Perceived overqualification and work engagement: the moderating role of organizational size

Testing two competing hypotheses based on different theories (i.e., the theory of career mobility and social exchange theory), this study examines whether organizational size enhances or buffers the negative relationship between perceived overqualification (i.e., having more education or qualifications, skills, and/or experience than one’s job requires) and work engagement (a positive, affective-motivational state of fulfillment characterized by vigor, dedication, and absorption). The study sample consisted of 107 university graduates. A time-lagged design with two data collection points was implemented. Data were analyzed using moderated regression. The results showed that organizational size enhanced the negative relationship between perceived overqualification and engagement (vigor, dedication, and absorption). This is the first study to show the moderator role of organizational size in the investigated relationship. To mitigate the dysfunctional consequences of perceived overqualification on work engagement, we suggest that large organizations should give overqualified employees the opportunity to develop long-term, high-quality relationships in their work environment.


Introduction
Overqualification (having more education and/or competences and skills than the job requires) is a major problem among recent graduates in Western countries. For instance, according to the Spanish Statistical Office (INE, 2020), around 37% of Spanish employed graduates under 30 years old are overqualified. The Organization for Economic and Co-operation and Development (OECD) reports that about 22% of Spanish adult (aged 16-65 years) workers are overqualified (OECD, 2019). This percentage of overqualified workers is similar in other countries, such as Greece, Germany, or the United States, or even higher in countries such as France, the United Kingdom, Canada, and New Zealand (OECD, 2019). These data are worrisome because previous studies indicate that perceived overqualification has negative consequences for employee well-being (Harari et al., 2017).
Little is known about the variables that might moderate this negative influence of perceived overqualification (Erdogan et al., 2011). This gap is problematic because it means that we do not fully understand under what conditions overqualification is especially dysfunctional, and when its impact can be attenuated. Therefore, it is theoretically and practically important to carry out studies that help to identify the variables that can moderate the negative effects of overqualification. From a theoretical perspective, we need to know under what conditions overqualification's negative effects are buffered or enhanced in order to better understand the relationship between overqualification and outcome variables. From a practical perspective, a better understanding of these relationships and their potential moderator variables will allow researchers to suggest ways to use effective strategies to prevent and reduce the dysfunctional effects of overqualification. This is particularly important because overqualification can have long-term negative effects on subsequent career trajectories (Erdogan et al., 2011). This study aims to contribute to filling this gap by deepening the understanding of the variables that can moderate the negative effects of overqualification.
Specifically, in this study, we focused on the relationship between perceived overqualification and work engagement, 1 3 analyzing the moderating role of organizational size. Work engagement has been defined as a positive state of fulfilment (Schaufeli et al., 2002). We focused on work engagement because it is a motivational state that has the capacity to influence a variety of important work outcomes, such as employee safety, turnover intention, organizational commitment, and employee performance, to name a few (Bakker et al., 2014).
We chose organizational size as a potential moderator of the relationship between perceived overqualification and work engagement for several reasons. First, organizational size is a pervading contextual factor that is always present. Thus, it is capable of influencing the phenomena and relationships that develop within organizational environments. Second, recent changes in firms' environments (e.g., technological advancements and globalization) have altered the impact of size on organizational phenomena (Josefy et al., 2015). Therefore, the influence of organizational size is not clear because this area of inquiry is characterized by inconsistent findings and alternative explanations (Josefy et al., 2015). This situation offers an opportunity to clarify the moderator role of size in an important relationship for employees and organizations. Third, previous research examining moderators of the relationship between overqualification and outcomes has paid a lot of attention to personal variables, whereas the potential influence of contextual organizational factors has been somewhat neglected (Erdogan & Bauer, 2021). Investigating the moderator role of organizational size provides an opportunity to fill this gap in the literature.
Thus, the goal of this study is to clarify the moderator role of organizational size in the relationship between perceived overqualification and work engagement (see Fig. 1). The following two research questions (RQ) are addressed: RQ1. Does organizational size moderate the relationship between overqualification and work engagement?; RQ2.
Does organizational size play a buffer role or an enhancer role in the aforementioned relationship? Consequently, we tested two alternative hypotheses, based on different theories, in a sample of university graduates. On the one hand, the theory of career mobility (Rosen, 1972;Sicherman & Galor, 1990) provides some reasons that organizational size may buffer the negative influence of overqualification on work engagement. The theory of career mobility predicts that workers may go through a temporary stage in their careers, during which they hold lower-level jobs where they acquire skills that will be useful in later higher-level jobs. Large organizations have more hierarchical levels than small ones and, thus, more opportunities to move to higherlevel jobs. Therefore, overqualified employees working in large organizations may have higher promotion expectations, and this factor might mitigate the negative impact of overqualification on work engagement. On the other hand, social exchange theory (Blau, 1964;Cropanzano & Mitchell, 2005) argues that organizational size may enhance the negative influence of overqualification on work engagement. In large organizations, it can be difficult to develop high quality social exchanges (Daspit et al., 2016). This means that employees in large organizations may lack an important protective shield, which might contribute to observing a greater negative relationship between overqualification and work engagement in large organizations than in small ones.
The present study aims to contribute to the overqualification literature in several ways. First, we respond to scholars' calls to identify the moderator variables that may help mitigate the negative consequences of overqualification on individual outcomes (Erdogan et al., 2011). By doing so, we contribute to extending the nomological network of perceived overqualification by specifying under what conditions the influence of perceived overqualification on work engagement is enhanced or buffered. Second, by testing two alternative hypotheses about the moderator role of organizational size, based on different theoretical frameworks, we begin to build some consensus in the knowledge about the investigated relationships (Hollenbeck, 2008). This is an effective way to reduce the uncertainty associated with relationships that have not been examined in previous research. Third, from a practical point of view, by identifying a variable that moderates the influence of perceived overqualification on work engagement, our study suggests ways to mitigate the dysfunctional consequences of this variable that managers and practitioners should consider.
This article is structured as follows. In the "Theoretical framework and hypotheses" section, we provide theoretical justification for the two alternative hypotheses tested in this study (i.e., the buffer vs. enhancer moderator role of organizational size in the relationship between perceived overqualification and work engagement). In the "Method" section, we present information about The study research model the sample and data collection procedure, measurement instruments, and data analysis. Then, in the "Results" section, we present the results obtained for the tested hypotheses. Finally, in the Discussion, we describe the theoretical and practical implications of our findings, the limitations and strengths of our study, and suggestions for future research directions.

Theoretical framework and hypotheses
The relationship between perceived overqualification and work engagement Work engagement has been defined as "a positive, affective-motivational state of fulfillment that is characterized by vigor, dedication, and absorption" (Schaufeli et al., 2002, p. 74). Vigor is characterized by a sense of energetic and effective connection with one's work activities, resilience when facing obstacles, and the willingness to invest effort in one's work. Dedication is characterized by having a sense of enthusiasm, inspiration, and challenge while being highly involved in the work. Finally, absorption is characterized by a sense of deep concentration and being fully immersed in one's work, without noticing that time passes and finding it difficult to detach from work. The relationship between perceived overqualification and work engagement can be explained by the theories of relative deprivation (Feldman et al., 1997) and personenvironment fit (Kristof-Brown & Guay, 2011). According to the relative deprivation theory, individuals who are in an optimal employment situation (e.g., they have a job that matches their education and/or competences and skills) are more likely to invest personal resources and effort in performing their jobs. In contrast, when individuals have jobs that require less education and/ or competences and skills than the ones they have, their jobs are not likely to provide them with enough purpose and meaning. Consequently, they will feel deprived, and their intrinsic motivation and work engagement will be negatively affected (Dumani, 2015). According to personenvironment theory, because of the mismatch between their qualifications and jobs and their unmet expectations, overqualified employees will be less willing to become involved in and dedicated to their work with vigor and absorption (Luksyte et al., 2020). Therefore, their work engagement will be low. The few studies that have examined the relationship between perceived overqualification and work engagement offer empirical support for the negative relationship between the two variables (Lou & Ye, 2019;Luksyte et al., 2020;Ma et al., 2020).

Organizational size as a buffer of the relationship between perceived overqualification and work engagement
As mentioned above, two theories suggest two alternative moderator roles of organizational size (buffer vs. enhancer) in the relationship between perceived overqualification and work engagement. First, we focus on the buffer role.
The theory of career mobility posits that "workers may temporarily work in jobs that provide them with skills to be used later in a different, higher-level job" (Sucherman, 1991, p. 103). Therefore, it can be positive to go through a temporary career stage in a lower-level job where the skills acquired will be useful in different future jobs. According to the theory of career mobility (Rosen, 1972;Sicherman & Galor, 1990), graduates may temporarily accept a job for which they are overqualified in order to obtain an initial income and gain some work experience to help them move up in their careers. Sicherman (1991) provided empirical evidence that overqualified workers have higher rates of occupational mobility. Thus, they are more likely to move to a higher-level occupation and a higher wage level than non-overqualified workers. Additionally, large organizations typically have more opportunities for advancement within the organization, and "promotion-based schemes will be used more in large organizations with many hierarchical levels than in smaller organizations with fewer levels" (Baker et al., 1988, p. 601). Accordingly, large organizations can often offer higher wages, more benefits, and other extrinsic motivators than smaller organizations (Brown & Medoff, 1989). Moreover, organizational size is positively associated with opportunities for objective career success (Nabi, 1999). Promotion opportunities are an indicator of the possibility of career advancement within the organization (Frenkel & Bednal, 2016), and promotion acts as a long-term incentive for workers (Takahashi, 2006).
These ideas lead us to propose that, because employees in large organizations have higher expectations for promotion than employees in smaller firms, these expectations of future advancement can buffer the negative influence of perceived overqualification on work engagement. Therefore, we hypothesize the following: H1. Organizational size positively moderates the negative relationship between perceived overqualification and the three dimensions of work engagement, vigor (H1a), dedication (H1b), and absorption (H1c), so that the relationship is weaker when organizational size is large than when it is small.
We posit the same relationship for the three dimensions of work engagement for the following reasons: first, they share a common motivational nature; and second, previous 1 3 research has shown that the three dimensions, although distinct, are highly correlated (Schaufeli et al., 2002(Schaufeli et al., , 2006.

Organizational size as an enhancer of the relationship between perceived overqualification and work engagement
Social exchange theory (Blau, 1964;Cropanzano & Mitchell, 2005) posits that "social exchange comprises actions contingent on the rewarding reactions of others, which over time provide for mutually and rewarding transactions and relationships" (Cropanzano & Mitchell, 2005, p. 890). Theorists agree that social exchange involves a series of interactions that are seen as interdependent and contingent on the actions of another person. According to social exchange theory, exchanges in organizations follow a continuum ranging from generalized to restricted (Daspit et al., 2016). Small organizations tend to follow a generalized social exchange system, characterized by long-term, high quality, and trustbased relationships and a collective identity orientation (Long & Mathews, 2011). In contrast, large companies would be more likely to follow a restricted social exchange system, characterized by time-limited, instrumental, and impersonal relationships, and encompassing economic values rather than socioemotional values (Daspit et al., 2016;Long & Mathews, 2011).
Previous research has shown that workers in small organizations report more supportive work environments (Mac-Dermid et al., 2001) and enjoy better relationships with their managers (Tsai et al., 2007). Moreover, close relationships with coworkers offer employees greater benefits (MacDermid et al., 1994). In larger organizations, employees have fewer opportunities to develop high-quality relationships (i.e., more informal, deeper, and with more personal contact) with co-workers and managers (Bencsik & Juhász, 2010). Moreover, large size in organizations promotes higher structural complexity, which is likely to negatively affect communication among employees (Jung, 2013). Research on worker alienation also suggests that employees in larger organizations experience fewer relationships based on close personal contact, which can result in a decrease in social integration (see Schminke et al., 2002). Therefore, based on previous literature, small organizations can be conceptualized as generalized social exchange systems, whereas large organizations can be seen as restricted social exchange systems.
Thus, we propose that the quality of the interpersonal relationships that arise in small companies creates a type of generalized social exchange that is valued by employees and mitigates the negative relationship between perceived overqualification and engagement. However, in large organizations, employees generally have a lower level of quality in their interpersonal relationships that cannot protect them from the negative influence of overqualification on work engagement. Accordingly, we expect that the larger the size of the organization, the stronger the negative relationship between overqualification and engagement will be.
H2. Organizational size negatively moderates the negative relationship between perceived overqualification and the three dimensions of work engagement, vigor (H2a), dedication (H2b), and absorption (H2c), so that the relationship is stronger when organizational size is large than when it is small.

Sample and procedure
The present study was part of a larger research project about perceived overqualification of recent university graduates. We recruited bachelor and master's degree students from a Spanish public university. A total of 10,307 students received emails containing the links to the online questionnaires approximately two months before graduation (Time 0). A total of 1,087 students answered the call and were informed that they would be contacted again after graduation. In order to reduce common method bias and establish a temporal sequence for the variables, we conducted time-lagged research with two data-collection points after graduation. Time 1: between six and nine months after graduation, and Time 2: four months after time 1. We chose these time frames because previous studies have suggested 6 to 9 months to be enough time for graduates to find a job (e.g., McArdle et al., 2007). In addition, a minimum of four months was considered enough time for participants to learn their roles and know about their organization. To increase the response rate, we offered a training course to the students who completed all the questionnaires over time, and we also sent several reminders to participants.
At Time 1, 503 graduates filled out the questionnaire. At Time 2, 360 of the graduates responded to the survey 1 . However, 253 of them changed jobs from Time 1 to Time 2. Therefore, only the 107 university graduates that kept the same job between Time 1 and Time 2 were included in the final sample of this study.
Of these 107 university graduates, 71% were female, and their average age was 26.6 (SD = 7.1). Regarding educational levels, 48.6% had completed a bachelor's degree, and 51.4% a master's degree. Finally, regarding the field of the degree obtained, 58% of the participants had completed a degree in social sciences, 14% in humanities, 18% in health sciences, 8% in natural sciences and mathematics, and 2% in engineering.

Measures
To translate the original scales into Spanish, we followed a translation/back-translation procedure (Brislin, 1970). First, a group of bilingual experts in employability translated the items from the original scale into Spanish. Second, the Spanish version was translated into English by a bilingual native English speaker. This last English version was compared to the original one, and only minor changes were made to improve the comparability of the English and Spanish versions. Finally, the Spanish version of the scales was responded to by a pilot sample of 10 university students to identify potential problems (e.g., item content difficult to understand). No problems were identified.

Perceived overqualification his variable was assessed at T1
with the scale of perceived overqualification (SPOQ) developed by Maynard et al.'s (2006), which consists of nine items (see Appendix 1). Using different samples, Maynard and Parfyonova (2013) showed that the SPOQ had adequate internal consistency and test-retest reliability. They also provided evidence of content and construct validity (in terms of dimensionality and convergent and discriminant validity). Additional studies in different countries have also supported the reliability and construct validity of the SPOQ in terms of dimensionality (e.g., Alfes et al., 2016). Participants responded to the items on a Likert scale ranging from 1 (strongly disagree) to 6 (strongly agree). Cronbach's alpha was 0.95.
Organizational size This variable was measured at T1. We asked subjects to indicate how many employees there were in the organization they worked in by selecting one of the following ordinal response categories: (1) up to 10 employees, (2) between 11 and 50, (3) between 51 and 100, (4) between 101 and 250, (5) between 251 and 500, and (6) more than 500 employees. These categories are similar to those used in other studies (e.g., Lai & Fu 2021).
Engagement This variable was measured at T2 with Schaufeli et al.'s (2006) 9-item version of the Utrecht Work Engagement Scale (UWES; Schaufeli et al., 2002). This version measures three underlying dimensions measured with three items each (see Appendix 1). Previous studies showed that this 9-item version of the UWES has adequate internal consistency, construct validity, and predictive validity (e.g., Balducci et al., 2010). Items were rated on a 7-point Likert scale (1 = not at all, 7 = completely). Cronbach's alphas were 0.93, 0.94, and 0.84 for Vigor, Dedication, and Absorption, respectively.
Controls Engagement is associated with employee age, gender, and educational level (e.g., Schaufeli et al., 2006). Therefore, we controlled for graduates' age, gender (male = 0; female = 1), and educational level (Bachelor's = 0; 1 = Master's degree). Finally, 76% of the 107 participants in our dataset were still employed at T2 in the same job they had at T1, whereas 24% were no longer employed. Thus, this 24% of graduates who were unemployed at T2 reported on their level of engagement (i.e., vigor, dedication, and absorption) in the job they had at Time 1. Because employees who were unemployed may have differences in their engagement levels, we controlled for "Employment status at Time 2".

Statistical analysis
Hypotheses were tested using Hayes' PROCESS macro for SPSS (Hayes, 2013). The two alternative hypotheses were tested for each work engagement dimension.

Results
Means, standard deviations, and intercorrelations are reported in Table 1.
Prior to testing our hypotheses, we conducted a confirmatory factor analysis (CFA) with Mplus (Muthén & Muthén, 2017) to examine the distinctiveness of the four study variables (perceived overqualification, vigor, dedication, and absorption). Considering the ordinal nature of the items, weighted least squares-means and variance adjusted (WLSMV) was chosen as the estimation method. The hypothesized four-factor model (4-FM) was tested and compared with two nested competing models: a one-factor model , in which all the variables loaded onto a single factor; and a two-factor model (2-FM, perceived overqualification and engagement), in which the three dimensions of engagement were combined into a single factor. The Focusing on the work engagement dimension of vigor, the multiple regression results obtained (see Table 2) showed that perceived overqualification was negatively related to vigor (-0.31, p <. 01). Moreover, the relationship between the interaction term (perceived overqualification x organizational size) and vigor was negative and statistically significant (-0.10, p < .05). Thus, the moderator role of organizational size in the relationship between perceived overqualification and vigor was supported. To further interpret this interaction effect, we represented how the relationship between perceived overqualification and vigor changed across the range of values of the moderator variable (see Fig. 2). We chose this representation because reporting simple slopes when examining an interaction effect has been seriously criticized by methodologists (Dawson, 2014;Finsaas & Goldstein, 2021) because simple slopes are computed for very few specific values of the involved moderator (e.g., ±1SD) (Dawson, 2014;Finsaas & Goldstein, 2021). Thus, "the significance or otherwise of these slopes is indicative of something of very limited use" (Dawson, 2014, p. 13). Acknowledging these limitations of simple slope analysis, and in order to fully understand an interaction effect, methodologists recommend representing the examined moderated relationship across the range of the moderator variable (Finsaas & Goldstein, 2021).
Focusing on Fig. 2, the x-axis represents the values of the moderator, organizational size, and the y-axis represents the values of the relationship between perceived overqualification and vigor. The dark line represents the value of this relationship conditional to the values of the moderator variable. The dot lines are the upper and lower bounds for the corresponding 95% confidence interval (CI). According to Fig. 2, the relationship between perceived overqualification and vigor was negative and significantly different from   Fig. 2 shows that as organizational size increased, the relationship became more negative, which means that organizational size enhanced the negative relationship between perceived overqualification and vigor. This result supported H2a instead of the alternative H1a. Focusing on dedication, the multiple regression results obtained (see Table 2) showed that perceived overqualification was negatively related to dedication (-0.50, p <. 01). Moreover, the relationship between the interaction term (perceived overqualification x organizational size) and dedication was negative and statistically significant (-0.09, p < .05). Therefore, the moderator role of organizational size in the relationship between perceived overqualification and dedication was supported. To further interpret this interaction effect, we represented how the relationship between perceived overqualification and dedication changed across the range of values of the moderator variable (see Fig. 3). According to Fig. 3, the relationship between perceived overqualification and dedication was negative and significantly different from zero across all the values of the moderator. Moreover, Fig. 3 shows that as organizational size increased, the relationship became more negative. Thus, organizational size enhanced the negative relationship between perceived overqualification and dedication. This result provided support for H2b, but not H1b.
Finally, focusing on absorption, the multiple regression results obtained (see Table 2) showed that perceived overqualification was negatively related to absorption (-0.28, p <. 01). Moreover, the relationship between the interaction term (perceived overqualification x organizational size) and dedication was negative and statistically significant (-0.09, p < .05). Thus, the moderator role of organizational size in the relationship between perceived overqualification and absorption was supported. To interpret this interaction effect, we represented how the relationship between perceived overqualification and absorption changed across the range of values of the moderator variable (see Fig. 4). Figure 4 shows that the relationship between perceived overqualification and absorption was negative and significantly different from zero across all the values of the moderator. Moreover, Fig. 4 shows that as organizational size increased, the relationship became more negative. Therefore, organizational size enhanced the negative relationship between perceived overqualification and absorption. This result supported H2c, but not H1c.

Discussion
In this study, we examined and clarified the moderator role of organizational size in the relationship between perceived overqualification and work engagement. Considering different theoretical frameworks, we tested two competing hypotheses that suggest two alternative moderator roles for organizational size in the aforementioned relationship. Hypothesis 1 tested a buffer role (based on the theory of career mobility), and Hypothesis 2 tested an enhancer role (based on social exchange theory). We obtained empirical support for Hypothesis 2 (H2a, H2b, and H2c). Specifically, the results supported the negative moderator role of organizational size in the relationship between perceived overqualification and the three dimensions of work engagement (vigor (H2a), dedication (H2b), and absorption (H2c)). We found that organizational size enhanced the negative relationship between perceived overqualification and work engagement, thus supporting the theoretical rationale offered by social exchange theory (Blau, 1964;Cropanzano & Mitchell, 2005).

Theoretical implications
The results of our study have several theoretical implications. First, the literature examining the implications of perceived overqualification for individual outcomes has highlighted the need to identify the moderator variables that may help to mitigate the negative consequences of overqualification (Erdogan et al., 2011). Additionally, the analysis of the potential moderator role of contextual organizational factors (e.g., organizational size) has been relatively neglected (Erdogan & Bauer, 2021). Our investigation contributes to this call and to filling this gap by highlighting the importance of organizational size as a contextual variable that moderates the influence of perceived overqualification on work engagement. Our results suggest that theoretical models on the consequences of perceived overqualification should incorporate organizational size as an important moderator. In addition, our results suggest other contextual variables that might be relevant in this regard. For instance, an organizational clan culture, in which employees' needs are considered and individual development is supported (Cameron & Quinn, 2006), should also buffer the negative consequences of perceived qualification.
Second, by comparing two alternative hypotheses based on different theories (theory of career mobility and social exchange theory), our study helps to understand why organizational size moderates the relationship between perceived overqualification and work engagement. Based on the results obtained and social exchange theory, we suggest that the generalized social exchange system typical of small firms, with its long-term, high quality, and trust-based relationships and supportive environment, mitigates the negative relationship between perceived overqualification and work engagement in these companies. However, the restricted social exchange system that is more typical in large companies, with its associated time-limited, instrumental, and more impersonal relationships, cannot protect employees from the dysfunctional consequences of perceived overqualification for work engagement in these organizations. The outcome is that, as organizational size increases, the negative relationship between perceived overqualification and work engagement also increases. Given that previous research has shown inconsistent findings and alternative explanations for the influence of organizational size on different organizational phenomena (Josefy et al., 2015), our results can contribute to building some knowledge consensus about the moderator role of organizational size in the investigated relationship.
Third, to date, few studies have examined the relationship between perceived overqualification and work engagement (e.g., Lou & Ye, 2019;Luksyte et al., 2020;Ma et al., 2020). This is surprising, given the prevalence of overqualification in Western countries and the importance of work engagement in models on work outcomes (Bakker et al., 2014;Lou & Ye, 2019) tested the mediator role of thriving at work, and Luksyte et al. (2020) tested the mediator role of relative deprivation and person-environment fit. Only Ma et al. (2020) examined potential moderators of the aforementioned relationship. They identified empowering leadership and role clarity as relevant moderators. Our study contributes to expanding the short list of moderators of the "perceived overqualification-work engagement" relationship by showing that a structural characteristic of organizations (size) enhances this negative relationship. By doing so, we contribute to strengthening the contingency approach to perceived overqualification consequences.

Practical implications
Our findings have important practical implications for managers and employees. First, managers would benefit from being aware that perceived overqualification is negatively related to work engagement, and that this negative relationship is more dysfunctional in larger organizations than in smaller ones. Based on our findings and social exchange theory, managers of large organizations in charge of overqualified employees should provide them with the opportunity to develop high quality relationships in their work environment as a way to mitigate the dysfunctional consequences of perceived overqualification on work engagement. Managers can cultivate these relationships by being accessible, showing concern for employee needs, and providing instrumental and social support. Given that managers are key role models in work environments (Zohar & Luria, 2004), they have the capacity to model supportive behaviors among the subordinates.
Second, individuals who are considering accepting a position for which they are overqualified should be aware that they are likely to develop lower work engagement than if they accept a position that fits their qualification. Additionally, they should be aware that the negative influence of overqualification will be stronger if the involved organization is a large one. Based on our findings and social exchange theory, overqualified employees working in large organizations should try to create a supportive social environment at work as a way to mitigate the dysfunctional consequences of perceived overqualification for work engagement. Overqualified employees can also be active agents and shape their own work environment by practicing organizational citizenship behaviors, altruism, and helping behaviors at work.

Limitations and strengths
The present study has several limitations. First, the sample used was composed of young Spanish university graduates at the beginning of their professional careers. Thus, the national context and the educational level of the sample may limit the generalizability of our results. Future research should test the relationships addressed in this study in samples from different contexts and with greater educational diversity. Nevertheless, we want to highlight that our study focuses on a target sample where overqualification is a relevant problem (INE, 2020).
Second, we had a modest sample size due to high attrition between measurement time points. However, our attrition analyses did not indicate systematic differences between individuals who remained in our sample and those who dropped out. Nevertheless, future studies should replicate our analysis in larger samples to increase confidence in the observed relationships.
Third, in this study we only used self-report data. Thus, common-method variance could be a potential problem. However, a strength of our study was that we measured the variables involved at two different time points separated by four months. This design allowed us to mitigate some of the threats associated with common method variance (Podsakoff et al., 2003).
Finally, our study focused on a single outcome (work engagement), which limits its scope. Future studies should expand our research model by including other attitudinal and behavioral outcomes, such as intention to quit the organization and actual quitting behavior. Doing so could add value by ascertaining whether the moderator role of organizational size observed here is generalizable to other consequences of perceived overqualification.

Conclusions
By comparing two alternative hypotheses based on different theories, our study showed that organizational size enhances the negative relationship between perceived overqualification and work engagement. We hope our study will contribute to extending our knowledge about the influence of contextual organizational factors on the relationship between perceived overqualification and its consequences.