Mediating role of job satisfaction in the relationship between organizational identification and organizational commitment: An empirical study on ship agency employees

Understanding the relationship between an organization and the employees in that organization is of paramount importance for the success and future of that organization. This is related to the measurement of the organizational attitudes and behaviours of employees and the use of these attitudes and behaviours for forward-looking predictions. This being said, the aim of the study is to explore the relationship between the organizational identification levels of employees in ship agencies, and their job satisfaction and organizational commitment. For the purpose of this study, the data were obtained from 265 employees of different ship agencies in the cities of Istanbul and Kocaeli, Turkey, through questionnaire. The data then were analyzed via the structural equation modeling analyses, which were conducted using AMOS v22 by Bootstrap resampling with 5000 replications; the results have showed that organizational identification has both a positive direct and an indirect effect on organizational commitment, the latter through job satisfaction. Further, this study revealed that organizational identification, together with job satisfaction, explained about 0.59% of the variation in organizational commitment.


Introduction
Understanding the psychological bond between the individuals and the organization is of a pivotal role in organizational behaviour studies as it shapes the attitudes and behaviours of employees (Rhoades and Eisenberger, 2002). Considering rapidly changing, competitive business environment; employees become the most significant and leading factor to achieve greater success, efficiency, and productivity (Nath and Agrawal, 2015;Güleç and Samancı, 2018). Organizations which desire to enhance their competitive power, need highly motivated, committed, satisfied and innovative human capital (Abou Elnaga and Imran, 2014). For this reason, determining employees' perceptions about their organization is critical to understand and control mechanisms that influence employees' attitudes, and behaviours toward their jobs (Tüzün, 2009).
In the organizational behaviour literature, organizational identification (OI), organizational commitment (OC), and job satisfaction (JS) concepts have received considerable attention since they provide strong, interpretive constructs for work related attitudes and behaviours. Previous studies have reported numerous favorable outcomes associated with these key topics. For instance, OI which reflects 'oneness' perception of members with organization, is known to increase organizational citizenship behaviour (Ashforth, et al., 2008;Tanghe et al., 2010), in-role behavior (Haslam and Ellemers, 2005;van Knippenberg, 2000); extra-role behavior (Riketta, 2005;Lee et al., 2015), job satisfaction (Feather and Rauter, 2004;Yang and Chang, 2008;Alegre et al., 2016), organizational commitment (Meyer et al., 2004;Marique and Stinglhamber, 2011;DeConinck, 2011) and many other desirable outcomes. Present study sets out to reveal the relationships between the OI, JS and OC of the employees working for ship agencies. Though a sizable number of studies exploring these subjects with different variables have been published over years, there has been few papers examining the relationships among all these three variables together (e.g. Arıkoğlu, 2019; Feather and Rauter, 2004;Mete et al., 2016, Sökmen, 2019. In maritime business context, no single study exists which investigates the linkage between those subjects, solely far too little attention has been paid to OC and JS relationship (e.g. Esmaeilpourand Ranjbar, 2018; Larsen et al., 2012;Yücel and Yorulmaz, 2016).
Ship agencies are one of the widely utilized intermediaries in the maritime sector just like freight forwarders and ship brokers. In international maritime trade, such intermediaries play a fundamental role with respect to effective service management by providing high quality, economical, effective, safe and secure logistics, and maritime services (Tuna, 1998). Deveci and Cerit (2007) emphasized that complexity of logistics and transportation services necessitates utilization of these supporting, and facilitating intermediaries such as ship agencies, and forwarders. These intermediaries carry out various crucial tasks including logistics arrangements, market research, legal advice, sales and post-sales services etc. (Ahn et al., 2011). Ship agencies, more particularly, represent ship owners and principles, and they perform multiple significant functions consisting commercial, operational, legal functions and so on (Özaydın, 2016). The primary objective of a ship agency is to protect owner's and ship's interests to accomplish undertaken ship operations with lowest cost and optimum efficiency (Baran and Arabelen, 2018). A wide range of services, from booking, marketing and sales, documentation to container tracking, port services, and customer relations, are offered by port agents (Deveci, 2002). Obviously, employees of ship agencies interact with several actors, and have crucial responsibilities to complete maritime services in the most effective manner. However, necessity of providing fast and high quality service in an intense competitive environment results in time pressure and job stress on agency employees. Unlike seafarers who have been subject to plenty of organizational researches (e.g. Ayap and Macalalad, 2016;Hult and Snöberg, 2014;Yorulmaz, 2018;Yuen et al., 2018) due to challenging, threatening, isolated living and working conditions, there is a lack of research exploring organizational behaviours of employees working for ship agencies. The importance of requirement of a research on ship agencies in Turkey also lies in the fact that the number of companies that received the shipping agency authorization certificate from T.C. Ministry of Transport and Infrastructure is 1251 (DTO, 2019). Although considerable amount of ship agencies, and thus employees are engaged in maritime sector; only few studies regarding organizational behaviour have been conducted (e.g. Baran

Organizational identification and organizational commitment
The psychological linkage between employees and their organization has been mainly conceptualized through OI and OC. The conceptual association of these two phenomena is ambiguous (Sass and Canary, 1991) since many authors have been asserted that both terms are synonym (Mottaz, 1989;Riketta, 2005) while others have argued against the idea by indicating the terms as distinct but related concepts (Gautam et al., 2004;van Knippenberg and Sleebos, 2006;Ashforth et al., 2008).
OI can be loosely defined as 'the degree to which a member defines him or herself by the same attributes that he or she believes define the organization' (Dutton et al., 1994). Another most frequently used definition made by Mael and Ashforth (1992) as 'the perception of oneness with or belongingness to an organization, where the individual defines him or herself in terms of the organization(s) in which he or she is a member.' They approach to OI as a particular cast of social identification. Both definitions are essentially rooted in Social Identity Theory (Tajfel and Turner, 1986) which suggests that individuals' identities that draw from group membership, are key influencers on their self-concept. Put it differently, when a member's belief and thoughts become self-defining, OI emerges (Karanika-Murray et al., 2015). Accordingly, employees internalize organization's norms and values, and perceive organizational membership as a part of their personalities. For instance, van Dick (2001) assert that OI is composed of cognitive, evaluative, affective, and behavioral aspects. Cognitive component refers to awareness of being a part of a certain social group; evaluative component attributes positive judgments regarding the organization; emotional affection to the group is implied by affective component, while behavioral component indicates participation in actions. Through identification with the organization, multiple needs of employees such as belongingness, selfenhancement, safety and uncertainty reduction can be satisfied (Ashforth et al., 2008). A sense of identification exhortates employees to display greater levels of altruism, cooperation, supportive actions, and devote more effort in favor of the organization (Demir, 2015). Moreover, perception of having similar goals, values, and interests with other members may avert employees from being alienated, and leaving the organization (Dutton et al., 1994;Riketta and van Dick, 2005). Feeling of attachment and loyalty to the organization stimulates employees' intrinsic motivation to support organizational processes, take more responsibility and improve job performance ( Ng, 2015).
OC, on the other hand, has also drawn a great deal of academic attention over the last few decades owing to its importance for organizational performance and effectiveness (Nath and Agrawal, 2015). The most prominent definition of the OC proposed by Mowday, Steers, and Porter (1979) as 'the relative strength of an individual's identification with and involvement in a particular organization'. For Wieselsberger (2004), it is a psychological and emotional state that shapes the relationship between the employee and his/her organization, in such a way that affects the decision about maintaining membership of the organization. Most of the researches conducted on OC have employed the Meyer and Allen's (1990) widely-accepted model which comprises three components: The first aspect is affective organizational commitment (AOC), and is identified as 'the employee's emotional attachment to, identification with, and involvement in the organisation'. The second aspect, continuance commitment (CC), indicates employees' unwillingness to quit job due to associated costs (i.e. time, status, benefits etc.) with leaving the organization. The third aspect is called normative commitment (NC) which stems from sense of obligation to keep working for the organization because of thought that is a moral duty (Uçanok and Karabatı, 2013;Yousef, 2017). These definitions may pose a confusing problem for distinguishing OI from OC, especially affective version due to conceptual overlap between the notions (Sass and Canary, 1999; Bergami and Bagozzi, 2000). However, there exist several distinctive landmarks to differentiate these two concepts. First of all, even both terms denote employee-organization relationship, OI explains the link in terms of self-referential concept while OC does not (Edwards, 2005). Employees who strongly identified with their organization always tend to give priority to organizational goals and values, also think and behave for sake of group since perceived those as a part of their self-concept. Secondly, OI is greatly flexible, contingents upon the salience of the social group and the context of communication with other groups. Whereas, OC is viewed as an attitude that, once developed, is relatively steady and long-lasting (Guatam et al., 2004). The sources of both concepts also bring about third distinctness. Perceived similarity and joint fate within the organization become grounds for identification (Mael and Ashforth, 1992), while OC is expounded via reciprocal benefits of employees' and their organization. In other words, satisfaction of socio-emotional needs (payment, recognition, support etc.) of the employees encourage them to be more committed (Rhoades and Eisenberger, 2002).
Reviewing the literature reveals that OI has a significant impact on work behaviours. For instance, Mete, Sökmen and Bıyık (2016) examined the link among OI, OC, JS, and person-organization fit with respect to IT specialists. They reported positive correlation among those four variables. Supportively, Lee and colleagues (2015), in their comprehensive meta analysis research, found that when individuals become identified with their organizations, it is more likely to followed by satisfaction and commitment. Authors asserted that having identified employees takes relatively precedence over employing those satisfied or committed individuals to improve organizational performance. Another meta analysis provided by Riketta (2005) confirmed that OI is closely associated with JS, OC, intent to leave, and many other several work-related outcomes. DeConinck (2011) posited that OI indirectly effects turnover rates of employees through OC. In the same vein, Ng (2015) noted the significance of OI and organizational trust as impetus of OC. He asserted that those two factors create strong motivation for employees to experience greater level of OC. A number of studies support those propositions by demonstrating that OI promotes OC

Organizational identification and job satisfaction
OI is vital for organizations since it affects why individuals decide to participate in or quit their organization (Ashforth et al., 2008). Extensive theoretical and empirical studies have consistently accentuated JS as an essential determinative factor of whether employees stay in or withdraw from an organization (Aziri, 2011;Bäker and Goodall, 2020). Being the most frequently investigated attitude within the organizational behavior literature, JS refers to the degree to which an employee likes his/ her job. It reflects pleasurable emotions stemming from judgment toward one's job experience based on how well his/her needs and expectations are met with what got from the job (Locke, 1969). There are several environmental (organization-related) and individual elements that influence the JS level of employees within the organization. According to Cumbey and Alexander (1998), JS occurs contingent upon not only expectations about work environment, but also interaction of employees, their personality traits and values. Five determinants the mostly mentioned in the satisfaction studies are management/supervision, relations of co-workers, pay levels, nature of the work, and promotion and carrier development (Peterson et al., 2003;Luthans, 2010; Baran, 2016). Understanding all elements that motivate and satisfy employees' needs is a key strategic move and the backbone for organizational success. For instance, Ayan (2005) maintained that OC, flexibility, human resources values, goal congruence, working with harmony, managerial skills, and ensuring JS are among the key determinants of ship agents' performances.
Employees who feel strong identification with their organizations are more likely to interpret work environment more favourably thanks to escalation of satisfaction. As strongly identified employees construe their jobs as an evidence of their membership to organization, those people tend to perceive their job positively. Put it differently, OI cognizantly or unconsciously leads employees to suppress or ignore detrimental, negative conditions regarding their jobs (van Dick et al.

The mediator role of job satisfaction
JS which is characterised by multi-faceted structure, has been investigated much more often compared to other job characteristics. An employee who is satisfied with his/her job may develop positive reactions toward his/her organization such as better performance, higher OC, more productivity, lower absenteeism and turnover intentions (Kim et  Concordantly, Qasim and Sayeed (2012) asserted that satisfaction perception makes employees more loyal, committed, productive and creative. Satisfied employees desire to keep working for their organization. However, based on their meta-analytical work, Güleç and Samancı (2018) suggested that both OC and JS levels of employees may differ from one sector to another. To illustrate, they found those variables are the lowest in the health sector. Studying JS of employees working in İzmir-based ship agents, Karaman (2009) revealed that satisfaction level of those employees are above average. Likewise, Baran and Arabelen (2018) stated that JS is of relatively higher prominence in service sectors, but different dimensions of satisfaction may lead to different levels of satisfaction. They concluded that co-worker relations and job security are the most satisfying components for ship agents' employees, while compensation and advancements are the least ones. Furthermore, they stated that increase in age, seniority and work experience brings about higher satisfaction. Another study carried out by Yumuşak  The conceptual model of the study points out the direct impact of OI and JS on OC. These expectations have been supported by various empirical evidences as stated above. Nonetheless, in addition to having direct influence, JS is also predicted to facilitate to clarify the linkage between OI and OC. Lok and Crawford (2001), for instance, found that a good deal of precursors of OC are either partially or entirely mediated through JS. It is not surprising when considering intersection position of JS for many models and theories which expound individuals' attitudes and behaviours (Mete et al., 2016). Due to explanatory power of the variable, JS has been used to clarify a degree of changes in work related outcomes. We propose JS as a mediator in the OI and OC relationship owing to several reasons. Firstly, plethora of evidence have presented positive correlations between OI and JS (e.g. Riketta 2008), on the other hand, revealed that emotional intelligence influences OC purely via JS. Since both OI and OC reflects psychological and emotional bonds, JS that reflecting pleasurable emotions is expected to bridge and support these two variables in some extent. Thereby, in the light of these arguments, we finally proposed that:

Data collection and sampling
This study seeks to reveal the casual relationship between OI, JS and OC in ship agencies based on the data obtained from 265 employees of ship agencies through face-to-face interviews and a questionnaire using Google form between May and August, 2020.

Measures instruments
The measurement tools used in this study were developed as 5-point Likert type instruments (1 = I certainly disagree, 5 = I totally agree); the reliability and validity of these tools were confirmed by previous studies. As these tools were applied to a different sample group, their reliability and validity were analyzed once again.
To measure OI, this study drew on the scale developed by Mael and Ashforth (1992). This scale consists of one dimension and six items. The scale of JS, which was previously used by Wright and Cropanzano (1998), has one dimension and five items. The explanatory factor analysis (EFA) performed to test the structural validity of the measurement tools, showed that the item that "If a story in the media criticized the ship agency that I work, I would feel embarrassed" in the scale of OI and the item that "All in all, I am satisfied with the supervision" in the scale of JS had low factor loads (0.27-0.32), and thus they were removed from the analysis. The scale of OC used in this study was developed by Meyer and Herscovitch (2001). The scale consists of one dimension and six items such as "I am willing to go the extra mile to make the agency that I work successful."

Statistical analysis
To determine the direct and indirect effect of the OI of employees in ship agencies on their OC and on JS, this study performed respectively frequency analyses, an EFA, normality and reliability analyses using SPSS Statistics program. Also, the structural validity of the measurement model was tested through confirmatory factor analysis (CFA); the research hypotheses were tested using AMOS 22 program and the Sobel test were conducted by Bootstrap resampling with 5000 replications.

Validity and reliability analysis
The results of the EFA using the principal components analysis and Varimax rotation, which was performed to determine the structural validity of the measurement model, demonstrated that there were two items with factor loads less than 0.50, which were then removed from the scale. The second EFA yielded a three-factor structure that consists of items with eigenvalues greater than 1 and grouped under the corresponding factors (KMO=0.897; χ 2 = 2191.304; df = 105; p < 0.001). Accordingly, OC explained 24.83% (factor loads varied between 0.627-0.807); JS explained 20.81% (factor loads varied between 0.774-0.839) and OI explained 19.50% (factor loads varied between 0.565-0.775) of the total explained variance of the factors in the measurement model and they overall explained 65.14%. Table 1 presents the standardized factor loads, which were yielded by the CFA performed after the EFA on the measurement tool, t-values and  Table 1 shows that the AVE values were higher than the ASV and MSV values and Table 2 shows that the square root values of the AVE values were greater than the Pearson's correlation coefficients, which confirms the discriminant validity of the tool as well (Fornell and Larcker 1981;Hair et al., 2014).
To test the reliability of the model, which was considered valid, the CR and Cronbach's Alpha (CA) coefficients of the measurement model were calculated, as can be seen in Table 1. It is clear that both CR and Cronbach's Alpha (CA) coefficients were above 0.70, and thus the model was considered as a reliable model (Hair et al., 2014).
The coefficients of skewness and kurtosis were measured to reveal whether the variables in the measurement model, which was already considered as structurally valid and reliable, were normally distributed or not. Also, the means of the variables as well as the Pearson's correlation coefficients are presented in Table 2. As the coefficients of skewness and kurtosis of the variables in Table 2 were in the range of ±1.5, it can be argued that the data were normally distributed. Further, Table 2 shows that the OI, JS and OC means of the employees were high and the correlation between them was moderate. Figure 2 shows the results of the structural equation modeling analyses, which were conducted to test the research model using AMOS v22 by Bootstrap resampling with 5000 replications. As seen in Figure 2, the goodness of fit indices (χ 2 /sd = 2.372; CFI = 0.945; TLI = 0.933; RMSEA = 0.050; SRMR = 0.072) in the measurement tool were within the acceptable values (Hu and Bentler, 1999;Hair et al., 2014); the total effect (β = 0.691; p < 0.01) and direct effect (β = 0.485; p < 0.01) of the OI on the OC were positive and statistically significant, and the positive effect of the OI on the JS was also statistically significant and explained 27.5% of the variation in the JS (β = 0.524; p < 0.01; R 2 = 0.275). The combined effect of the OI and JS on OC was also examined and it was found that the effect of OI was persistent but decreased, and the JS had a positive impact on the OC (β = 0.393; p < 0.01) and they together explained about 59% of the variation in the JS. Based on these results, Hypotheses 1, 2 and 3 were all supported. The mediating role of JS was assessed by considering the confidence intervals obtained by the Bootstrap analysis measuring indirect effect. The confidence intervals in the Bootstrap analysis were statistically significant (β = 0.206; p < 0.05; 95% CI [0.116; 0.314]) and did not include 0; from this it follows that JS is an mediating variable between OI and OC (MacKinnon, Lockwood and Williams, 2004). Further, the intermediary role of JS was also analyzed through the Sobel test. The results showed that its intermediary role is statistically significant (z = 3.798; SE = 0.067; p < 0.001). Based on this result, Hypothesis 4 was supported as well.

Conclusion and discussion
This study seeks to reveal the casual relationship between organizational identification, job satisfaction and organizational commitment in ship agencies based on the data obtained through face-to-face interviews and questionnaire using Google form between May and July, 2020. The data, gathered from 265 employees in ship agencies in the cities of Istanbul and Kocaeli, were analyzed through path analyses in structural equation modeling; the results supported the four research hypotheses.
The findings of this study reveal that the OI of the employees in ship agencies have both a positive direct and an indirect effect on their OC, the latter through their JS. That is to say, the JS levels of the employees, who identify themselves with their organizations, increase, which in turn results in increased OC. Because the results of this present study highlight that OI both directly increases OC and indirectly enhances JS, which positively affects OC. This also accords with Lee et al. (2015)'s findings suggesting that sense of identification followingly ma-sense of identification followingly makes individuals more likely to develop JS and OC since perception of organizational identity as self-referential, motivates individuals to react more positively towards both their job, and also organization. The results of this study are congruent with results of considerable amount of past researches which have studied organizational identification as a precursor to organizational commitment (e.g. Bergami  ted that JS either totally or partially mediates the reliationship of OC with a great number of antecedents of it. The result is not surprising when considering Social Exchange Theory. Accordingly, fulfillment of employees' socioemotional needs stemming from OI affects their loyalty, effort, involvement, and evaluations. To the extent that employees are satisfied with this reciprocal relation, they will encouraged to be organizationally committed. In other words, perceived quality of exchange relationship leads employees to increase their commitment. Consistent with this reasoning, all hypotheses were supported.
The value of this research emanates from its novel findings and contributions to the both maritime and management fields. Being an international, dynamic, busy and customer-oriented circle of trade, understanding the motivational factors of ship agency employees is essential for sustainable competitive advantage, effective and efficient work processes, and also customer satisfaction. Obviously, managers who are aspiring to achieve sustainable organizational success are expected to be more aware, supportive and motivating. Therefore, this study provide a promising basis not only theoretical but also practical implications for the future researches.
The sample of this study consists of only employees in ship agencies, which operate in Istanbul and Kocaeli, and this is certainly a limitation of this study. For this reason, further studies may include larger sample groups with all employees in ship agencies across Turkey or test the research model with employees in different sectors.