The Moderating Roles of Remote, Hybrid, and Onsite Working on the Relationship between Work Engagement and Organizational Identification during the COVID-19 Pandemic

Abstract

Flexible working practices have become commonplace due to the emergence of the turbulent environment that emerged during the COVID-19 pandemic, which forced organizations to change their business models, structures, processes, and policies. In this context, organizations have started to reconfigure work in terms of flexible working practices that enable them to use the full potential of their employees and to provide the conditions for well-being at work and, as a result, competitive sustainability. This study aimed to explore the relationship between dimensions of work engagement, namely, vigor at work, dedication to work, and absorption in work, and organizational identification under the moderating roles of different working practices, namely, remote, hybrid, and onsite working, during the COVID-19 pandemic. In this cross-sectional study, 200 randomly chosen employees from the public insurance industry in Turkey formed the research sample. The results indicate that each dimension of work engagement, namely, vigor at work, dedication to work, and absorption in work, is positively associated with organizational identification. In addition, when the moderation effects of different working practices on this relationship were analyzed, it was apparent that the relationship between an employee’s absorption in their work and organizational identification was weaker in those working onsite, stronger in those working in a hybrid context, and strongest in those working remotely. Therefore, we suggest that work redesign towards remote working practices enhanced positive psychological and behavioral changes in employees, i.e., well-being at work, resulting in a strengthened relationship between absorption in work and organizational identification during the COVID-19 pandemic.

1. Introduction

The sustainability of organizational activities is closely related to the extent to which the organization produces effective, efficient, and profitable products and services. Indeed, the production of quality products and/or services at the optimum level in every sense and the launching of these products and/or services to the market under highly competitive environmental conditions depend on many organizational factors, such as organizational resources, especially human resources, the organizational structure, procedures, climate, processes, goals, objectives, strategies, and policies. In order to sustain its competitiveness, the organization should manage all these factors in harmony so as to adapt to rapidly changing environmental conditions. In particular, achieving the goals and objectives of the organization depends on the attitudes and behaviors of the employees toward their work. In this respect, it is crucial for the organization to succeed in retaining its qualified employees and in ensuring that working conditions allow for qualified employees’ engagement with their work and organizational identification [1]. Herein lies a challenging issue concerning the ability of the organization to enhance the development of organizational performance [2] and to assure the existence of an environment that is essential for its employees to engage themselves in their work as well as to identify with their organization [3]. In this regard, an organization with employees who are vigorous, dedicated, and absorbed while doing his or her work, namely, those who are engaged in their work as a permanent, cognitive state rather than one that is momentary [4] and who identify themselves with their organization, is an important cornerstone in achieving the goals and objectives related to company performance and thus its sustainability.

However, the pandemic has forced organizations to reconfigure work, so flexible working practices have become an alternative model for all organizations on a global scale. Remote working, in particular, has started to be applied as a common model in order to prevent the spread of the disease [5]. In this context, understanding how to enhance well-being at work through reconfiguring work not only promotes higher productivity but also offers sustainable growth to the organizations that realize this. In this respect, it is necessary to understand which working practices, namely, remote, hybrid, or onsite working, enhance employees’ well-being at work in terms of their work engagement, namely, vigor, dedication, and absorption, and their organizational identification in order to gain higher productivity as well as sustainable growth. In numerous studies, work engagement dimensions and organizational identification are researched either together or separately, but no study explaining when and in which working practices the relationship between dimensions of work engagement and organizational identification can be strengthened can be found. To fill this gap in the literature, this study aimed to reveal when and in which direction the relationship between dimensions of work engagement and organizational identification will emerge and the changes in remote, hybrid, and onsite working under the current COVID-19 pandemic conditions. Thus, this study contributes to the related literature both theoretically and in practice in the context of shedding light on reconfiguring work in terms of remote working practices that enhanced well-being at work and that resulted in a strengthened relationship between absorption in work and organizational identification during the COVID-19 pandemic.

2. Theoretical Framework and Hypothesis Development

2.1. Work Engagement and Its Dimensions

Work engagement is a prominent concept in industrial psychology and management studies; its theoretical background and relationship with other related concepts have been investigated by many academics and researchers. In the relevant literature, the term work engagement is used interchangeably with employee engagement in some studies; nevertheless, it is emphasized by many authors that they are different concepts in terms of their focus and scope [6,7]. Employee engagement, first conceptualized by Kahn, refers to channeling all of one’s energy, physically, emotionally, and cognitively, to labors [8]. As many authors have noted about the difference between these two concepts, employee engagement is a broader concept comprising feelings about one’s organization and work [9], whereas work engagement is a specific concept related to feelings about one’s work [4,10]. Since the aim of this study is to investigate the psychological context of work that encourages individuals to put effort into their work [7], the concept of work engagement is used in this study.

Work engagement is viewed by most authors as a work-related positive mind-set leading to positive work-related outcomes [11]. Schaufeli et al. define work engagement as “a positive, fulfilling, work-related state of mind that is characterized by vigor, dedication, and absorption” [10] (p. 74). In this definition, three dimensions of work engagement are highlighted: its behavioral dimension is expressed by vigor, its emotional dimension is expressed as dedication, and its cognitive dimension is expressed as absorption [12]. Herein, vigor refers to high levels of energy and mental resilience during work, the eagerness to put effort into one’s work tasks, and perseverance when facing difficulties [10]. Dedication indicates a strong involvement in one’s work through a sense of meaning, enthusiasm, pride, and challenge [10,13]. Absorption refers to being fully concentrated and deeply immersed in one’s work with pleasure, whereby time flies, thus causing difficulties in detaching oneself from work [10,14]. In addition, some authors argue that work engagement is theoretically grounded in burnout research as a positive antithesis/antipode of burnout [9,15]. However, there is still a debate that the dimensions of burnout are the opposite of work engagement dimensions. Herein, Maslach and Leiter [16] and Schaufeli et al. [10] emphasize that only two dimensions can be assessed as opposites: exhaustion vs. vigor and cynicism vs. dedication. They argue that the other burnout dimension, inefficacy, cannot be considered the opposite of absorption in work.

In the literature, many predictors of work engagement at the individual and organizational level are emphasized. Personality traits, emotional intelligence, psychological capital, and cognitive status are identified as individual level predictors of work engagement, enabling employees to fully reflect his/her potential [17]. Organizational support, job characteristics, organizational climate, and organizational justice are perceived to be organizational level predictors [14,18,19]. Thus, the presence of perceived organizational support, positive organizational climate, perceived organizational justice, job resources such as performance feedback, and autonomy leads to work engagement [12,17,18,19,20]. On the other hand, work engagement is associated with a number of individual and organizational outcomes. At the individual level, work engagement is related to employee’s mental and physical health and their well-being at work. Studies show that employees who have high work engagement experience positive emotions and achieve good mental and physical health [21,22,23]. Work engagement is also related to organizational level outcomes such as organizational commitment, job satisfaction, intention to leave, organizational citizenship behavior, and task performance [24]. Studies prove that employees who are highly engaged in their work feel more committed to the organization [25,26], derive more satisfaction from their job [27,28,29], have a low level of intention to leave the organization [28,30], have a high level of organizational citizenship behavior [28,31,32], and demonstrate better task performance [33,34,35]. However, research indicates the existence of some negative consequences of work engagement, namely, a very high level of absorption in work without enjoyment and pleasure, which may lead to workaholism [36], thus possibly leading to mental and psychosomatic health complaints and work–family conflict [37,38].

In this context, employees with higher levels of work engagement in terms of vigor, dedication and absorption are likely to be more proactive and endeavor to enhance an organization’s productivity, overall performance, and its competitiveness [39,40,41]. These positive outcomes indicate that work engagement is a crucial factor in the sustainable management of organizations [18] as well as individual well-being. Consequently, examining dimensions of work engagement to identify different working practices that enhance their levels may have managerial and practical implications for both employees and organizations, especially in the public sector.

2.2. Organizational Identification

Organizational identification has long been viewed as an important aspect of organizational life owing to a number of benefits for the sustainability of organizations. Organizational identification is grounded in the “Social Identity Theory” [42], which outlines a specific form of social identification. Within this theory, social identification denotes an individual’s sense of belonging to certain groups, while organizational identification refers to defining himself/herself in terms of his/her membership in a particular organization, that is, an individual’s psychological bond with or belonging to the organization [43,44]. Ashforth and Mael [45] define organizational identification as “perception of oneness with or belongingness to an organization” (p. 21). Thus, organizational identification implies establishing a desired connection with the organization to which an employee belongs as a result of the harmonization of his/her personal goals with the organization’s goals [45], giving behavioral support to his/her organization in such a way as to share success or failure through a sense of unity and solidarity and integrating with the organization by seeing himself/herself as a part of that organization [46]. Van Dick et al. [47] consider identification in organizational contexts through its cognitive, affective, evaluative, and behavioral dimensions, similar to social identity theory. The cognitive dimension signifies the employee’s awareness of belonging to a certain group, the affective dimension is the emotional attachment to that group, the evaluative dimension describes the value attributed to that group, and the behavioral dimension is the individual behaving in the way that is expected of him/her [48]. In this context, once the employee feels that he/she is bound up with a group cognitively (awareness of belonging to the organization; internalizing the values of the organization) or emotionally (a sense of pride in membership), his/her sense of organizational identification rises [49,50].

In the literature, it is seen that a variety of predictors and outcomes are linked to organizational identification both at the individual and organizational level. Factors such as job tenure, a need for affiliation, and collectivism are stated as individual factors for strengthening organizational identification [51]. In addition, total rewards perceptions [52], organizational trust [53], and perceived organizational support [51] are found to be predictors of organizational identification. Thus, higher levels of perceived total rewards, organizational trust, and organizational support lead to a higher level of organizational identification. Moreover, research also shows that work engagement is associated with organizational identification both as an antecedent and as an outcome, due to their reciprocal relationships [27,52,54]. On the other hand, research indicates that organizational identification leads to positive outcomes such as higher levels of job satisfaction [49,51], job performance [55,56], job involvement [49], and a lower level of intention to leave [57,58]. Hence, the contribution of organizational identification to these positive outcomes specifies its importance in organizations achieving their goals regarding sustainable competitiveness. Thereby, studying organizational identification to determine which working practices strengthen its link with dimensions of work engagement may have practical implications for both employees and organizations, especially in the public sector.

2.3. Working Practices

During the pandemic, it has become a necessity for organizations to develop strategies by adopting information technologies under time pressure, to redesign their systems and business models, to switch to online services and e-commerce, to enter new business areas due to declining markets or outdated businesses, and to create alternative working practices as a result of lockdowns, stay-at-home orders, and social distancing measures [59]. Hence, organizations took faster steps to meet the requirements of the digital age in order to survive and maintain their competitiveness in unexpected situations [60], which resulted in flexible working practices becoming widespread.

Flexible working, which requires an intensive use of information and communication technologies (ICTs), refers to the freedom or right to determine where and when the employees work or not [61]. In this study, we discuss the most common working practices, namely, onsite working, remote working, and hybrid working. Onsite working, as a traditional way of working, refers to the employee’s performed work in the workplace assigned by the organization in accordance with the determined working hours. Remote working, as a flexible and independent form of working, is defined as the employee’s carrying out of his/her work from home or from a place outside the workplace via technological communication tools, within the scope of the work organization. Hybrid working is an integrated working practice that is composed of traditional and remote working practices, partly—for 2–3 days—onsite and partly—the other days—working from home or somewhere else interchangeably.

Research shows that each working practice has different advantages and disadvantages. Fully onsite working does not provide any flexibility, job control, or autonomy in the production of products or services for employees spatially. Nevertheless, fully implementing onsite working for some occupational groups emerges as a necessity rather than a preference [62]. Thus, it is inevitable that production-oriented work, which is carried out physically and by the workforce, is performed at the workplace [63]. Remote working is viewed as beneficial due to its many advantages, such as cost advantages in workplace and transportation, the flexible production of goods and services, employment of the best employees without time–space limits, reduction in absenteeism and in work time losses, and an increase in job satisfaction and in motivation of the employees [64]. However, the fact that remote working, which was partially and optionally applied before the pandemic, became mandatory for most organizations during the COVID-19 period, necessitates the redesign of many jobs in order to maintain its positive effects [65]. Remote working offers advantages also for the employee, e.g., by enabling him/her to control time [66], increasing organizational commitment and job satisfaction by reducing stress [67,68], achieving autonomy, creating a work–life balance and providing work–life satisfaction, saving time spent in transportation from work to home, reducing costs on energy and expenses [69], and positive effects on physical and mental health [70]. For housewives or househusbands, remote working, when carried out at home, establishes an advantage in terms of fulfilling housework and family responsibilities [68]. On the other hand, hybrid working provides the flexibility to benefit from the advantages of both onsite and remote working, and ensures employee satisfaction when it occurs in line with the goals and interests of the employee and the organization. However, hybrid working may lead to dissatisfaction for some employees, as it partially includes, for employees who prefer to fully work remotely or onsite, the option that they do not prefer.

Besides various advantages, there are some disadvantages of both remote and hybrid working. When considered from the point of view of the organization, the disadvantages of fully remote or hybrid working might be listed as causing an increase in recruitment and training expenses, causing controlling, safety, and health problems, and weakening the belongingness of the employees, whereas from the point of view of the employees, the disadvantages of fully remote or hybrid working might be stated as losing the corporate culture, loneliness, decreased social skills and limited learning and promotion opportunities, conflict between work and family roles, increased work intensity, and inconvenient job safety and working conditions as a result of social isolation [61,64,66,68].

Although each working practice has different pros and cons, it is important to reveal which working practices strengthen employees’ work engagement in terms of vigor, dedication, and absorption and their organizational identification in order to effectively reconfigure work as well as ensure well-being at work. Therefore, exploring working practices in relation to this issue may offer practical implications for both managers and employees.

2.4. Work Engagement Dimensions and Organizational Identification

Work engagement and organizational identification are interrelated concepts in organizational life. Employees who are passionate about their work will be strongly integrated with their work, will feel meaningful and important, will try to fulfill their role in the organization with a high-energy spirit, will strive in their work with great enthusiasm, and will focus completely on their work owing to the positive feelings they have for their organization and work [1]. Integration with the organization enables employees to consider their work meaningful and establishes the necessary psychological base for their identification with their work. If the work represents the reason for the existence of the organization with which the employee is integrated, the employee will identify with his/her organization as he/she absorbs his/her work, will internalize the values of the organization by absorbing it, and ultimately will be connected to his/her organization, since the employee considers his/her work as a part of the organization which he/she identifies with, and the work will thus be performed in the best way [71]. In the literature, numerous studies [28,54,72,73] examining the relationship between work engagement and organizational identification point to a positive link between work engagement and organizational identification.

Furthermore, Chughtai [74] and Freeney and Tiernan [75] recommend that each dimension of work engagement should be examined separately to more accurately detect the strengths and weaknesses in terms of each dimension of work engagement and to reveal which dimension comes into prominence and is comparatively important. Following these suggestions, the relationships between each dimension of work engagement and organizational identification are examined separately within this study, using a three-factor model of work engagement.

Employees who have high levels of energy and mental resilience at work also have positive emotional reactions to their work. These positive attitudes towards work also shape their attitudes towards their organization, so employees who have vigor at work will also identify themselves with their organization through having these positive attitudes, namely, through emotionally attaching to their organization as well as through accomplishing their work tasks, namely, in behavioral terms. In this context, Balcı and Ağ [76], in their research involving 138 private hospital employees, found a positive relationship between employees’ vigor and their identification with their organizations. Nevertheless, Aktaş and Akdemir [54] concluded that there is no relationship between the vigor of employees and their organizational identification. On the other hand, employees who experience burnout due to job dissatisfaction and generally unfavorable environmental conditions such as economic, health, and family problems feel psychologically and physiologically under pressure, and their mental resistance and vigor to make an effort decrease. Burnout experienced by the employee decreases his/her success as a result of the decrease in his/her vigor and emotionality about work, and thus emerges as a factor that reduces his/her organizational identification and commitment [77]. The emotional reactions that occur as a result of the employee’s feelings about his/her work, and the level of well-being and satisfaction with his/her job also have effects on the level of job stress [78], thus determining the level of vigor by affecting his/her physical reactions, that is, his/her attitudes and behaviors. Job requirements and job resources turn into stressors when they force the employees to exert too much energy and when they increase the negative attitudes of employees towards their work [79]. As a result, a positive or negative attitude towards work shapes an employee’s attitude towards the organization accordingly. Thus, considering the predominant opinion that there is a positive relationship between an employee’s work engagement and his/her identification with his/her organization, it is expected that the vigor of the employees is positively associated with their organizational identification. These arguments lead us to propose the following hypothesis:

Hypothesis 1.

Vigor at work is positively related to organizational identification.

An employee’s dedication to work with a high sensitivity, as well as his/her emotional integration with his/her work, allows him/her to see himself/herself as a part of the organization. Furthermore, employees who are highly dedicated to their work also adapt to their organizations by emotionally attaching to that organization with a sense of pride in being a member of that organization, so their identification with the organization and their commitment to the organization increases [80]. Supporting this view, Aktaş and Akdemir [54] and Balcı and Ağ [76] also found a positive relationship between dedication and identification with the organization in their studies. Hence, it is assumed that an employee’s dedication to his/her work contributes to his/her identification with the organization. Hence, we formulate the following hypothesis:

Hypothesis 2.

Dedication to work is positively related to organizational identification.

Employees who are fully concentrated and deeply immersed in their work with pleasure fulfill the requirements of their job and work towards organizational goals. Thus, they internalize the values of the organization by prioritizing the organizational goals over their personal interests. As a result, employees who absorb their work become identified with their organizations. Otherwise, if employees find their work meaningless and are not happily engrossed in their work, i.e., they are not absorbed in their work, they become alienated from their organization. In the study by Aktaş and Akdemir [54], it was determined that there is a positive relationship between employees’ absorption in their work through fully concentrating on it and their identification with their organizations. Additionally, Balcı and Ağ [76] found a positive link between absorption and organizational identification in their study. Similarly, Riketta [49], in his meta-analysis research, revealed that there is a positive relationship between employees’ absorption in their work and their integration with their organizations. Thus, it is expected that absorption in work is positively associated with organizational identification. In light of this discussion, we formulate the following hypothesis:

Hypothesis 3.

Absorption in work is positively related to organizational identification.

2.5. Work Engagement Dimensions, Organizational Identification, and Working Practices

Working practices could have effects on employees’ level of vigor and their identification with their work. Flexible working practices provide various conveniences allowing for the employees to meet their personal needs first and foremost. Working in an environment where employees feel under stress affects their integration with the organization by reducing their motivation towards work [81]. In this context, flexible working practices are viewed as a factor that reduces the perceptions of employees that they are working under pressure. Scandura and Lankau [82], in their study on 160 managers, revealed that employees’ evaluations of their organizations in flexible working conditions and their attitudes toward work increase their job satisfaction and person–organization fit. Owing, in particular, to flexible working practices, employees can establish a better work–life balance and channel the energy they would have spent on commuting to work. In addition, by means of flexible working practices, employees can become more vigorous by reducing their stress levels due to the loss of surveillance from managers, which may put pressure on them to complete work; thus, they have positive feelings and emotions toward their organizations for providing this opportunity [83]. In this manner, it is expected that flexible working practices have a positive effect on the identification of employees with their organizations by making them more vigorous. This effect is expected to occur more in remote working contexts, which provides the most flexibility to the employee compared to onsite and hybrid working contexts, and it is expected to occur more in hybrid working contexts than in onsite working contexts. Hence, we offer the following hypothesis:

Hypothesis 4.

Working practices moderate the positive relationship between vigor at work and organizational identification such that (a) the positive relationship between vigor at work and organizational identification is stronger in remote working contexts than in onsite contexts, in turn, (b) this positive relationship is stronger in hybrid working contexts than in onsite working contexts.

Flexible working practices could increase employees’ dedication to their work through providing an environment that enables them to be strongly involved in their work with a sense of enthusiasm and inspiration [67] and thus could affect their level of organizational identification. Wiesenfeld et al. [84], in their study on 250 employees working in the sales division of a large technology organization, found that flexible working practices that are attractive for employees enhance their identification with their organizations. Similarly, Taborosi et al. [70], in their study on 265 employees working in various organizations in Serbia, Bosnia and Herzegovina, Montenegro, and Croatia found that employees’ identification with their organizations was significantly higher in those working remotely than in those working onsite. In addition to this, as highlighted by Halgin et al. [85], employees who were engaged in terms of having high levels of connectivity with their work achieved strong ties to coworkers, thus building social relationships in distributed, flexible work settings, which, in turn, may potentially help them to establish a psychological bond with their organization. Moreover, Zafari et al. [86], in their study on 481 flexible employees from different sectors in both Austria and Spain, found that flexible working practices which serve autonomy for employees lead them to have more willingness to devote to their work and high identification with their organization. These considerations lead us to suggest the following hypotheses:

Hypothesis 5.

Working practices moderate the positive relationship between dedication to work and organizational identification such that (a) the positive relationship between dedication to work and organizational identification is stronger in remote working contexts than in onsite contexts, in turn, (b) this positive relationship is stronger in hybrid working contexts than in onsite working contexts.

The working conditions and the policies of an organization affect the attitudes of its employees towards work; that is, the policies of organizations that aim to meet employee needs and expectations ensure that the employee feels valued by the organization, is happily engrossed in his/her work, and thus is absorbed in his/her work with great enthusiasm. Moreover, the fact that the employee does his/her work with pleasure enables him/her to see himself/herself as a part of an organization that he/she cares about and thus he/she adapts to that organization. In particular, the work environment and the physical conditions of the workplace have effects on the employee’s full concentration and his/her ability to immerse himself/herself in his/her work with pleasure and are essential to the employee performing his/her work willingly and remaining motivated and concentrated [69]. As a result of 31 meta-analyses, Baltes et al. [66] found that flexible working practices have positive effects on job satisfaction and employee satisfaction and reduce absenteeism, which is an indicator that the employee is deprived of work engagement and does not feel united with the organization. Giving more freedom and autonomy to employees through flexible working practices has effects on their perspectives on work and increases their absorption in that work [87] and their identification with their organization [86]. In this context, offering an employee flexible working conditions with the option of choice is a factor that increases the well-being of that employee. The employee focuses more, concentrates fully on his/her work, and becomes absorbs in his/her work, to the extent that he/she has the freedom to control work-related time and determine which flexible working practices can provide a work–life balance. Thus, the employee who is more absorbed in his/her work through flexible working contexts internalizes the values of the organization and identifies himself/herself with the organization at a higher level. On the basis of the above discussion, we offer the following hypotheses:

Hypothesis 6.

Working practices moderate the positive relationship between absorption in work and organizational identification such that (a) the positive relationship between absorption in work and organizational identification is stronger in remote working contexts than in onsite contexts, in turn, (b) this positive relationship is stronger in hybrid working contexts than in onsite working contexts.

2.6. Theoretical Model

The theoretical model we propose in line with the discussions and hypotheses above is depicted in Figure 1.

The model informing our research posits that work engagement, with its three dimensions, namely, vigor, dedication, and absorption, is positively associated with organizational identification and that this relationship is moderated by working practices, namely, onsite, remote, and hybrid working.

2.7. Aims and Significance of the Study

The aim of this study was to reveal the relationship between employees’ work engagement, namely, vigor, dedication, and absorption, and their identification with their organizations, taking into account the moderation effects of both onsite working as a traditional working practice and flexible working practices, namely, remote and hybrid working, which have become prevalent during the pandemic and are thought to become permanent after the pandemic. In other words, in our study, we aimed to determine when and in which direction the relationship between dimensions of work engagement and organizational identification will emerge and change in remote, hybrid, and onsite working practices. In this context, our research questions are as follows: Do working practices moderate the relationship between dimensions of work engagement and organizational identification? If so, in which working practice will this relationship be strongest, and in which will it be weakest?

In the literature, we have not encountered a study directly examining the moderating roles of working practices on the relationship between dimensions of work engagement and organizational identification. In this context, we expect that our study will fill the gap in the literature and contribute to it by examining the moderation effects of working practices on the relationship between dimensions of work engagement and organizational identification.

3. Materials and Methods

3.1. Samples and Procedures

This cross-sectional study was carried out from October 2021 to the end of March 2022 and involved voluntarily participating employees in the headquarters and regional offices of companies in the public insurance sector in Turkey. We reached 500 employees chosen by a simple random sampling method via e-mail and/or phone and requested their participation in our online survey that we created via Google Forms. We included the valid answers received from the remaining 200 employees after omitting the questionnaires with incomplete answers. Thus, our sample consisted of 200 employees. We analyzed demographic questions according to frequency, as follows: 61% of participants were female; 75% of them were 0–40 years old; 65% of them had a bachelor’s degree; 67% of them were married; 47.5% of them had 0–5 years tenure in the current workplace; 28% of them had over 10–15 years of experience in the sector; 73.5% of them worked over 39 h a week; 17% of them worked remotely; 51.5% of them worked in a hybrid context; and 31.5% of them work onsite.

3.2. Measures

In this research, we used two different scales, and their details are as follows:

Work Engagement Scale: To measure dimensions of work engagement, the shortened, 9-item version (UWES-9) [4] of the 17-item Utrecht Work Engagement Scale (UWES) developed by Schaufeli et al. [10] was used. Translated into Turkish and validated by Özkalp and Meydan [72], the UWES-9 includes three subscales reflecting dimensions of vigor with three items (e.g., “At my job, I feel strong and vigorous”), dedication with three items (e.g., “I am enthusiastic about my job”), and absorption with three items (e.g., “I am immersed in my work”). In the analysis, this three-factor model was used.

Organizational Identification Scale: The Organizational Identification Scale developed by Mael and Ashforth [43] was used to measure organizational identification. Adapted to Turkish and validated by Tak and Aydemir [88], this scale has 6 items (e.g., “When I talk about my organization, I usually say ‘we’ rather than ‘they’”) in one dimension.

Participants were asked to rate each of the items using a 5-point Likert scale (1 = strongly disagree; 5 = strongly agree).

Working Practices: To measure working practices, namely, remote, hybrid and onsite working, we asked employees to mark whether they work at the workplace assigned by the organization in accordance with the determined working hours, i.e., onsite working, whether they carry out their work from home or from a place outside the organization via technological communication tools, i.e., remote working, and whether they partly worked onsite and partly worked remotely, i.e., hybrid working. ‘Working practices’, a categorical variable with three categories (1 = remote working, 2 = hybrid working, and 3 = onsite working), was transformed into two variables each with two categories. In other words, this variable was dummy coded into two variables, one called remote working and one hybrid working: if working practices = remote working, then the column remote working was coded with a 1 and hybrid and onsite working with a 0; if working practices = hybrid working, then the column hybrid working was coded with a 1 and remote and onsite working with a 0; and if working practices = onsite working, then the column onsite working was coded with a 1 and remote and hybrid working with a 0. This dummy coding was automatically performed by SPSS Statistics. In this study, we used onsite working as the reference category and it was not included as a dummy indicator variable for the working practices moderator construct. Hence, the dummy-coded variables of remote and hybrid working were included in the analyses.

3.3. Data Analysis

Partial least squares structural equation modeling (PLS-SEM), a second-generation multivariate statistical method, was used, and analyses were performed via the SmartPLS 4 statistical package software in this study. Non-parametric resampling (bootstrap) was used to calculate the estimated standard error values, t-statistics, and confidence intervals. All models in this study were obtained with 10,000 resamples, as recommended [89], and the significance level in the analyses was accepted as 0.05.

Since data were collected from a single respondent, it is important to check the common method bias. Although a traditional approach is to use Harman’s single factor test [90] to check for common method bias–in this study, Harman’s test results showed that no single factor emerged in the unrotated solution (68.92% of the total variance) and the first factor did not explain most of the variance (31.45%)–many researchers, such as Gaskin [91], state that Harman’s single factor test is no longer widely accepted and is considered an outdated and inferior approach. Therefore, a common method bias test in SmartPLS 4 was performed using the VIF collinearity approach with a random dependent variable, namely, the full collinearity test developed by Kock and Lynn [92], which is strongly recommended by researchers. According to this approach, first, a new variable with random values was created. Second, a new model where all of the latent variables point to a construct with this new random variable was created. Third, the PLS procedure was run, and the inner model VIFs were checked. The inner model VIFs, i.e., remote working = 1.392, hybrid working = 1.288, vigor = 2.391, dedication = 2.420, absorption = 1.383, organizational identification = 1.087, were not above the threshold of 3.3; therefore, it was concluded that collinearity, i.e., common method bias, was not present in the model. Moreover, Siemsen et al. state that, if the interaction terms are investigated and significant interaction effects are established, a study should not be criticized for common method bias, because interaction effects diminish or deflate as a result [93]. Since the main purpose of our study was to reveal the moderation effects of working practices, taking the common method bias test using the VIF collinearity approach with a random dependent variable results, common method bias is not likely to be a concern in this study.

Following the suggestions of Hair et al. [94], this study used a two-step process for PLS-SEM, namely assessment of the measurement model and the structural model. In this study, assessment of the measurement model included evaluation of overall fit of the saturated model examining the values of the discrepancy measures, and evaluation of the reflective measurement model examining indicator reliability, internal consistency reliability, i.e., Cronbach’s alpha, reliability coefficient ρA, composite reliability ρC, and validity, i.e., content, convergent, and discriminant. Nevertheless, assessment of the structural model comprises evaluation of the overall fit of the estimated model examining the values of the discrepancy measures, evaluation of the path coefficients and their significance levels, hypotheses testing, the model’s explanatory power, i.e., the coefficient of determination (R²), and effect size ($f^2$), and the model’s predictive power ($Q_{\mathrm{predict}}^2$).

Furthermore, in order to detect the moderating effects of working practices on the whole model, a multigroup analysis (MGA) using SmartPLS 4 was also performed. Once the groups (remote, hybrid, and onsite working practices) were generated, measurement invariance was tested following the measurement invariance of composite models (MICOM) procedure, which was developed by Henseler et al. [95] and is strongly recommended to be performed before testing multigroup comparisons (PLS-MGA) in order to prevent the misleading group comparisons conducted by many researchers [89,96]. The three steps of the MICOM procedure comprise determining configural invariance, compositional invariance, and the equality of composite mean values and variances [95]. In this study, configural and compositional invariance were established and the composites had equal mean values and variances across the groups of working practices, thus full measurement invariance was established. Henseler et al. [95] recommend that if full measurement invariance is established, the model should be analyzed using the pooled data, namely, a moderation analysis should be run by including the interaction effects that serve as moderators representing the grouping variable, which can lead to an increase in the statistical power and generalizability of the model. Since full measurement invariance was established in this study, it was concluded that rendering PLS-MGA was unnecessary and running a moderation analysis using the pooled data could be advantageous.

4. Results

4.1. Assessment of the Measurement Model

This research used a reflective approach for theoretical model constructs, i.e., dimensions of work engagement and organizational identification. Following the recommendations of Benitez et al. [97], the assessment of the reflective measurement model begun with the evaluation of the overall fit of the saturated model, that is, with confirmatory factor analysis which is useful to test the adequacy/validity of a reflective measurement model. The values of the discrepancy measures and 95% quantiles of their corresponding reference distribution for our model are given in

Table 1. Results of the confirmatory factor analysis.

Discrepancy	Overall Saturated Model Fit Evaluation
	Value	HI95	Conclusion
SRMR	0.029	0.038	Supported
dULS	0.252	0.521	Supported
dG	0.106	0.214	Supported

Note. This table was adapted from Benitez et al. [97] and generated on the basis of their recommendation.

.

SRMR value was below the recommended threshold value of 0.080 [98], and all discrepancy measures were below the 95% quantile of their reference distribution (HI₉₅) [99], thus indicating that the model has a good fit [100].

As recommended by Hair et al. [89], it should be noted that single-item variables which have measurement values of 1.00, in this study dummy-coded moderator variables, i.e., remote and hybrid working, should not be reported with measurement model results, namely, measures of validity and reliability, otherwise the interpretation of these constructs’ validity or reliability values is not meaningful, thus these were not reported.

Next, following the suggestions of Hair et al. [94], the reflective measurement model was evaluated to determine indicator reliability, internal consistency reliability, i.e., Cronbach’s alpha, reliability coefficient ρA, and composite reliability ρC, and validity, i.e., content, convergent, and discriminant (see

Table 2. Measurement model results.

Constructs	Items	Convergent Validity			Internal Consistency Reliability			Discriminant Validity
		Loadings	Indicator Reliability	AVE	Cronbach’s Alpha	Reliability ρA	Composite Reliability ρC	HTMT
		>0.70	>0.50	>0.50	0.60–0.90	0.60–0.90	0.60–0.90	Significantly Lower Than 0.85?
Vigor	Vigor 1	0.948 ***	0.898	0.814	0.883	0.898	0.901	Yes
	Vigor 2	0.946 ***	0.894
	Vigor 3	0.804 ***	0.646
Dedication	Dedication 1	0.933 ***	0.870	0.838	0.900	0.900	0.903	Yes
	Dedication 2	0.941 ***	0.885
	Dedication 3	0.871 ***	0.758
Absorption	Absorption 1	0.635 ***	0.403	0.544	0.572	0.579	0.780	Yes
	Absorption 2	0.747 ***	0.558
	Absorption 3	0.818 ***	0.669
Organizational Identification	Org_ Identif 1	0.704 ***	0.495	0.528	0.825	0.835	0.870	Yes
	Org_ Identif 2	0.753 ***	0.567
	Org_ Identif 3	0.725 ***	0.525
	Org_ Identif 4	0.730 ***	0.532
	Org_ Identif 5	0.780 ***	0.608
	Org_ Identif 6	0.661 ***	0.436

Note. This table was adapted from Hair et al. [89] and generated on the basis of their recommendation. *** p <0.001.

).

First, item loadings, indicator reliability (squared item loading), and average variance extracted (AVE) were computed to examine the constructs’ convergent validity. Table 2 depicts the results of the measurement model assessment. Hair et al. [89] recommend that item loadings should be 0.70 or higher and indicator reliability and AVE values should be at least 0.50. One item each from absorption, i.e., Absorption 1, and organizational identification, i.e., Org_ Identif 6, had item loadings and indicator reliability below the recommended thresholds. But these items were retained on the basis of their contributions to content validity and AVE value [89,97]. Depicted in Table 2, the AVE values for all constructs were above the threshold value of 0.50, thus satisfactory convergent validity was achieved for all constructs in this study.

Second, Cronbach’s alpha, reliability coefficient ρA, and composite reliability ρC were computed to examine the constructs’ internal consistency reliability. Values of 0.60 to 0.90 are recommended for Cronbach’s alpha, reliability coefficient ρA, and composite reliability ρC [89]. Absorption had Cronbach’s alpha and reliability coefficient ρA below the recommended thresholds. However, its Cronbach’s alpha value of 0.572 and reliability coefficient ρA value of 0.579 were accepted in combination with its composite reliability ρC, i.e., 0.780, higher than 0.60 [89,97,101], and keeping in mind that internal consistency coefficients of absorption vary across countries and values between 0.55 and 0.90 have been calculated [4,72], so all items were kept to better compare the results.

Third, the heterotrait-monotrait ratio of correlations (HTMT), with a threshold of 0.85, was computed to examine the discriminant validity of the constructs [102]. Depicted in

Table 3. Discriminant validity results.

Constructs	1	2	3	4	Inner VIF
1.  Vigor	-				3.276
2.  Dedication	0.827	-			3.208
3.  Absorption	0.625	0.632	-		1.281
4.  Organizational Identification	0.615	0.611	0.695	-	-

Note. Discriminant validity is established at HTMT_0.85 [103]. VIF = variance inflation factor.

, all the correlation values among the constructs were below the recommended threshold value of 0.85, thus all the constructs fulfilled this criterion. In addition, all the independent variables, i.e., vigor, dedication, and absorption, had variance inflation factors (inner VIF) lower than the threshold value of 5.0 [89]. Hence, empirical evidence for discriminant validity was obtained.

Since the reliability, validity, and good fit of the measurement model were established, the analysis and the evaluation of the structural model was proceeded with to test the hypothesized relationships.

4.2. Assessment of the Structural Model

In this study, a bootstrapping technique with 10,000 bootstraps [103] was employed to generate path coefficients and their significance levels and for hypotheses testing, as proposed by Hair et al. [89]. The assessment of the structural model comprises the evaluation of the overall fit of the estimated model examining the values of the discrepancy measures, evaluation of path coefficients and their significance levels, hypotheses testing, and the model’s explanatory power, i.e., the coefficient of determination (R²), effect size ($f^2$), and the model’s predictive power ($Q_{\mathrm{predict}}^2$) [89,97]. Table 3 and Figure 2 present the assessment of the structural model.

Initially, to analyze the moderating roles of working practices, first, dummy-coded variables of remote and hybrid working practices were included in the analyses, keeping in mind onsite working practices as the reference category. Subsequently, following the recommendations of Hair et al. [89] and Becker et al. [104], interaction terms, i.e., “Remote Working × Vigor”, “Hybrid Working × Vigor”, “Remote Working × Dedication”, “Hybrid Working × Dedication”, “Remote Working × Absorption”, and “Hybrid Working × Absorption”, were created applying the two-stage approach with standardized data when conducting moderator analyses. It should be noted that, unlike previous versions, SmartPLS 4 automatically performs the two-stage approach for moderator analysis which estimates a main effects model, i.e., without the interaction term, in stage 1 and a simple effects model when the interaction term is calculated and is included in stage 2 [105].

In evaluating the structural model, primarily the overall fit of the estimated model through the bootstrap-based test of overall model fit and the SRMR as a measure of approximate fit was examined, as proposed by Benitez et al. [97]. The SRMR value, i.e., 0.031, was below the recommended threshold value of 0.080 [98], and all discrepancy measures were below the 95% quantile of their reference distribution (HI₉₅) [99], thus indicating that the model has a good fit [100] (see

Table 4. Structural model results.

Hyp.	Relationship	Path Coefficient (β)	t-Value	Effect Size $\left({\mathit{f}}^2\right)$	Results
H1	Vigor → Organizational Identification	0.178 ***	2.870	0.022	Supported
H2	Dedication → Organizational Identification	0.134 ***	2.447	0.020	Supported
H3	Absorption → Organizational Identification	0.282 ***	3.921	0.028	Supported
H4a	Remote Working × Vigor → Organizational Identification	0.253	0.718	0.003	Not Supported
H4b	Hybrid Working × Vigor → Organizational Identification	0.085	0.341	0.001	Not Supported
H5a	Remote Working × Dedication → Organizational Identification	−0.572	1.575	0.006	Not Supported
H5b	Hybrid Working × Dedication → Organizational Identification	0.030	0.129	0.000	Not Supported
H6a	Remote Working × Absorption → Organizational Identification	0.531 **	5.700	0.058	Supported
H6b	Hybrid Working × Absorption → Organizational Identification	0.121 *	2.034	0.017	Supported
	Endogenous Variable	Coefficient of Determination (R²)	Predictive Power $(Q_{predict}^2)$
	Organizational Identification	0.426	0.203
	Overall fit of the estimated model	Value	HI95
	SRMR	0.031	0.038
	dULS	0.274	0.533
	dG	0.109	0.215

Note. * p < 0.05, ** p < 0.01, *** p <0.001.

).

For hypotheses testing, all path coefficients and their significance levels in the model were evaluated. Hypothesis 1 to Hypothesis 3 posited that vigor at work, dedication to work, and absorption in work would be positively related to organizational identification, respectively. As shown in Figure 2 and Table 3, vigor at work, dedication to work, and absorption in work were positively related to organizational identification (for vigor at work: β = 0.178, t = 2.870, p = 0.000; for dedication to work: β = 0.134, t = 2.447, p = 0.000; for absorption in work: β = 0.282, t = 3.921, p = 0.000). Therefore, Hypothesis 1 to Hypothesis 3 were supported at p < 0.001. When the moderation effects of the working practices on the relationship between vigor at work and organizational identification were analyzed, the results show that there was no statistically significant change in the level of identification with the organization of those working remotely, hybrid, or onsite, depending on their working practices. Thus, Hypothesis 4a, which predicted that the positive relationship between vigor at work and organizational identification would be stronger in remote working contexts than in onsite contexts (β = 0.253, t = 0.718, p = 0.473), and Hypothesis 4b, which predicted that the positive relationship between vigor at work and organizational identification would be stronger in hybrid working contexts than in onsite working contexts (β = 0.085, t = 0.341, p = 0.733), were not supported. The results indicate that working practices did not moderate the positive relationship between dedication to work and organizational identification. Thus, Hypothesis 5a, which posited that the positive relationship between dedication to work and organizational identification would be stronger in remote working contexts than in onsite contexts (β = −0.572, t = 1.575, p = 0.115), and Hypothesis 5b, which posited that the positive relationship between dedication to work and organizational identification would be stronger in hybrid working contexts than in onsite working contexts (β = 0.030, t = 0.129, p = 0.898), were not supported.

The results (see Figure 2 and Table 3) show that working practices moderated the positive relationship between absorption in work and organizational identification. Hence, Hypothesis 6a, which predicted that the positive relationship between absorption in the work and organizational identification would be stronger in remote working contexts than in onsite contexts (β = 0.531, t = 5.700, p = 0.007), and Hypothesis 6b, which predicted that the positive relationship between absorption in the work and organizational identification would be stronger in hybrid working contexts than in onsite working contexts (β = 0.121, t = 2.034, p = 0.033), were supported. The slope plot for the link between absorption in work and organizational identification moderated by remote and onsite working showed that the positive relationship is stronger in remote working contexts (i.e., the slope is highly steeper) than in onsite contexts (i.e., a flatter line) as depicted in Figure 3.

Accordingly, the slope plot for the relationship between absorption in the work and organizational identification moderated by hybrid and onsite working showed that the positive relationship is stronger in hybrid working contexts (i.e., a slightly steeper line) than in onsite working contexts (i.e., the slope is flatter), as illustrated in Figure 4.

Since the structural model assessment also entails evaluating the model’s explanatory power, the most commonly used measure, the coefficient of determination (R²) value, was examined first [89]. The R² value of the model is 0.426, thus, the model explains 42.6% of the variance in the organizational identification of employees, using unexplored moderator variables for organizational identification (i.e., working practices: remote, onsite, and hybrid working). Regarding the R² value ranging from 0.315 to 0.771 in prior research on the reciprocal relationship between the dimensions of work engagement and organizational identification [106] and the originality of our moderator variables’ effects on the link between the dimensions of work engagement and organizational identification, an R² of 0.426 seems to be a good value. Second, the effect sizes ($f^2$) of the relationships between the constructs were examined to assess the relevance of the significant relationships [97]. According to Cohen’s [107] guidelines, for the direct effects, $f^2$ values of 0.02, 0.15, and 0.35 indicate a predictor construct’s small, medium, or large effect, respectively, on an endogenous construct. In our model, all dimensions of work engagement, namely, vigor at work $(f^2$= 0.022), dedication to work $(f^2$= 0.020), and absorption in the work $(f^2$= 0.028) contributed to the explanation of the organizational identification with small effect sizes (see Table 3). These results also indicate that, among all dimensions of work engagement, absorption in the work had the largest effect size, that is, absorption in the work comes into prominence compared to the others in terms of its effect size. The relevance of the moderation effects was assessed examining the $f^2$ value proposed by Kenny [108] and Hair et. al. [89]. Although for assessing the $f^2$ value Cohen’s [107] guidelines are widely used, Aguinis et al. [109] have found that the overall mean $f^2$ value in tests of moderation is only 0.009, thus Kenny [108] and Hair et. al. [89] recommend using an $f^2$ value of 0.005, 0.010, and 0.025 indicating small, medium, and large effects, respectively, for assessing the strength of the moderation effect, i.e., how much the moderation contributes to the explanation of the endogenous construct. In our model, the results of the moderation effect of remote working compared to onsite working on the link between absorption in the work and organizational identification indicate a large effect $(f^2$= 0.058) whereas the results of the moderation effect of hybrid working compared to onsite working on this relationship indicate a medium effect $(f^2$= 0.017) (see Table 3). These results also indicate that, among all working practices, remote working has the largest effect size, whereas onsite working has the smallest effect size, that is, the positive relationship between absorption in the work and organizational identification is strongest in remote working contexts and is weakest in onsite working contexts in terms of their effect sizes. Furthermore, results show that the positive relationship between absorption in the work and organizational identification is medium in hybrid working contexts in terms of its effect size.

Aside from assessment of the model’s explanatory power, in this study the model’s predictive power ($Q_{\mathrm{predict}}^2$) was also assessed using the PLSpredict procedure developed by Shmueli et al. [110] that indicates out-of-sample prediction for assessing the model’s practical relevance. To assess the model’s predictive power, Shmueli et al. [111] and Hair et. al. [89] recommend to first evaluate the $Q_{\mathrm{predict}}^2$ statistic (i.e., $Q_{\mathrm{predict}}^2$ values should be greater than zero), followed by a comparison of the RMSE (root mean square error) values produced by the PLS-SEM analysis and the LM (linear regression model) estimations. In our model, $Q_{\mathrm{predict}}^2$ values > 0, and 4 out of 6 (a majority) of organizational identification’s indicators in the PLS-SEM analysis yielded smaller prediction errors compared to the LM, thus indicating a medium predictive power [89]. Overall, our model’s predictive power: $Q_{\mathrm{predict}}^2$ = 0.203 (see Table 3). These results indicate that our model produced generalizable findings.

5. Discussion

5.1. Theoretical Implications

The present investigation attempts to examine the moderating roles of remote, hybrid, and onsite working on the relationship between dimensions of work engagement, namely, vigor at work, dedication to work, and absorption in work and organizational identification during the COVID-19 pandemic. This paper aims to explore and reveal the working practices in which the relationships between employees’ vigor at work, dedication to work, and absorption in work and their identification with their organizations occur, and in which working practice these relationships are stronger. In this vein, we conducted a cross-sectional study in the public insurance sector from the midpoint to the late period of the pandemic in Turkey.

The theoretical framework of our research model (see Figure 1) proposes that each dimension of work engagement is associated with organizational identification. Moreover, it is also proposed that the relationships between each dimension of work engagement and organizational identification are moderated by working practices, namely, remote, hybrid, and onsite working. As anticipated, the results of this study support many of the hypotheses drawn from this theoretical framework regarding the moderating roles of each working practice in the relationships between each dimension of work engagement and organizational identification. Our findings provide evidence for the importance of work redesign in terms of remote working practices, which enhances well-being at work, resulting in a strengthened relationship between absorption in work and organizational identification during the COVID-19 pandemic.

First, following the suggestions of Chughtai [74] and responding to the call of Freeney and Tiernan [75], the relationships between each dimension of work engagement and organizational identification was examined separately within this study to determine which dimension employees are more inclined toward, and to reveal which dimension is most prominent. As expected, vigor at work, dedication to work, and absorption in work were each positively associated with organizational identification. These findings support the validity of previous research findings examining the relationships between each dimension of work engagement and organizational identification [28,54,76]. When we compare our research findings with the findings of other studies in order to reveal which dimension stands out compared to the others in terms of the strength of the relationship, we found that, in studies conducted in western countries, such as a study by Karanika-Murray et al. [28] conducted in the UK, vigor at work was most associated with organizational identification; on the contrary, in our study and Aktaş and Akdemir’s study, both conducted in Turkey, absorption in work was most related to organizational identification. A possible explanation for this result may be differences in the contextual factors, such as the national culture, in which the research is conducted. In this regard, many studies [112,113,114] emphasize that, in research on organizational behavior, national culture should be considered an important factor that influences individual attitudes, perceptions, and behaviors by regulating organizational policies and practices through its value system [114]. Although this point of view can be criticized for being too general, it is suggested that research findings, especially regarding an individual’s work-related attitudes, perceptions, and behaviors, be interpreted by considering cultural value differences [112]. For example, in a study of Lee et al., it is stated that the individual–organization identity connection and its level varies according to whether it is aligned with the cultural value system adopted in a particular national context [114]. Moreover, in a meta-analysis study by Taras et al., it was found that cultural values have a strong predictive power compared to other variables and that individuals’ work-related attitudes, perceptions, and behaviors differ in the context of different cultural values [113]. Thus, the reason for the prominence of different dimensions of work engagement, in the relationship between the dimensions of work engagement and organizational identification in studies conducted in different countries, may be cultural value differences.

Second, unexpectedly, remote, hybrid, and onsite working practices do not significantly moderate the relationship between either vigor at work or dedication to work and organizational identification. It is worth noting that, without any previous findings, these findings could not be compared with related studies. The possible reason for these findings could be due to environmental conditions. As stated in recent studies showing the negative influences of the pandemic on employees’ attitudes and perceptions towards work [115,116,117], there is reason to suspect that anxieties and worries about the COVID-19 pandemic as well as the deeply felt economic crisis in Turkey may influence employees’ emotional states and attitudes toward working. Thus, in these hostile environmental conditions, employees’ levels of energy, interaction with other employees, mental resilience, and ability to cope with challenges while working may deflate [118], and they may not be fully integrated with their work. Therefore, employees’ identification with their organizations may also be low due to their low level of vigor and dedication. Hence, this situation may not be changing, with remote, hybrid, or onsite working practices, in this hostile environment.

Third, the most important contribution of this paper is that the moderation effects of working practices on the relationship between absorption in the work and organizational identification are revealed. As hypothesized, working practices were found to moderate the positive relationship between absorption in the work and organizational identification such that the positive relationship between absorption in the work and organizational identification is stronger in remote working contexts than in onsite contexts, in turn, this positive relationship is stronger in hybrid working contexts than in onsite working contexts. The results also indicate that, among all working practices, remote working has the largest effect size, whereas onsite working has the smallest effect size, that is, the positive relationship between absorption in the work and organizational identification is strongest in remote working contexts, is medium in hybrid working contexts, and is weakest in onsite working contexts in terms of effect sizes. As mentioned above, due to the lack of research findings on this subject, it was not possible to compare these findings with other studies. The fact that working practices have moderation effects in the relationship between absorption in the work alone and organizational identification requires an explanation by considering working practices in terms of the meaning of absorption in this relationship. That is to say, it is highly probable that the relationship between absorption in the work, which is characterized as concentrating completely on one’s work and immersing oneself in it with pleasure [10], and organizational identification is expected to be highest in remote working contexts and lowest in onsite working contexts. This is because waking up early in the morning to prepare to arrive at the workplace at a certain time every day, and spending time and effort to do so, spending time on transportation from work to home along with its stress and burden, and, during the COVID-19 pandemic, the anxieties and worries associated with the risk of being infected from being with other employees can be thought to reduce and hamper an onsite employee’s concentration and the likelihood that he/she can become deeply immersed in his/her work with pleasure. Indeed, many studies provide evidence that flexible working practices, especially remote working, eliminates these drawbacks and offers the employee benefits, e.g., time control, autonomy, reduced stress, an improved work–life balance, cost advantages in energy and expenses, and positive effects on physical and mental health [66,67,68,69,70]. Therefore, as a result of these benefits of remote working, the degree of an employee’s identification with his/her organization increases most as he/she becomes more absorbed in his/her work in remote working contexts. Nevertheless, it can be argued that mid-moderation effects of hybrid working occur between absorption in the work and organizational identification due to it comprising some of the disadvantages of onsite working.

5.2. Practical Implications

The findings of this study offer some practical implications both for individuals and organizations. First, managers should recognize the importance of reconfiguring work in the face of the current turbulent environment. For public sector managers, this study identifies the importance of remote working to enhance an employee’s absorption in his/her work as well as his/her identification with his/her organization. There is a need for human resources (HR) managers in the public sector to redesign jobs in terms of remote working, instead of working in a completely traditional way, i.e., onsite working, to enhance an employee’s concentration on his/her work, his/her ability to immerse himself/herself with pleasure, and his/her identification of himself/herself with his/her organization, these will lead to well-being at work, employee retention, decreases in absenteeism, cost savings in energy and expenses, higher productivity, and the sustainable growth of the organization [66]. This study also highlights another important issue that managers need to be aware of in reconfiguring work in terms of hybrid working practices, which is another form of flexible working. That is, when remote, hybrid, and onsite working practices are assessed together, in the context of this study, although hybrid working has some of the disadvantages of onsite working, an employee’s concentration, ability to immerse himself/herself with pleasure, and identification with his/her organization are stronger in hybrid working contexts than in onsite working contexts, albeit much lower than in remote working contexts. When compared with remote working, hybrid working has a lower effect on the employee’s absorption in his/her work and his/her identification with the organization. In this respect, managers should decide, in mutual agreement with the employee, on a way of working that will pave the way for their employees’ well-being and productivity.

HR professionals also need to redesign the work offered by their organization, especially by creating mechanisms that ensure that remote employees are fully concentrated on their work and identified with their organization. Within the scope of this study, although the employee’s absorption in their work has come to the forefront among the dimensions of work engagement, it is suggested that training that encourages remote employees to be energetic at work and to be dedicated to work be provided and that the job be designed by integrating practices such as management by objectives (MBO), which will encourage employees and increase the degree of their identification with the organization. In a recent study by Costantini and Weintraub, the first of its kind in the field, it is emphasized that, for work engagement to prevail, “managers should encourage remote workers to set goals for themselves and at least some of these goals should be related to social expansion” [119] (p. 11). In this respect, remote employees should also set goals for themselves and be involved in MBO programs, taking into account their contribution to work engagement, i.e., vigor at work, dedication to work, and absorption in their work.

Apart from this, it is suggested that employees pay attention to one more aspect in their identification with the organization by absorption in their work. When employees, especially remote workers, are fully immersed in their work and find it difficult to detach from it, they should be careful not to cross the line of work being fun for them and move into a phase of workaholism that constantly drives them to work day and night. This is because workaholism does not lead to positive work-related outcomes such as employee well-being and identification with the organization as in the case of absorption in the work; on the contrary, it causes negative outcomes such as work stress, social dysfunction, burnout, and work–family conflict [36]. In this respect, it is recommended that this boundary be adjusted by the employees in a way that leads to their well-being. On the other hand, as stated by Spagnoli et al. [120], it is suggested that managers monitor the risk of workaholism, especially in remote employees, with supportive and empowering leadership—in Di Fabio’s words, “decent leadership” [121]—and make the necessary arrangements.

5.3. Limitations and Directions for Future Research

This study provides insightful contributions to the current literature both theoretically and practically, but these contributions should be considered in light of the limitations of the study. Primarily, the methodological limitations of this study, such as the self-reported data, sample size, and cross-sectional research design, should be taken into account. First, data were collected from a single respondent during the pandemic, as in many studies, so a risk of response bias may be present. However, common method bias is not a major concern in this study in light of some of the precautions taken, as outlined in the Methods section. In addition, this study could have collected data from multiple respondents, e.g., managers in addition to employees; however, as argued in many studies, for individual facts, such as information about each dimension of the employee’s work engagement and his/her identification with his/her organization, the self-reported data collection method, instead of using multiple respondents, is preferred to avoid any prompting bias [27]. Second, the sample size was relatively small, but this constraint is overcome by the methodological preference of PLS as a component-based approach, with minimal requirements of sample size to achieve sufficient statistical power [89]. Furthermore, the sample is representative of the Turkish public sector population which is limited to one country, and this model produced generalizable findings due to its medium predictive power. Nonetheless, we recommend researchers to interpret the findings of our study in the context of countries that share similar cultural values and environmental factors with Turkey. Third, the use of a cross-sectional research design, which limits causality inferences, is another limitation. A longitudinal research design may be preferred in future research in order to clearly understand the causal relationships between each dimension of work engagement and organizational identification and to reveal how this relationship develops over time. A limitation of the scope of this study is that it does not cover all variables that potentially predict organizational identification. Therefore, the results are specific to our model, and another model with other predictor variables may yield different patterns.

This study presents several avenues for future research. Since our study has revealed the moderation effects of remote, hybrid, and onsite working on the relationship between only one dimension of work engagement (namely, absorption) and organizational identification, different research models that consider the moderation effects of working practices are suggested for future research. First, future research can adapt the moderation effects of job tenure in the current workplace, sector experience, and age to our research model and take into account all three dimensions of work engagement to determine which dimension will stand out in its relationship with organizational identification. Thus, by determining in which direction and to what extent different working practices, job tenure in the workplace, sector experience, and age affect this relationship, in reconfiguring work, practices that will pave the way for employee well-being at work will be determined by taking into account these demographic characteristics in employees. Second, future research may also investigate the moderation effects of remote, hybrid, and onsite working on the reciprocal relationship between other types of work-related well-being, e.g., job satisfaction [10,12,13,14,15,74] or workaholism, burnout, and organizational identification. Thus, the determination of the direction and strength of the reciprocal relationships between these variables in different working practices will also guide the reconfiguration of work. Third, another avenue for further research may be to extend our research model to the antecedents of work engagement by adopting a job demands–resources (JD-R) approach. So far, although job resources, such as social support, supervisory coaching, performance feedback, and autonomy, and personal resources, such as resilience, optimism, and self-efficacy, have been investigated as antecedents of work engagement [12] in the relevant literature, some studies have predicted each sub-dimension of work engagement with these antecedents [11,122,123]. In addition, there is no research that reveals the moderation effects of all three working practices, i.e., remote, hybrid, and onsite working, on the relationship between the antecedents and dimensions of work engagement or on the relationship between all of them and organizational identification. By adapting these antecedents to the model, such a study will help to determine which dimensions of work engagement are related to what factor in the background and to explain how all three working practices moderate their relationships with organizational identification. Additionally, the findings of such a study will also provide an insightful point of view for managers intending to redesign jobs in a way that will lead to higher levels of well-being at work as well as positive job outcomes. Fourth, further research can also investigate the consequences of each dimension of work engagement, such as job satisfaction, intention to quit, and organizational citizenship behavior [28,124], and can adapt these as mediators to our model to reveal how these relationships will change in remote, hybrid, and onsite working practices. The findings of such a study will open a new door to effective human resources practices and will reveal how work should be reconfigured in the context of obtaining positive outcomes. Furthermore, such questions will offer a richer portrait of reconfiguring work in terms of flexible working practices that enable the sustainability of organizational activities as well as the adaptation to rapidly changing environmental conditions. Fifth, future research can examine, apart from dimensions of work engagement, how antecedents of organizational identification, i.e., individual-based and/or organization-based predictors of organizational identification, such as positive and negative affectivity, individualism, perceived organizational support, the strength of organizational identity, the perceived external image of the organization, person–organization fit, the need for affiliation, and organization-based self-esteem [125,126], predict organizational identification under the moderating roles of working practices. Finally, since this research focuses on the public insurance sector in Turkey, further studies can replicate our findings at private sector companies in different countries. It is worth noting that, until the onset of this pandemic, none of the public sector organizations had ever adapted flexible working practices, namely, remote and hybrid working, in Turkey. The fact that flexible working practices became a necessity in public institutions with the pandemic triggered this research. This is because onsite working, known as the traditional way of working, is an operationalization of the long-standing classical management style of public organizations in Turkey. In this respect, since we conducted our research during the midpoint and final period of the pandemic, we recommend that future studies evaluate and compare our research findings in the context of this “new normal”, considering that employees adapted to these new flexible working practices in a period of only approximately 1.5 years.

6. Conclusions

This study explores the moderating roles of remote, hybrid, and onsite working on the relationships between each dimension of work engagement and organizational identification during the COVID-19 pandemic. Our findings support the validity of previous research findings examining the relationships between each dimension of work engagement and organizational identification. In particular, contrary to what has been found in western countries, in our study conducted in Turkey, absorption in work was most related to organizational identification. Our results provide support for a weaker relationship between absorption in work and organizational identification in those working onsite, a stronger relationship in those working in a hybrid context, and the strongest relationship in those working remotely. These findings suggest that flexible working practices, especially remote working, eliminate the possible disadvantages of onsite working, such as the loss of time on transportation from work to home, along with its stress and burden, and, during the COVID-19 pandemic, anxieties and worries about the risk of being infected from physically being with other employees, and these disadvantages reduce and hamper an onsite employee’s concentration on his/her work and prevent him/her from being deeply immersed in his/her work with pleasure. These results also imply that HR managers in the public sector need to redesign jobs in terms of remote working to enhance an employee’s concentration ability to immerse himself/herself in his/her work with pleasure and to improve his/her identification with his/her organization, which will lead to well-being at work, employee retention, decreases in absenteeism, cost savings in energy and expenses, higher productivity, and the sustainable growth of the organization. Future research should adapt the moderation effects of job tenure in the current workplace, sector experience, and age to our research model and take into account all three dimensions of work engagement to determine which dimension will stand out in its relationship with organizational identification.