Perceived organizational support and work engagement: the role of psychosocial variables

Purpose – This study aims to test the role that organizational sociopsychological variables may play in in ﬂ uencing jobstress andworkengagementin anorganizationalidentitychange scenario. Design/methodology/approach – On a sample of 118 employees of an Italian company in the personnel training services sector, multivariate statistical analysis tests a pattern where organizational variables such as work support (by supervisors and coworkers, independent variables) – moderated by corporate identi ﬁ cation (moderating variable) – and mediated by organizational trust (mediating variable) – boosts employeework engagementandlowerspsychosocial risks (dependentvariables). Findings – The mediating effect of “ organizational trust ” is signi ﬁ cant in the relationships of “ supervisor social support ” and “ coworker social support ” with the “ absence of psychosocial risks. ” Moreover, an increase in supervisor social support can lead to a statistically signi ﬁ cant increase in work engagement. This occurs only for employees with lowormediumidenti ﬁ cationandnotin highly identi ﬁ ed individuals. Originality/value – The ﬁ ndings from the analysis on moderation are of primary importance because they show us a new perspective that can play the role of a guiding and practical principle on how to act on an organization ’ s humanresources, speci ﬁ callytargeting those with loweror medium corporate identi ﬁ cation.


Introduction
In the organizational context, there are multiple dimensions that can be articulated in concrete and tangible or intangible dimensions. It is already well known that nurturing intangible dimensions, such as human capital and social capital, will have a positive impact on very concrete outcomes, such as performance and financial capital (Luthans et al., 2004). A more in-depth analysis of intangible aspects can include a slightly different interpretation of those assets: for example, social capital is referred to the network of contacts and relationships, but it can also involve "processes of empathy, trust, commitment, satisfaction and happiness that articulates negative and positive emotions towards supervisors, colleagues or organizational processes" (Lirios et al., 2018) (p. 11). The dimension of psychological capital, which includes four positive psychological states (confidence, hope, optimism and resilience), can be a core element of work engagement (Stajkovic, 2003;Benevene et al., 2018). As some authors have noted (Luthans et al., 2004), a higher level of the four components of psychological capital can provide employees with more job satisfaction and performance.
Psychological capital is framed in the theoretical context of positive psychology, and indeed, there are evidences of its positive impact on job attitudes and behaviors (Avey et al., 2011): we will frame both individual and collective positive dimensions of work in the context of psychological capital. The intangible dimensions that will be taken into account are explained in the following paragraphs.
Specifically, this research contributes to the existent literature by highlighting the central role of perceived organizational support in the improvements of work engagement and jobrelated stress coping; in terms of more practical implications, it will be discussed how this kind of research contribution could be the basis to develop possible training interventions oriented to work engagement raising.
Therefore, the objective of this study is to fill a gap in the literature consisting in the lack of a model relating the abovementioned variables. Basically, this research fills the gap regarding how these intangible assets can influence the performance and health of a company's employees and how particularly social relationships and the working and organizational climate are relevant aspects for promoting a healthy and productive work environment.
In addition to advancing the literature on the subject and proposing a comprehensive model to assess the social-psychological variables examined in this study (organizational trust and corporate identification), this contribution also discusses a possible practical implication of the results, in the terms of training which could be developed for organizational personnel such as specifically the managerial roles.
Below the relevant constructs are defined and, on the basis of the existing literature evidences, their relations are summarized to show the need of specifically testing how organizational trust and corporate identification are intangible relevant psychosocial features capable of, respectively, mediating and negatively moderating the effects that workplace social support (both from the manager or supervisor and from colleagues) exerts Psychosocial variables both by boosting the employees' work engagement and by buffering their psychosocial risks.
1.1 Perceived organizational support Social support has a relevant both preventive and protective function that can reduce general stress (Kenny et al., 2014;Kenny and Hage, 2009). Framed in the work context, social support is defined as perceived organizational support, and it is considered as the commitment of the organization toward the employees. It consists of two complementary dimensions: the colleagues' and the supervisor's perceived support. The colleagues' perceived support includes practical support and information related to tasks, other than socio-emotional support and empathy (Rousseau et al., 2009). The perceived supervisor support has been defined as the employees' perception of how much their supervisors are supportive and caring about the workers' well-being . Empirical researches demonstrated that support from colleagues is positively related to both job satisfaction and job and personal commitment (Chiaburu and Harrison, 2008;Karatepe et al., 2010;Galanti et al., 2021). Motivation in the workplace can also be positively affected by the colleagues' support (Karatepe et al., 2010). With regard to the supervisors' perceived support, there are some studies conducted by Malik et al. (2015) that found that a good relationship between supervisor and employee can directly and positively affect job satisfaction. It then emerged how a good supervisor-collaborator relationship also affects job satisfaction indirectly, through the mediation of motivation. Situational leadership theory (Hersey and Blanchard, 1969) also supports this line of research. This leadership style, characterized by the figure of the leader as a partner who relates horizontally with individual employees trying to meet their needs, promotes the development of motivation and performance and reduces stress levels in employees.

Organizational trust
Another relevant theme, which is central in the definition of the organizational climate, is organizational trust, defined as the employees' trust toward the management with the expectation to receive more positive than detrimental actions (S ims ek and Tas çı, 2004; Erdem and _ Is bas ı, 2000). There are several studies in the literature stating that the perception of organizational support plays an important role in providing organizational trust. Indeed, if employees feel support from their colleagues, their trust toward the organization may increase (Canipe, 2006;Polat, 2010;Narang and Singh, 2012;Duffy and Lilly, 2013). The same effect has also been tested with supervisors' support (Kara et al., 2015).

Psychosocial risks in workplace
Numerous studies on social exchange Stamper and Johlke, 2003;Viswesvaran et al., 1999) have suggested that social support should also help employees to cope with stress. In fact, Rhoades and Eisenberger (2002) showed that perceived organizational support has a moderate negative relationship with job strain. Thus, it can be inferred that high perceived organizational support indicates to workers that emotional support and help are available from the organization when they would face high job demands, reducing their psychological stress. Moreover, there are several studies that corroborate the positive relation between perceived organizational support and the reduction of job-related stress (Jung and Yoon, 2013;Loi et al., 2014;D'Alterio et al., 2019), especially when workers find themselves in social minority situations or when they encounter tasks that are extremely complicated. Another indirect relation is mediated by JWL 34,5 personal job motivation in decreasing the risk of job-related stress. Past studies have already analyzed the relation between organizational support, organizational trust and stress, and it has been tested a moderator effect of trust and support in the relation between perceived organizational politics and job-related stress (Vigoda-Gadot and Talmud, 2010). However, their moderation effects are not equally addressed.

Organizational identification
According to social identity theory (SIT), people tend to classify themselves and others according to various social categories such as organizational affiliation, religious affiliation, gender and age (Tajfel and Turner, 1985). SIT also argues that individuals identify themselves with social categories to enhance their self-esteem (Hogg and Turner, 1985;Tajfel, 1978).
The organization represents an essential reference point for the individual to answer the fundamental question "who am I?," to shape and support the process of social and personal identity construction (Ashforth and Mael, 1989).
Ellemers has taken up and applied the assumptions of SIT to the study of organizational identification with particular attention to group dynamics (Ellemers, 1993;Ellemers et al., 1999). Unlike Ashforth and Mael (1989), Ellemers et al. (1999) did not make a distinction between identification and commitment: in essence, according to the authors, identification, involvement and commitment to one's own work are closely related and mutually reinforcing.
Several studies highlighted the positive impact of organizational identity on job motivation (van Knippenberg, 2000;van Knippenberg and van Schie, 2000); other past researches found that a good relation between supervisor and employee can positively affect organizational identity (Katrinli et al., 2008).
There have been some studies investigating the link of organizational identification with the other considered constructs: for example, Dai and Qin (2016) empirically tested that perceived organizational support can play a role in employee engagement through organizational identification. In their study (Dai and Qin, 2016) they found that organizational identification has a positive mediating effect between perceived organizational support and work engagement. Since a moderating effect of organizational identification on the other two variables has never been tested, it can be hypothesized that less identified individuals need more support from their supervisors so that there is an increase in identification, which in turn can lead to a subsequent increase in work engagement, and that therefore this moderating relationship may be negative.
In support of this hypothesis, there are some previous studies, which have analyzed the nonlinearity of organizational identification in correlating with other psychosocial variables such as transformational leadership and employee adaptability (Wang et al., 2017) but also with variables such as supervisor support and commitment to change (Zappalà et al., 2019).
Indeed, in Zappalà et al.'s (2019) study, results showed that supervisors can indirectly decrease employees' perceptions of cost, risk or pressure to adhere to change by increasing their sense of identification with the organization.
Finally, Wang et al. (2017) showed a negative interaction between the dependent variable (transformational leadership) and the moderating variable (organizational identification) of Psychosocial variables their model in predicting employee adaptability and indirectly job crafting. Thus, transformational leadership appears to be more effective in indirectly promoting job crafting through increasing employee adaptability, especially when employees are less identified with the organization. According to our literature review, the negative moderation effect of the organizational identity between perceived support and work engagement has never been tested before. Schaufeli and Bakker (2004) described work engagement as an active, positive, work-related state that includes three dimensions: vigor, dedication and absorption. Past studies have confirmed the relevance of work engagement also with regard to organizational citizenship behaviors: for example, it was found that lower levels of work engagement are associated with higher levels of turnover intentions and deviance, especially when employees do not perceive organizational support (Shantz et al., 2016). At the same time, work engagement is affected by several organizational dimensions, such as culture, communication, leadership and corporate reputation (Köse, 2016).

Work engagement
Regarding the link among work engagement and the other considered constructs, we can cite some previously reviewed studies (van Knippenberg, 2000;van Knippenberg and van Schie, 2000) in which the positive relationship between organizational identification and work engagement is emphasized. The study by Dai and Qin (2016) cited above also relates our same variables, demonstrating how greater perceived organizational support can predict an increase in organizational identification and, in turn, work engagement.
There are also other studies that relate these variables in a different way from the one cited above and from how we hypothesized. For example, the study by Köse (2016) showed that greater work engagement can lead to a greater perception of support from one's colleagues and superiors. Karanika-Murray et al.'s (2015) study, on the other hand, highlights how an increase in organizational identification can translate into an increase in performance and how this relationship is positively mediated by work engagement. Finally, Najeemdeen et al.'s (2018) study shows that perceived organizational support has a significant positive influence on employee engagement.

Social support, organizational trust and work-related stress
Various studies, such as Vigoda-Gadot and Talmud's (2010) one, highlight through a correlation analysis how variables like social support in the organizational environment, organizational trust and work-related stress are significantly related to each other. In fact, this study specifically shows a significant positive relationship between social support and organizational trust, while it shows a significant negative correlation between work-related stress and both social support and organizational trust. These correlations led us to develop hypotheses H1a and H1b, expressed below.
1.7 Social support from supervisor, organizational identification and work engagement There are various researches that have related these variables to each other. Leadership style has been associated several times with motivation, such as by House (1971), who, theorizing his path-goal theory, states that the leader's behavior is contingent on some psychosocial variables such as job satisfaction or motivation. In particular, he emphasizes that leaders who are "supportive" and therefore able to assist their subordinates are those who are most successful in raising the levels of engagement in their employees.
It is also known as found in Ötken and Erben's (2010) study that organizational identification correlates positively with work engagement and that supervisor support can JWL 34,5 moderate this relationship. This shows us, although the constructs are examined differently than in our study, the close relationship between organizational identification and supervisor support in predicting work engagement. International literature underlines the key role played by organizational support, especially during organizational socialization, on employees' engagement (see, for example, Saks, 2006). For what concerns, stricto sensu, the supposed moderation played by organizational identity, this took origin from a niche literature that stresses the nonlinearity relationships between organizational support and several individual positive outcomes (see, for example, Rosen et al., 2017;Harris and Kacmar, 2018).
In light of the past literature and in line with the theoretical framework of psychological capital, the logical-theoretical reasoning that led us to formulate the following hypotheses derives from the results of previous studies on the subject. In fact, with regard to hypotheses H1a and H1b, referring mainly to the study of Vigoda-Gadot and Talmud (2010), we started from the correlations examined in that study to hypothesize a mediation model in which organizational trust (intrinsic variable in the individual worker) can mediate an external variable to the worker (perceived organizational social support) with the aim of predicting lower psychosocial risks. This is very important to understand how the health of individual workers is strongly influenced by the work environment and by the relationships they establish with their colleagues and supervisors or managers. However, it is important not to underestimate the personal sphere composed of emotions, feelings and evaluations that very often can influence the effects that work environment and colleagues have on the individual.
As for hypothesis H2, we have placed organizational identification as a moderator because the impact that the supervisor behavior has on his/her coworkers also depend on the level of maturity, identification and experience of the latter. Indeed, there are several studies that suggested various moderators or mediators in the relations between the leadership style and followers' behavior. The follower maturity is strictly related to the individual attitude to take responsibilities, to the commitment to organization and to the level of professional development (Harber and McMaster, 2018). There are three levels of maturity based on different amounts of competence and commitment; according to Harber and McMaster (2018), followers at a lower level of maturity are better supported by a transactional servant leadership approach, which can improve followers' development and autonomy. Moreover, the study by Zhu et al. (2009) tested a positive relation between transformational leadership and followers' work engagement; they also assessed that this relation is moderated by some followers' individual characteristics, such as creativity, innovation, proactivity and learning orientation (Zhu et al., 2009).
Therefore, we hypothesized that support from supervisor could increase the employee's engagement to work, but that this increase was significant only in employees who were poorly identified with the organization.
In line with this, we have formulated the two following main hypotheses, presently not yet fully statistically tested by previous existing references: H1a: We do hypothesize a direct link between the perceived colleagues' support and the organizational trust, which in turn could reduce the risk of job-related stress (trust mediation effect of the colleagues' support on stress).
H1b: We do hypothesize a direct link between perceived supervisors' support and organizational trust, which in turn could reduce the risk of job-related stress (trust mediation effect of the supervisor's support on stress).

Psychosocial variables
Said in different terms, employees who perceive higher organizational support (from either the supervisor and/or the colleagues) will develop trust in the organization, which will in turn make it more difficult for them to experience stress: H2: We do hypothesize that organizational identity has a moderator role in the relation between perceived supervisor's support and work engagement. In other words, we suppose that the support from the supervisor succeeds to increase the work engagement of the workers who are less identified with the organization rather than those who are more identified with it, as the first ones would need it more than the second ones (as they are already work engaged on the basis of their higher organizational identification).

Materials and methods
This research has taken place during the year 2018, on the basis of the request made by an organization providing employment services: their expressed need was to identify aspects that could increase the social identity at both the teamwork level and the overall organizational group, due to the change that the company was experiencing (i.e. a brand change and market positioning). Such a turmoil was a challenge especially in terms of identity, since the company group was going to acquire a new corporate image, which was also going to be externally reflected (on customers). An online questionnaire had been administered through the Qualtrics platform: all employees were invited to anonymously participate with an official launch e-mail to orient employees on the survey in progress and to keep a high employee engagement.
Tables 1 and 2 describe the sociodemographic characteristics of the sample. The survey was named "When the company changes name," and it was developed to analyze the impacts of organizational change on the psychosocial sphere of work relations, on organizational well-being and on social identity. This questionnaire consists of six evaluation scales, for a total of 64 response items. The scales included in the questionnaire are described as follows: Organizational identification, taken from the scale six items measure of organizational identification of Mael and Ashforth (1992), where an example item is "When someone criticizes the company I work for, I perceive it as a personal insult". Work engagement was measured through the nine-item Utrecht Work Engagement Scale by Schaufeli and Bakker (2003), with items like "When I work I feel full of energy". Organizational trust was divided into eight dimensions: goals, valuing, listening, information, relationships, operations, health and safety and social utility, and all dimensions refer to the organizational health questionnaire by Avallone and Paplomatas (2002), formed by items such as "The organization finds appropriate solutions to the problems it faces". Perceived supervisor social support, using a scale from Caplan's (1975) Job Stress Questionnaire (JSQ) called social support from supervisor, features items such as "In this time of change, how much help do they offer when faced with a serious work problem?". Perceived colleagues social support was measured through a scale of the JSQ of Caplan (1975) called social support from others at work, with item like "At this stage of change, how willing are they to listen to your professional concerns?". Absence of psychosocial risks (stress) is a multidimensional scale, and it includes negative indicators, indicators of psychophysical discomfort, absenteeism; the scale is taken from various scales of the organizational health questionnaire by Avallone and Paplomatas (2002), and an example item is "Impatience in going to work: existence of a daily difficulty in going to work." In Table 3, the descriptive characteristics of the scales are reported.

Analyses
Preliminary analyses were oriented to explore the sample and the measurement scales; these were carried out with the software IBM statistical packaging for social science (SPSS) Statistics 25.

Psychosocial variables
H1a and H1b are aimed at investigating the effects of the different mediation models, and they were analyzed one at a time by taking either "Social support of supervisor" or "Social support of co-workers" as independent variables, with "Organizational trust" as the mediator variable and "Absence of psychosocial risks (stress)" as the dependent variable.
To carry out these analyses the macro "PROCESS v3.5 by Andrew F. Hayes" of the software was used; more precisely, the two mediations were carried out using the "model number 4" of the macro.
With regard to the moderation hypothesis (H2), the aim is to investigate the moderating effect of the variable "organizational identification" (moderating variable) on the relationship between "supervisor's social support" (independent variable) and "work engagement" (dependent variable). In this case, the "model number 1" of the macro by Hayes was applied.

Results
To test the theoretical models, first, the normal distribution and reliability of all the investigated constructs was verified through the calculation of Cronbach's a (see Table 3).
With regard to the factors "Health and Safety Protection" and "Social Utility," it was not possible to calculate the reliability through Cronbach's a, as they are composed of only one item.
For the multifactorial variable "absence of psychosocial risks," consisting of 15 items, the reliability of the factors "indicators of psychophysical discomfort" and "absenteeism" could not be calculated, since both consisted of only one item.
3.1 Mediating effect of "organizational trust" between "social support from colleagues" and "absence of psychosocial risks (stress)" In this hypothesis, which analyzes the mediating effect carried out by "organizational trust" between "social support from colleagues" and "absence of psychosocial risks (stress)," the results showed a total mediating effect. In fact, while "social support from colleagues" does not directly predict the "absence of psychosocial risks" (coeff = 0.0583; p = 0.3803), the latter is predicted by "organizational trust" (coeff = 0.7164; p = 0.0000). As shown in Table 4 and Figure 1, for the entire regression model, the total effect is positive and significant (effect = 0.2914; p = 0.0000). Considering that an increase in "social support of colleagues" predicts an increase in "organizational trust" (coeff = 0.3255; p = 0.0000), this will also lead to a lower chance of developing psychosocial risks (stress). The fundamental role played by the mediator variable organizational trust is thus confirmed with respect to the relation among colleagues' support and absence of psychosocial risks.
3.2 Mediating effect of "organizational trust" between "social support from supervisor" and "absence of psychosocial risks (stress)" From the analysis, it emerged (Table 5 and Figure 2) that with regard to this hypothesis the regression model presents a "total" mediation. In fact, the independent variable directly predicts only the mediator variable (coeff = 0.2858; p = 0.0000) and not the dependent variable (coeff = 0.0849; p = 0.1004). The inclusion of the mediator variable, which significantly predicts the dependent variable (coeff = 0.6233; p = 0.0000) makes the entire model significant (effect = 0.2631; p = 0.0000). This means that the "social support from supervisor" does not directly predict the "absence of psychosocial risks"; however, with the inclusion of "organizational trust" in the model, it manages to indirectly predict the absence of stress. JWL 34,5 3.3 "Organizational identification" moderation effects on relationship between "social support from supervisor" and "work engagement" This hypothesis considers the moderation of "organizational identification" in the relationship between "social support from supervisor" and the employees' "work engagement." Results (Table 6 and Figure 3) showed that the moderating effect of "organizational identification" is significant and negative (coeff = À0.2033; p = 0.0027). Considering that the relationship between "social support of the supervisor" and "work engagement" as well as the relationship between "organizational identification" and "work  Psychosocial variables engagement" were both significant (respectively, coeff = 1.0496, p = 0.0003; and coeff = 1.2562, p = 0.0000), we can draw important conclusions. In fact, these results show us that in the subjects with a lower degree of organizational identification, an increase in social support from the supervisor can lead to a statistically significant increase in work engagement. Therefore, as depicted in Figure 4, the positive effects of the supervisor social support upon the employees work engagement happens only for employees who are poorly or averagely identified, while on the other hand, in the highly identified subjects, the increase in work engagement due to the supervisor's support is nonsignificant (coeff = 0.29,

Discussion
The results of the study confirmed all our hypotheses. Starting from the H1a hypothesis, as verified by numerous studies (Canipe, 2006;Polat, 2010;Narang and Singh, 2012;Duffy and Lilly, 2013), it was found that support from colleagues in the workplace can predict an increase in organizational trust in employees. Similarly, a direct relationship between organizational trust and lower psychosocial risks has been verified as previously found in the study by Lambert et al. (2021), showing that certain types of organizational trust, such as trust in supervisors or in colleagues, have positive effects on employee stress. Finally, the analyses revealed that the mediating effect by organizational trust is "total." In fact, while the direct relationship between the independent variable and the dependent variable is nonsignificant, the model gains significance with the inclusion of the mediating variable (organizational trust). We can therefore highlight the importance of organizational trust in the relationship between the   -Gadot and Talmud (2010). Unlike the existent literature, this study emphasizes how the interaction between support from colleagues and organizational trust is essential to try to reduce stress in the workplace. In fact, peer support alone does not appear to have a significant effect on employee stress if employees do not have feelings of trust in their organization and management. With regard to the results that emerged from the analyses on the H1b hypothesis, as already exposed by Kara et al. (2015), support from supervisors turns out to be a direct predictor of organizational trust. The latter then, as also tested in hypothesis H1a and in the study by Lambert et al. (2021), was found to have a positive and significant correlation with the absence of psychosocial risks. Finally, as with hypothesis H1a, the direct relationship between the independent variable and dependent variable was not found to be significant. However, with the inclusion of the mediating variable (organizational trust), the entire model turns out to be significantly predictive of the absence of psychosocial risks. This is consistent, as with the H1a hypothesis, with the study by Vigoda-Gadot and Talmud (2010). Again, as found in the analyses of the previous hypothesis, the central role of organizational trust is highlighted. As with the H1a hypothesis, these results are innovative and significant as such a model has never been tested in the literature. The latter suggests the importance of developing a feeling of organizational trust in employees in such a way to make them more receptive to the positive stimuli and support they might receive from their organization, with the final outcome being positive repercussions both for individual employees and for the organization itself.
Finally, the analyses performed to test hypothesis H2 revealed the most interesting and innovative results of our study. In fact, both supervisors support and organizational identification were found to be positively correlated with an increase in work engagement. This is in line with the literature on the topic, as Malik et al. (2015) had already emphasized a direct relationship between supervisors support and work engagement. Also regarding the Moderating effect of "organizational identification" in the relationship between "support from supervisor" and "work engagement JWL 34,5 relationship between organizational identification and work engagement, this was previously tested by van Knippenberg (2000) and van Knippenberg and van Schie (2000) with similar results. However, the most innovative result emerges from the analysis of the interaction between the independent variable (supervisor support) and the moderating variable (organizational identification) in predicting the dependent variable (work engagement). In fact, unlike some previous studies (Ötken and Erben's, 2010), the results showed a negative interaction between supervisors support and organizational identification. This means that less identified employees benefit more from supervisors support. This is plausible and in line with some leadership theories such as situational leadership (Hersey and Blanchard, 1969) or the concept of servant leadership (Stone et al., 2004), which emphasizes that supervisors must establish different relationships based on the employees with whom they relate. These results are innovative from a scientificacademic point of view, since they present a previously untested model and propose a negative interaction between supervisors support and organizational identification not hypothesized and adequately tested by the existing literature. They are also innovative from a practical application point of view. Indeed, the results that emerged from the analysis regarding the moderation of "organizational identification" in the relationship between "social support from supervisor" and "work engagement" are of primary importance because they show us a novel perspective that has not been previously tested in the literature. This leads to several conclusions and implications, which could help to interpret various organizational phenomena and properly manage human resources, over and above the general indication to nurture psychological capital and the organizational intangible assets.
The results show that an increase in social support can create an optimal climate in an organization by promoting health, work engagement, trust in the organization and, above all, the organizational identification of employees.
The study has some limits: first of all it is cross-sectional, and data cannot be generalized, also because the sample is from a very specific business sector (employment services). Another limitation presented by this research is the small sample size, which did not allow for different analyses than those carried out using the SPSS macro "PROCESS v3.5 by Andrew F. Hayes." As far as future prospects are concerned, we could try to make the data more generalizable by expanding the number of subjects in the study and investigating companies in wider sectors of activity. Expanding the sample would also allow for different types of more accurate analysis, such as structural equation modeling.
Among the results obtained, the most interesting and innovative is the one that emerged from the analysis of the moderation because it is the one that can play the role of a driving and practical principle on how to act.
There are some practical implications derived by our results, first of all, it could be hypothesized a future intervention inspired by the theory of situational leadership. Referring to the theory of situational leadership (Hersey and Blanchard, 1969), we could propose to the supervisors of the company a training intervention to adapt their leadership style in relation to the employees with whom they are in contact. In fact, in situational leadership theory, the leader loses the connotation of a boss and becomes a partner. In this way, it is the leader who, together with the subordinate, understands the leadership style (or partnership) that is most suitable for that specific employee at that specific time or for that specific task. In particular, it will be important to understand the level of identification of the employees, since the supervisors of this firm will have to use a supportive style necessarily when dealing with individuals or groups of employees with little or no identification. It can be Psychosocial variables hypothesized that more identified employees may prefer a directive or delegating leadership style; however, they seem to be less dependent from the supervisor's support to keep a proper work engagement (which is probably already adequately fueled by their own higher degree of organizational identification). The analysis of results suggested a possible future perspective related to a training intervention aimed at "transforming" the supervisors of the company into coaches. It could be interesting to investigate the presence of coaches in the organization and implement a training path for supervisors; becoming a coach means evaluating his or her own ability to listen to and to understand the point of view of others, with the intention of putting himself or herself at the service of the team, trying to grasp what he or she can do for the group of subordinates rather than the other way around; see the servant leadership concept too (Stone et al., 2004).
The coaching-oriented supervisor aims to transmit to the subordinates the values of trust, autonomy and self-determination, not through imposition (top-down) but through shared discussion.
For the organization, transforming its supervisors into coaches can have numerous advantages such as an improvement in the organizational climate and well-being and make the information flows more fluid.
The coaching process is fundamental in enhancing managerial skills to manage change. Moreover, a transaction of supervisors in coaches is consistent with the results of the analysis of both mediations and moderation, which provide us with a fairly clear overview of how a supportive leadership style can improve organizational performance and well-being.