = 0.45-0.68) and Table 1: Zero-order associations (Pearson's r) for continuous predictors and outcome variables (Listwise deletion, n = 199). vertical cohesion between the two time points (r = 0.72). We also note that the pre-and post-measures of horizontal cohesion are lower at mid-deployment and on homecoming (T3 and T4) than before deployment (T1 and T2). The correlations matrix also shows that only the status dimension threat and the horizontal cohesion T2 and T4 are significantly correlated with changes in operational identity. Temporarily employed soldiers report significantly lower levels of professional identity before deployment (T1) than permanently employed soldiers. However, there are no significant relationships between employment and change in operational identity. In addition, temporarily employed soldiers show higher levels of horizontal cohesion than permanently employed soldiers before deployment (T1), decreasing upon deployment (T2) and mid-deployment (T3). On homecoming (T4), there are no significant differences between horizontal cohesion and employment.

Mixed Model Analysis
Use of a two-level hierarchical mixed model regression analysis enabled an assessment of the effects of change in professional identity by repeated measurement time (level 1) and subject level (level 2). Level 1 was defined as the repeated measurement time at T1 and T6 of military identity, giving us the possibility to examine change in identity over time. Only professional identity and individualism were examined. Level 2 was defined as individuals, giving us the possibility to investigate individual variables (status dimensions and cohesion). A baseline model was established by using the unstructured covariance matrix for the repeated measures and maximum likelihood (ML) estimation. Predictors of the dependent variable were entered in three different blocks. First, the variable time was entered as a fixed factor. Second, T2 and T4 horizontal cohesion were entered and, third, threat was entered as fixed factor. Based on the principle of parsimony, no significant variables for each block were discarded. The baseline model had a -2LL = 912, df = 7. Besides time, the following fixed effects were significant: Threat F (1, 164) = 28.16, p < .001. To improve the model further, inspection of the dependent variable change in professional identity showed a fairly constant variance for the two time points. Interclass correlations of the repeated measurements were also recorded (r = .43). Thus, we chose the covariance matrix compound symmetry as it prescribes constant variance at each time point and constant correlation between measurement times at level 1. This did not improve the model, -2LL = 914, df = 6, and nor did other variance matrixes. Then, a random intercept was added to the model, including the significant covariate predictor threat as random effects, but this did not improve the model either. The results of the baseline model are as described in Table 2 (n = 199). Table 2 shows that there is a significant relationship between T1 and T6 in operational identity. Subjects who are exposed to one more standard unit of threat, have a predicted increase in operational identity of 0.58, 95% CI [.38, .78]. The variables horizontal cohesion (T2) and horizontal cohesion (T4) were not found to be significant or improve the baseline model. Furthermore, a mixed model analysis assessed the effects of change in individualism by repeated measurement (time). Change in individualism was not significant (n = 199), F (1, 198.00) = 3.02, p > .05).

Discussion
The results showed that operational identity increased after service in international operations, and that there was no change in individualism. This partly confirms our first prediction. A deployment to international operations includes carrying out live military operations and a chance to prove and strengthen your competence and skills (Godal et al., 2016). By conducting military operations abroad, soldiers perform as a group in a stressful environment and develop their military skills, and this is described as meaningful experiences (Schok, Kleber & Boeije, 2010). Thus, such experiences comply with the primary characteristics of professionalism and might explain the strengthening of operational identity in the Model fit -2LL = 939, df = 6. current research. For the military organisation, a strengthened operational identity is an advantage. Soldiers with a strong operational identity represent military personnel skilled in warfare (Johansen et al., 2013a;2013b). This is favourable in case of crisis and war and is valuable in the training of new personnel when experienced personnel return to regular service. Furthermore, the willingness to participate in international operations makes it easy to recruit qualified personnel for the next mission. Participating in international operations is probably the single most important factor to professionalise the Armed Forces further (Eriksson, 2007). The fact that individualism does not increase during a deployment is also favourable for the military organisation. The results in the current research indicate that the experiences during a deployment do not promote individualistic values. Seemingly, payment, individual rights and working conditions are not factors that get a lot of attention in this context. To some, individualism is regarded as a threat to operational identity (Johansen et al., 2013b). However, soldiers who deploy to international operations often find meaning in experiences that promote individualistic goals and values. For instance, some personnel are motivated to deploy in order to test themselves in stressful situations (Lien et al., 2016). Such individualistic values are just one of many other aspects of service that soldiers find meaningful (Lien et al., 2016;Schok et al., 2010), and personnel who search for self-fulfilment, possibilities and challenges probably do not pose a huge threat to operational identity. Indeed, the Norwegian Armed Forces recruit personnel to join the Armed Forces and to deploy to international operations by appealing to such values (Eriksson, 2007).
In the current research, changes in idealism were not measured as the factor analysis of the NPIS scale did not show a consistent factor structure over the two time points T1 and T6. This might be explained by the fact that idealism has a complicated structure and is difficult to operationalise (Johansen et al., 2013a). It might also reflect a heterogeneous sample. The sample used in the current research represents different age groups and branches within the Armed Forces. Furthermore, the sample consists of several units with different objectives that have been exposed to stress in varying degree. Those soldiers who have seen much suffering might become less compassionate about the people they are to help, decreasing their affective motivation to serve, even if it is temporarily (Braender & Andersen, 2013). Thus, those who become less compassionate might score unequally on emotionally versus cognitively based values of idealism, while others might score equally on emotionally and cognitively based values of idealism.
We also predicted that status dimensions are predictors of change in operational identity, and this was partly confirmed. The variable threat (risky operations and number of missions) was related to strengthening of operational identity. Some experiences might secure one's social position in a group (Ellemers et al., 2002), and we believe threatening experiences confirm one's identity as a professional soldier. This adheres to descriptions of professional soldiers who presume first and foremost that they are what they do (King, 2010). It also adheres to description of social identity as a process of comparison on social status dimensions (Ellemers et al., 2002). Similarly, Britt, Adler, & Bartone (2001) defined meaning as both war-related experiences and adoption of a peacekeeper identity for US soldiers performing peace support operations on the Balkans. Consequently, military leaders should avoid that their personnel are stuck in a camp during their entire deployment. Rotation systems might ensure that most soldiers get involved in live military operations during their deployment in order to find positive meaning in their experiences, thus confirming their operational identity. However, it might raise several ethical questions if conducting military operations and experiencing threat becomes a goal in itself. In the process of training to conduct live military operations lies an expectation of potential threatening situations during a deployment. Thus, it might create an inherent pressure on military leaders to conduct military operations independently of the overall objectives of the mission.
The variable action (combat and IED attacks) was not a significant predictor of change in operational identity even though one might believe that combat is perceived as a strong status dimension for most soldiers (Lien et al., 2016). Combat involves a level of controllability (Linley & Joseph, 2004). In contrast, IED-attacks might be perceived as uncontrollable, enhancing feelings of helplessness. Thus, it might be perceived as a weak status dimension for the soldiers even though it could be regarded as action as described by Lien et al. (2016). Another aspect is that experiences such as combat might be traumatic events (Larner & Blow, 2011), and the meaning of such experiences might change if focus shifts from status to the risk involved (Lien et al., 2016). Thus, this might explain why the variable action used in the current research is not a predictor of change in operational identity as otherwise assumed. It also emphasises that the motivation of the soldiers might change, dependent on their experiences during a deployment.
We predicted that cohesion is a predictor of change in operational identity, but this was not confirmed. This prediction was mainly based on descriptions of operational identity (Johansen et al., 2013a), but some researchers question whether the ethos of military units really reflects a development of cohesion, particular in so called temporary or instant units (Ben-Shalom, Lehrer, & Ben-Ari, 2005). The current research did not distinguish between task cohesion and social cohesion. The PCA analysis showed that items from both task and social cohesion loaded on the same factor, indicating they were measuring the same construct. We made a division of horizontal and vertical cohesion, but neither of these variables increased from pre-to post-deployment. The fact that professionalism increases from pre-deployment to post-deployment leads one to expect that cohesion also will increase during live military operations. Normally, common experiences, familiarity with other team members and external threats strengthen cohesion within military units (Bartone et al., 2002). However, the results showed no increase in cohesion and this might be due to a reduction in number of operations in the last part of the deployment. Most of the personnel belonged to the PRT 19 in Maymaneh, a unit that handed the camp over to the Afghans and redeployed to Mazar-e-Sharif two months prior to their departure back home. They considerably reduced the number of conducted security operations the last two months and thus also the common experiences that are related to cohesion, such as backing and caring for other team members (Lien et al., 2016). Moreover, if cohesion is an expression of commitment to both the group and its objectives (Mudrack, 1989), changing the objectives in the last part of the deployment might further have reduced the development of cohesion.
We also investigated whether there were any differences between permanently and temporarily employed soldiers with respect to change in operational identity. In the current study, temporarily employed personnel on average did not show lower levels of operational identity from pre-deployment (T1) to post-deployment (T6). However, temporarily employed soldiers score significantly lower at pre-deployment on professional identity. According to Griffith (2009b), reservists might distance themselves from the military profession due to competing social roles. However, there were no differences at homecoming between temporarily and permanently employed soldiers with respect to operational identity. This indicates that temporarily employed soldiers were well integrated into their units and did not experience being alienated from the profession during the deployment.
The fact that temporarily employed soldiers report higher levels of horizontal cohesion at pre-deployment and during the deployment might be due to their previous references of cohesion. Temporarily employed soldiers might be less experienced members of professional military units and might report high levels of cohesion if they have previously not felt a strong sense of belonging to a unit. The fact that they report the same levels of horizontal cohesion at homecoming might be seen as a result of an adaption to the context, where their job is to fit into a new group no matter what, since the group is all, they have (Winslow, 2004). It takes almost six months until they report the same levels of cohesion. Thus, one could argue in favour of the importance of training new teams and units before deployment in order to establish primary groups that enable needed bonding and relevant interaction between soldiers (King, 2013). However, ad hoc units might establish cohesiveness quickly when action and shared missions foster member interaction, cooperation and swift trust among group members (Ben-Shalom et al., 2005). Thus, one might believe that the soldiers of the current research developed cohesion within their units more slowly since they experienced a limited amount of action during the deployment.

Limitations and Further Research
The current research has used a prospective design, but can only document an association between status dimensions and an increase in operational identity. Further research might make it possible to develop a questionnaire designed to measure status dimensions for military personnel serving in international operations and not only measure stress exposure as in the current research. Such a measurement might include possible other meaning categories that are perceived as status dimensions, and not only the experience of stressors as included in the current study. Further research might also explore in more depth why some stressors (threat) are associated with operational identity and others are not (action). Besides, even though cohesion seems to be closely connected to operational identity (Johansen et al., 2013a), the results in the current study showed no such association. Further research might reveal whether other samples can establish such an association, for instance samples of standard military units being deployed to international operations.
The results of the factor analysis of the NPIS in the current research excluded idealism and reduced items of individualism. This calls for further investigation of these dimensions. Moreover, the current study found differences between temporarily and permanently employed personnel. Similar studies of temporarily employed personnel might explore why motivation differs between those samples and better address possible measures to counter these effects. Further research might also confirm that this category of personnel is less experienced than permanently employed soldiers as we have claimed in the current research, and whether they always report high levels of horizontal cohesion when they join ad hoc units.

Conclusion
The purpose of the current study has been to explore the motivational aspects of soldiers participating in international operations. The study was carried out by examining changes in military identity. As predicted, the results showed no difference in individualism from pre-deployment to post-deployment. Idealism was not tested due to an invalid factor structure between the two time points, something that should be explored further. The findings also showed that operational identity is confirmed during such service, and that a strengthened identity is associated with the experience of certain stressors (threat). This calls for soldiers to be exposed to threat during their deployment as a part of finding meaning (Lien et al., 2016). The lack of association between operational identity and cohesion in the current study should be investigated further. The findings also revealed a difference in the level of operational identity between temporarily and permanently employed soldiers. Thus, future studies might explore the motivation of soldiers, including temporarily employed soldiers, in order to better design pre-deployment training procedures for all categories of personnel.