Initially, COR theory seeks to provide a new research perspective in understanding “what stress is” (Hobfoll, 1989, 2011). It contends that individuals tend to conserve, protect, and acquire resources, and on this basis, both potential and actual resource losses can cause stress at the individual level (Hobfoll, 1989; Hobfoll et al., 2018). In comparison to veteran employees, newcomers usually have few opportunities to gain resources because of their relatively short tenure in the new organization. Moreover, new employee groups have limited resource reserves, and integration into the new company requires a large amount of personal and organizational resources. More often than not, newcomers are frequently seen to suffer from resource loss and a relative resource imbalance. The resource gain paradox states that in the face of resource loss, timely resource replenishment is more crucial and valuable to individuals (Hobfoll et al., 2018). At this point, replenishing those already low on resources with fresh ones can help them promptly and do more to stop the ongoing loss of resources (Hobfoll et al., 2018). For individuals with scarce resources, the replenishment and supply of resources are helpful for relieving stress and tension.

First, from the perspective of COR (Hobfoll et al., 2018), job resources influence work engagement in a motivating way. A valid orientation program is an important form of job resources, and it enables newcomers to accumulate needful resources following their entry into the employment organization. This initial gain is favorable to developing resource gain spirals (Hobfoll et al., 2018), which means that a person’s initial resource accumulation enhances later resource gain (Hobfoll et al., 2018). In other words, employees with more initial individual resources will have an advantage of greater resource gain.

Second, according to the crossover effect of resources (Hobfoll et al., 2018), the impact of orientation training on work engagement is likely to be realized through personal resources (PsyCap). In this respect, SRT theory (Saks and Gruman, 2018) also notes that socialization resources affect personal resources, which in turn have an influence on newcomers’ distal socialization outcomes including work engagement organizational commitment, and turnover intention, among others. As a result, this work incorporates newcomers’ PsyCap as a particular mediator in the model, which will aid in elucidating the principles governing resource cluster operation.

Last but not least, it is also clear how LMX and education level fit into the job engagement paradigm. The leader-member exchange theory (Graen and Uhl-Bien, 1995) postulates that “in-groups” have a greater quality of leadership member exchange than “out-groups”; and that, as a result, employees may receive more support in the form of psychological or material resources. While low-quality LMX is likely to obstruct the access to the flow of resources and negatively impact engagement, high-quality LMX acts as wide resource caravan routes to facilitate the smooth passage of resource caravans (Hobfoll et al., 2018; Liu et al., 2023). Furthermore, education can be seen as an energy resource that serves to help newcomers acquire more resources (Hobfoll, 1989) and preserve and enhance personal resource pools, so it is likely to be a vital moderator at the individual level.

According to Luthans et al. (2007), PsyCap is highly malleable and inclusive and, therefore, can be consciously cultivated through certain specific measures. On this basis, orientation training is one of the essential methods for developing PsyCap (Saks and Gruman, 2011). In contrast, from the existing literature, there still needs to be more evidence on the relationship between orientation training and PsyCap. However, the causal relationship between orientation training and the lower-level components of PsyCap has made some progress. For example, Ashforth et al. (2012) argue that orientation training keeps newcomers optimistic and hopeful by enhancing their job competencies. Moreover, a recent investigation by Huang and Hung (2022) in high-tech companies notes that orientation training provided by organizations has a positive impact on developing employees’ motivation and self-efficacy and helps to improve their cognitive level and job performance. This is because newcomers can quickly grow accustomed to the workplace and become competent in their job duties through orientation training. Such positive experiences, in turn, give newcomers courage to face and overcome difficulties in career development. Based on the foregoing, it is reasonable to predict that orientation training is likely to positively impact PsyCap as well. This is because PsyCap is the caravan resource of the four dimensions including self-efficacy, resilience, hope, and optimism (Luthans et al., 2008). Based on the foregoing, the first hypothesis of our study is developed:

The COR theory states that individuals who possess sufficient resources can better manage stress and have a greater desire for work productivity (Hobfoll, 1989). Individuals will also attempt to nurture resource gain spirals to enhance their resource reserves (Wang et al., 2019). As one’s positive psychological resources, PsyCap may provide employees the inner strength and security required to work (Li et al., 2015). For instance, employees with high PsyCap have greater confidence in their talents and abilities (Zhou et al., 2019). They also maintain optimism about the present and the future and are less likely to become trapped in challenging circumstances, demonstrating higher work engagement (Zhou et al., 2019). In addition to the objective benefits that psychological capital brings to employees mentioned above, employees with high PsyCap are more willing to devote themselves to work that brings resource returns to accelerate the resource gain spiral, thus showing high dedication. Therefore, it is reasonable to speculate in our study that the higher the PsyCap of newcomers, the more engaged they will be at work. Accordingly, our study proposes the second hypothesis:

In the literature, PsyCap is often considered as a mediator between resources and outcomes in the organizational context. Ngo et al. (2023), for example, investigate the mediating role of PsyCap in the relationship between employees’ development policies and their reported well-being. In the context of organizational socialization, newcomers might use orientation training to foster their PsyCap (noted earlier) on the one hand; and on the other hand, individuals with high PsyCap may invest in resources and show high engagement at work to increase the possibility of obtaining potential future resources (e.g., organizational recognition, income, and job promotions). Jointly, these would suggest the mediating role of PsyCap in the process of developing and maintaining newcomers’ work engagement. This concurs with Luthans et al.’s (2007) theoretical notion that PsyCap is often treated as a mediator in the relationships between job resources and outcomes. This also concurs with Saks and Gruman’s (2018) argument that socialization resources nourish newcomers’ PsyCap and further maintain and develop work engagement. A review of the literature indicates that there has been a lack of empirical evidence on the mediating role of PsyCap in the foregoing relationships. We, therefore, propose the third hypothesis:

According to COR theory, resources are constantly present in the environment, climate, and leadership that are unique to the organization. These factors can either support and nurture the resources in the “passageway” or obstruct and restrict them (Hobfoll et al., 2018). Among them, LMX is considered a resource caravan passageway (Hobfoll et al., 2018). Specifically, employees who maintain high levels of exchange with their leaders are frequently perceived as “insiders,” with interactions characterized by higher levels of support, trust, and respect. In contrast, employees who exhibit low exchange relationships tend to be viewed as “outsiders,” and their interactions are only within the formal work context (Ke and Ding, 2020).

Therefore, our study hypothesizes that newcomers with high quality LMX are more likely to access and utilize orientation training resources and thus exhibit higher levels of work engagement than those with low-quality LMX. In a high-quality LMX, leaders provide tangible and intangible job support to their subordinates, making employees feel more trusting and secure (Ke and Ding, 2020). It is crucial for organizations to enable their new associates to feel comfortable with the orientation training, which in turn promotes higher level of work engagement. Besides helping newcomers grasp the training material, high-quality LMX can give the employees more positive feedback on their work. This in turn helps motivate newcomers to exhibit high level of work engagement. In the literature, LMX is usually considered as a moderator. Xu et al. (2022), for instance, document that LMX moderates the relationship between social undermining and employee silence. In socialization literature, there has been a lack of empirical evidence regarding the moderating role of LMX in the relationship between orientation training and work engagement. Based on the foregoing discussion, the fourth hypothesis of this study is therefore developed:

Education is an indicator of individual competence (Wang and Wang, 2016). In general, employees with higher education degrees typically possess more advanced cognitive, learning, information-seeking, and analytical decision-making skills than those with lower levels of education. For example, education is highly correlated with employees’ innovative behavior; and employees with higher education can better facilitate and engage in creative activities in the company (Kong, 2017). In the present study sample, hotel newcomers’ education level is likely to moderate the relationship between orientation training and PsyCap, meaning that those with higher education are more likely to make better use of organizational resources (e.g., effective orientation training) to foster their personal resources (e.g., PsyCap). In particular, higher education level frequently denotes more vital personal abilities (e.g., learning, cognition, etc.), which aids in enabling new hotel employees to learn and comprehend the pertinent training materials more quickly, effectively converting the resource advantage into a psychological advantage and having higher PsyCap. In this respect, most early studies (e.g., Kong, 2017) concentrate on the impact of education on corporate performance. In the domain of organizational socialization research, Liu et al. (2023) report that income moderates the relationship between task mastery and work engagement. An argument could be extended to expect that income moderates the relationship between orientation training and work engagement. Cai et al. (2021) and Harris et al. (2022) also call for more empirical investigations on the moderating effect in newcomers’ organizational socialization research. Therefore, the fifth hypothesis of this study is developed:

The questionnaire contains four core constructs (orientation training, PsyCap, LMX, and work engagement) whose measurement items are detailed in Appendix A. The four constructs each are measured by using a well-established Likert scale. Orientation training was measured by using Zheng’s (2018) scale ranging from 1 (very poor) to five (very good). Respondents were requested to indicate the levels of training effectiveness on, for example, orientation to organizational culture and history. PsyCap was measured using a 6-point Likert scale (1 = strongly disagree, 6 = strongly agree) developed by Luthans et al. (2007). An example is, “I always feel confident when presenting in front of many colleagues,” etc. LMX was measured by Graen and Uhl-Bien (1995) using a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree), such as “I am clear about whether my leader is satisfied with my performance,” etc. Work engagement was measured by Schaufeli et al. (2006) using a 7-point Likert scale (0 = never, 6 = always), such as “I feel motivated at work,” etc. Finally, the study collected newcomers’ demographic information.

It should be mentioned that we took Greenbaum et al.’s (2018) suggestion by measuring LMX from employees’ perspectives for two main reasons. One is that our study focuses on exploring the behaviors and attitudes of a newcomer, and therefore his or her perceptions of LMX are critical. The other is that previous scholars (e.g., Minsky, 2002) measured LMX for leaders and subordinates but they found perceptual differences between leaders and associates. The differences are mainly due to communication problems.

Prior to formal data collection, we conducted a pilot test and collected 20 copies of questionnaires from hotel newcomers. The Cronbach alpha value of each latent constructs in the questionnaires was greater than 0.90. In the main study, respondents must meet the following two preconditions: (a) non-managerial employees from 3- to 5-star hotels and (b) work tenures between 1 to 12 months following entry into the employment organization. We used snowball sampling technique to collect the empirical data. The formal data were collected between December 2021 and April 2022. We adopted the strategy of online data collection by using the “Wenjuanxing” platform ² for two notable reasons. One is that this platform is widely used in China (Wang et al., 2021). The other is that offline data collection was impossible due to the COVID-19 problem. We motivated respondents’ engagement to fill out the questionnaires by using a random bonus package ranging from 1–10 RMB yuan. As a result, 674 questionnaires were returned, among which 77 invalid questionnaires (e.g., responding time being less than 90 s) were removed, and 567 valid questionnaires were retained. The socio-demographic characteristics are presented in Table 1.

To minimize the effect of common method variance/bias (CMV/B), Podsakoff et al. (2003) suggest the use of different scales for different constructs. Therefore, the points of the Likert scales among the four constructs in the present study differ from each other. Moreover, respondents’ anonymity was assured, and their engagement in filling out the questionnaire was motivated (noted earlier). All these strategies are effective and helpful in minimizing the effect of CMV/B. Apart from the foregoing, two more methods were adopted to detect the potential problem.

Specifically, one post-hoc test (i.e., one-factor method) was conducted to detect the CMV/B problem. All the measurement items for the four latent constructs were entered in SPSS 26.0, and principal component factor analysis was analyzed. The principal component factors turned out to be not a single common factor but multiple factors. The first factor explains 18.80% of the total variance, which is far below the threshold level of 50% (Harman, 1976).

The other post-ho test involves the unmeasured latent method construct (ULMC, Richardson et al., 2009) method. In particular, we have built four different models using AMOS 24.0: (a) model A (the trait-only model, χ²/df = 4.155), (b) model B (the method-only model, χ²/df = 9.853), (c) model C (the trait/method model, χ²/df = =3.655), and (d) model D (the trait/method -R model, χ²/df = 3.077). The results show that the trait-only model (i.e., model A) fits better (∆χ² = 4373.231, p = 0.000) than the method-only model (i.e., model B), suggesting that observed variance in the independent and dependent constructs is not because of method alone. The trait/method model (i.e., model C) fits better than (∆χ² = 227.700, p = 0.000) the trait-only model (i.e., model A), which shows that trait-based and method variance is presented in the data. Finally, the revised trait/method model (i.e., model D) fit better (∆χ² = 700.438, p = 0.000) than the trait/method model (i.e., model C), providing no substantial evidence of bias because of CMV.

The data were analyzed using SmartPLS 3.0. This method is suitable for complex models such as the one (with both moderation and mediation) in the present study. In addition, most of the causal paths in the present study are exploratory, and PLS-SEM is, as per Hair et al. (2017), a suitable method. Generally, the reliability and validity of the theoretical constructs have been guaranteed.

First, the factor loading of each construct’s specific items was tested, and their values were between 0.732 and 0.864 (see Table 2), higher than the threshold level of 0.7 (Hair et al., 2017). Second, the composite reliability (CR) ranges from 0.936–0.955 (see Table 2), which is greater than the criterion of 0.7, indicating that each latent construct exhibits a high level of internal consistency (Hair et al., 2017).

In terms of convergent validity, Hair et al. (2017) suggest that the AVE value of a construct should be greater than 0.50. The present study’s AVE values range from 0.635 to 0.675 (see Table 2), all greater than the foregoing threshold level. With regard to discriminant validity, we used the following two approaches to assess it. The first one concerns the square root of each AVE value. Table 3 indicates that the square root of each AVE value on the diagonal is greater than the correlation coefficients between the two corresponding constructs. This would suggest that all four variables in our study have good discriminant validity according to the criteria suggested by Ab Hamid et al. (2017) and Hair et al. (2017). The second discriminant validity assessment involves, as per Henseler et al. (2015), the HTMT (Heterotrait-Monotrait) Ratio test. In particular, when HTMT value is less than 0.85, the corresponding construct can be considered to have good discriminant validity (Henseler et al., 2015). In the present study, HTMT values in Table 4 show that all the four latent constructs exhibit acceptable level of discriminant validity.

Turning to the model fit, Hair et al. (2017) suggest that the SRMR (Standardized Root Mean Residual) value should be less than 0.08. In the present study, SRMR values in the Saturated and Estimated models are 0.029 and 0.032, respectively. These results suggest that the model fits the data well in the present study. Finally, Hair et al. (2017) suggest that VIF values of measurement items should be less than 5.00. In the present study, VIF values range from 1.889 to 3.700. This would suggest that VIF problem is not an issue in our study.

We tested all hypotheses in this study using 567 parent samples with 5,000 complete bootstraps enabled in SmartPLS 3.0. Specifically, we chose bias-corrected complete bootstrapping to test research hypotheses at the significance level of 0.05 (2-tailed). One advantage of this method, as per Hair et al. (2019), lies in its ability in working around the potential estimation bias caused by multivariate nonnormal data. As a result, our hypothesis testing results are less likely to be biased as per Hair et al. (2019) and the results are summarized in Table 5. In particular, H1 concerns the effect of orientation training on PsyCap, and it gains empirical support (β = 0.608, p = 0.000) in this study. H2 involves the effect of PsyCap on newcomers’ work engagement, and it has also been supported (β = 0.432, p = 0.000) in the present empirical data. H3 relates to the mediating role of PsyCap, which has proven it to be a substantial (β = 0.257, p = 0.000) mediator between orientation training and work engagement. Moreover, PsyCap’s mediating role has been further confirmed by the following situations. As shown in Table 5, when controlling for PsyCap as a mediator, path C was not significant (β = −0.020, p = 0.638); but when PsyCap is not being controlled in the model with only one antecedent (i.e., orientation training) and one outcome (work engagement), path C′ was, however, significant (β = 0.432, p = 0.000).

H4 and H5 are moderation hypotheses. Specifically, H4 hypothesizes that LMX moderates the relationship between orientation training (independent variable) and work engagement (dependent variable). Results in Table 5 indicate that the interaction variable (i.e., LMX*orientation training) significantly (β = 0.065, p = 0.025) influences work engagement. The moderation plot is shown in Figure 1. A post-hoc group regression analysis was conducted using PLS-SEM3.0, and the results show that in terms of the positive effect of orientation training on engagement, new employees with high LMX have been more affected by engagement (β = 0.236, p = 0.000) than those with low LMX (β = 0.219, p = 0.000). In other words, results reveal that newcomers with high LMX were more effective in using social resources such as organizational orientation training to increase their engagement than their counterparts with low LMX. Jointly, the foregoing results would suggest that H4 gains empirical support in this study.

Likewise, H5 hypothesizes that education plays a moderating role between orientation training (independent variable) and PsyCap (dependent variable). Table 5 shows that the variable (i.e., education*orientation training) significantly (β = 0.096, p = 0.027) affects PsyCap. The moderation plot is shown in Figure 2. Furthermore, a post-hoc group regression analysis was conducted using PLS-SEM3.0. The results show that in terms of the positive effect of orientation training on PsyCap, new employees with education at college and above were more affected by PsyCap (β = 0.674, p = 0.000) than those with education at high school and below (β = 0.560, p = 0.000). In other words, newcomers with higher education (college and above) are more likely to make good use of socialization resources (i.e., orientation training) to maintain and develop their PsyCap levels than their counterpart group of lower education (high school and below). Taken together with the foregoing results, H5 is empirically supported in the present data.

In the domain of organizational socialization, the research framework in this study is one of the few studies (e.g., Liu et al., 2023) that are theoretically and jointly built on COR theory and SRT theory. In general, our research framework as well as its study findings helps managers and the academic community understand how, why, and under what conditions orientation training affects newcomers’ work engagement. Our empirical findings add to the body of knowledge on organizational socialization and broaden and deepen the explanatory capacity of COR theory and SRT theory on the phenomena of newcomers’ organizational integration in several ways.

First, there has been a lack of documented report on the direct and positive effect of orientation training on newcomers’ PsyCap; and our study provides the empirical evidence on this research hypothesis (H1). This would suggest that Saks and Gruman’s (2018) theoretical notion of this particular causal linkage gains empirical support successfully. This also lends empirical support for the argument of COR: organizational resources have spillover effect on personal resources (Hobfoll et al., 2018). The finding regarding the direct linkage between orientation training and PsyCap is empirically exploratory and thus contributes to the literature substantially. Apart from orientation training, future studies should both empirically investigate more organizational and job resources as listed in Saks and Gruman’s (2011) work and examine their impact on newcomers’ PsyCap. Future studies are particularly warranted to examine other types of orientations—e.g., Wen et al.’s (2023) cultural orientation—as well as their influences on socialization process and outcomes.

Second, COR theory proposes that resources do not exist individually and separately, but usually present themselves in a cluster (Hobfoll et al., 2018). In line with this proposition, the four newcomers’ four HERO resources should be considered and examined as a cluster of PsyCap. However, more often than not, most previous socialization research examines HERO as four independent resources only (e.g., Self-efficacy, Bauer et al., 2007), with a neglect of treating HERO as a cluster of PsyCap with only a few exceptions (e.g., Ngo et al., 2023). The empirical findings regarding PsyCap as well as its relationship with work engagement, have been accordingly very rare. For instance, Karatepe and Karadas (2015) investigate only self-efficacy as well as its effect on work engagement. The present study differs from most previous socialization research. That is, we theoretically consider PsyCap, as per Luthans and Youssef-Morgan (2017), as a theoretical construct which is organically built on its four HERO components. In this study, PsyCap has proven itself to be a theoretical construct with both convergent and discriminant validity. As such, we lend empirical support for both COR (Hobfoll et al., 2018) and PsyCap (Luthans and Youssef-Morgan, 2017) theories in that PsyCap is a resource caravan rather than four independent HERO resources. This drops an important theoretical implication that future research should continue to consider PsyCap as a distinct and legitimate core construct and take a step further to examine more nomological network of PsyCap (e.g., employee development policies, Ngo et al., 2023) for one main reason. That is, SRT theory emphasizes that the four individual HERO resources “operate synergistically, and their overall PsyCap may demonstrate the most vital relationship with socialization outcomes (Saks and Gruman, 2011, p. 6).”

Third, in their SRT model of engagement pathway to organizational socialization, Saks and Gruman (2018) propose that socialization resources (e.g., orientation training) have both direct and indirect effects on work engagement. This proposition has only been partially supported in the present study. Specifically, Table 5 indicates that orientation training predicts work engagement directly without controlling for the mediator of PsyCap (β = 0.432, p = 0.000). After controlling for the mediator, the direct effect of orientation training on work engagement disappears (β = −0.020, p = 0.683). This reveals that PsyCap is a full, rather than partial, mediator in the relationship between orientation training and work engagement. This suggests that SRT model of engagement pathway to socialization (Saks and Gruman, 2018) needs to be refined when socialization resources and personal resources are narrowed down to the specific resources of orientation training and PsyCap. The refinement concerns the foregoing direct versus indirect effect of orientation training on work engagement. It is particularly felt considering that SRT approach to socialization research is a new paradigm, and SRT is still in its infancy (Ashforth et al., 2012; Cai et al., 2021).

Finally, in response to the relative lack of moderating effects in organizational socialization (Cai et al., 2021), we have examined the moderating roles of LMX and educational level in the research framework (Figure 3). In particular, we consider LMX as a resource caravan passageway in the process of organizational socialization, and its quality is likely to regulate the speed of various resource fleets on the “socialization path” and its impact on work engagement. Our empirical result indicates that the effect of orientation training on work engagement is stronger when LMX is high than when it is low. This finding is exploratory and thus significantly contributes to the body of literature. This provides empirical support for one notion of COR (Hobfoll et al., 2018): “LMX is a resource passageway, and it plays an important role in follower engagement” (p. 9).

In terms of the positive effect of orientation training on PsyCap, our empirical result reveals that this effect is stronger when education is high than when it is low. Education is a type of energy resource, as opposed to LMX, which makes up the general environment of resource passageway (Hobfoll et al., 2018). The finding regarding the moderation role of education in the present study is essentially exploratory. Therefore, it is suggested that energy resources, as well as their roles in the socialization process, should be examined and included in future SRT models.

In the present study, the effect of orientation training on work engagement is fully mediated by PsyCap. In other words, organizational resources (e.g., orientation training) affect personal resources (e.g., PsyCap), which in turn affect work engagement. This would suggest that orientation training must be ensured to be positive and effective. Otherwise, ineffective training leads to lower level of PsyCap, which in turn results in disengagement. As such, orientation training programs should be well designed, implemented, and evaluated from time to time so that both trainers and trainees are well engaged in achieving the effectiveness of orientation training.

In practice, hotel training is oriented, more often than not, to newcomers’ skills and abilities but not to their psychological state, such as PsyCap. In the present study, PsyCap has proven itself as a resource caravan and it influences work engagement positively. One practical implication is that following their entry into the organization, newcomers should be tested for their PsyCap levels from time to time. This is due to PsyCap’s nature of plasticity, malleability, and openness to change (Luthans and Youssef-Morgan, 2017). In case of new hires with low PsyCap levels, managers should diagnose and intervene in such problems. The major objective of PsyCap development programs would not necessarily be building new knowledge and skills, but intervening and enhancing, for instance, newcomers HERO in that they can do well by means of using existing and potential resources. Special programs like positive experiences, performance feedback, and effective coaching/modeling could be developed and tailored to augment newcomers’ PsyCap levels. In short, PsyCap is subject to change, and once higher levels of newcomers’ PsyCap have been developed and maintained, higher levels of work engagement will be accordingly developed.

The present study has examined two moderators including LMX and education. In particular, newcomers with college and above education make better use of training resources than their counterparts with senior middle school and below education to cultivate PsyCap. This leads to the following implication—It is important to recruit employees, particularly the front-line staff, with an education background of college or above. The other moderator is LMX, a resource caravan passageway (Liu et al., 2023). LMX is a way for managers and subordinates to achieve resource flow. High-quality LMX can provide newcomers with various explicit resources (e.g., information bias) and implicit resources (e.g., supervisor support). Low-quality LMX means managers and subordinates are often filled with dissatisfaction, suspicion, and even resistance (Graen and Uhl-Bien, 1995). The results of our study suggest that LMX moderates the relationship between orientation training and newcomer engagement. This finding is exploratory in organizational socialization research domain. But it also echoes Xu et al. (2022) finding regarding the moderating role of LMX. Namely, in organizational context the interactive effect between LMX and social undermining affects employee silence significantly. As such, we extend the literature substantially. In the present study, the relationship between orientation training and PsyCap is specifically stronger when education is high than when it is low. Therefore, it is necessary for organizations to promote resource exchanges between managers and newcomers. The hotel can also set up a comparable prevention and intervention system so that when newcomers report low levels of LMX, hoteliers should diagnose the problem by conducting in-depth interviews, for example, to identify the root causes and quickly implement corresponding remedies.

Our empirical data are self-reported and cross-sectional, which may lead to common method variance (CMV/B). These data collection methods have their limitations but are reasonable for two main reasons. First, Podsakoff et al. (2003) argue considerably on decreased or increased causal relationships. However, Lance et al. (2010) document that the common method’s inflationary bias is mitigated mainly by attenuation from measurement inaccuracy. Bauer et al. (2007) consider cross-sectional data a reasonable alternative strategy to capture the dynamics of newcomers’ organizational socialization phenomenon. To address CMV/B issue, we, therefore, conduct post-hoc statistical analysis to diagnose the potential CMV/B problem. As noted earlier (Section 3.3 Common Method Variance/Bias), the statistical results indicate no substantial evidence of bias because of CMV. Furthermore, to lessen method bias, techniques such as ensuring respondent anonymity, as advised by Podsakoff et al. (2003), were utilized at the response reporting stage. Second, all the latent constructs in our overall models are generated from psychological and subjective impressions that are intangible but perceptible. As a result, the self-reporting approach is suitable, as per Song and Chathoth (2011), in that people’s true perceptions and intents can typically be reported by no one else but themselves.

Because most of our study findings are exploratory, the generalizability of our study findings is unknown. Given this limitation, it is necessary to verify our study findings in future studies with longitudinal data and in other service industries, national cultures, and geographical areas. Future studies should investigate the roles of more socialization resources (e.g., organizational support) as listed in SRT (Saks and Gruman, 2018) and personal resources (e.g., core self-evaluations) in the development of newcomers’ work engagement. Finally, more moderators (e.g., emotional intelligence) should be investigated in future SRT models.