A Meta-Analysis on Employability
“A matter of Satisfaction”
Business Administration MSc’s ThesisAmsterdam Business School University of Amsterdam
June 2016
Author: João Pedro Rocha Ferreira Andrade Diogo Supervisor: Dhr. Dr. Stefan T. Mol
Statement of originality
This document is written by João Pedro Rocha Ferreira Andrade Diogo who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
“A Love affair with Knowledge will never end in Heartbreak.”
Table of Contents Abstract ... 4 Introduction ... 5 Literature review... 6 Satisfaction Work Satisfaction Job Satisfaction
Career Satisfaction (+ Professional Satisfaction) Life Satisfaction
Employability
Research Question ... 15 Method ... 20
Employability proxies
(Overall) Career adaptability
Protean Career Orientation & Boundaryless career orientation Organizational mobility preference (OMP)
Perceived Employability Age Procedures Meta-Analysis ... 30 Results ... 32 Discussion ...455 Summary of evidence Limitations Conclusions Acknowledges ... 50 Reference List ... 51 Appendix ... 60
Abstract
As a sub-product of a Meta-analyzes on Employability, this study is motivated by the changes in the dynamics of the labor market that resulted in a change from lifelong careers dedicated to one or only a few companies to careers in which employees are more mobile, moving across jobs in the same company or moving from one company to another. Therefore, the ability to stay mobile is of great importance. We will focus then is the analyses of Satisfaction because, as Employability, also Satisfaction has a central role in individual’s life as each and every one of us seeks for happiness.
Using 9 studies with a total sample size of 7.186 people, this paper uses meta-analytic techniques to investigate the relationship between Employability and Work/Life Satisfaction. Hypotheses were tested and although neither the relationships between Employability and Work Satisfaction are significant, nor the relationship between Employability and Life Satisfaction is; results do show that the moderation of age is of great importance to Life Satisfaction, being this impact of the moderation of age positive and significant in the particular case of Life Satisfaction.
There is an underlying attempt on this study to understand if the meaning of Job Satisfaction is the same as Career Satisfaction or if the difference is only semantic. To that, this study raises questions regarding if they are indeed two different concepts.
Introduction
Over the last couple of decades many studies have been published on employability, being this abstract idea of Employability something that is not new in the literature (Feintuch, 1955). During the last half century many definitions have been suggested by scholars and those are not always aligned with which other.
Also to the concept of Satisfaction some attention has been given in the past few decade. Although the philosophical concept of satisfaction goes back to ancient Greece, in the modern age one can identify a different focus as the focus is in the operationalization of the concept more than in the philosophic application of it. Since the late 30’s with Roethlisberger and Dickson’s publication of Management and the
Worker (1939) the concept of Satisfaction has been of interest for the study of human
behavior and business related research. Nowadays, wellbeing is a very important part of the western’s culture, contributing to the importance of its understanding.
With the objective of creating an overview of what is the state of the art in the employability field a meta-analysis is to be conducted. This will allow not only scholars, Human Resources practitioners and managers to have a clear vision of the meaning and better use the concept of employability, as we will discuss, but it also gives the employee a more complete tool to assess and manage his/hers capabilities and opportunities, which one will reflect also on the discussion section. In particular, the focus will stay in Satisfaction, either Work Satisfaction or Life Satisfaction, as it is, more than ever, a concern present in today’s society.
Literature review
Satisfaction
Satisfaction is a feeling of fulfilment of one’s expectations, needs, wishes or the sensation of contentment derived from this. It includes emotions of happiness, sense of wellbeing, joy, delight and enjoyment. It is a well-known concept and the “pursuit of satisfaction” is a commonly shared goal in the Western’s way of life.
Satisfaction gives people the motivation, reasons to live and it can be a measure of personal success. It is so important that it is said to be the most frequently studied variable in Organizational Behavior Research (Spector, 1997). It allows companies to keep turnover low, influences the company’s reputation and achieving both by keeping employees satisfied. Its importance creates interest not only in people working in organizations but also to the research community, naturally. Many studies were conducted in the subject of job satisfaction and in this project alone one was able to identify hundreds of studies that, one way or another, touch the subject of satisfaction in the work environment and in life.
Before going further with the understanding of Satisfaction it is important to mention that for the purpose of this study, Satisfaction was separated in two parts: Work Satisfaction and Life Satisfaction. Work Satisfaction is to be analyzed with the aggregation of Job Satisfaction and Career Satisfaction, having the latter an aggregation of Professional Satisfaction.
Work Satisfaction
As Spector (1997) said, for many people work has a central position in life, both social and personal life. Many people use work to fulfill not only their daily agenda but also to feel they contribute to something. Some, however, are not that fond of work
and, using Spector’s own words, they “hate to work” (1997, page VII) and they only work because they have to. The attitude of getting gratification from work, is called Work Satisfaction.
Job Satisfaction
Job satisfaction was defined by Smith, Kendall, and Hulin (1969) as the feelings or affective reactions to a given work-related situation and its facets. It is how people feel about their jobs and the different aspects that come with it (Spector, 1997). Aspects include possible benefits like health insurance, job resources, flex time, retirement options, vacation time or personal development opportunities, just to mention a few. The lack of benefits is also an aspect impacting Job Satisfaction.
Either focusing on the organization or on the individual employee, there are many relevant reasons to consider job satisfaction. Within the reasons, and making use of the reasoning presented by Spector (1997), the humanitarian perspective states the importance of fair treatment and respect, which by itself, is a good reason for us to care about this type of satisfaction. In addition, Job satisfaction is a measure of treatment, so high job satisfaction reflects, to some extent, good treatment. In opposition to the humanitarian perspective, the utilitarian perspective has, naturally, a rationale that refers to the relationship between satisfaction (job satisfaction in particular) and behaviors that influence organizational performance and basic organizational functioning. A study conducted by Cheri Ostroff (1992) with data from 298 schools and more than 13.000 teachers reports results that support the positive relationship between employee satisfaction and performance at the organizational level. Consequences of employee satisfaction can be translated in either desirable or undesirable behaviors.
As the reasons mentioned before are individually relevant, it is of no surprise that organizations, employers, researchers and employees have a shared concern regarding job satisfaction. It is important under many perspectives, so the motivations can be humanitarian and/or utilitarian but they can always be of use to those interested in the concept of job satisfaction.
Among job satisfaction dimensions, one can find appreciation, communication, coworkers, fringe benefits, job conditions, nature of work itself, organization, organization’s policies and procedures, pay, personal growth, promotion opportunities, recognition, security and supervision (Spector, 1997). These dimensions, called facets by Spector, are what one would call “different sides of the same coin” that is job satisfaction. Locke (1976) divides any possible dimension in four areas: other people, rewards, nature of the work and organizational context. That is, according with Lock, job satisfaction will always be related to one of the aforementioned dimensions and people get satisfaction from different sources that are always sub dimensions of those four areas. For example, one works in a company where there is a high level of companionship between colleagues; this will bring satisfaction on the dimensional level of other people which translates in a positive job satisfaction input to the individual.
Career Satisfaction (+ Professional Satisfaction)
A second sub dimension of Work Satisfaction, is Career Satisfaction. This attitudinal and subjective facet of Work Satisfaction is seen as the emotional reaction to positive and pleasurable appraisal of one’s job or job experiences (Locke, 1976) or simply how content one is with one’s job (Spector, 1997). It is, indeed, a very similar concept to Job Satisfaction but the difference resides in the fact that Job Satisfaction is focused on a particular job/role, while Career Satisfaction takes into account all the roles and
jobs practiced so far and how those jobs/roles contributed for the actual career situation.
Career Satisfaction has been linked to organizational outcomes such as intention to leave, turnover (Igbaria, 1991), organizational change and its support (Gaertner, 1989) and organizational commitment (Carson, Carson, Phillips, & Roe, 1996; Igbaria, 1991).
Because both Career Satisfaction and Professional Satisfaction aimed to measure the same in the studies analyzed, one will assume the concepts target the same facet of Work Satisfaction and therefore will be referred to only as Career Satisfaction.
As mentioned before and aligned with it, for the purpose of this study, Professional Satisfaction falls under the umbrella of Career Satisfaction as both refer to the path chosen by one regarding one’s occupation. The two concepts then fall under Work Satisfaction because that concept it a broader one. Career Satisfaction differs from Job Satisfaction because although both are equally part of Work Satisfaction, the last is more concerned with the job itself, while Career Satisfaction is more concerned with the overall satisfaction of the present result of different choices made regarding the career as agglomeration of jobs, as pointed before.
Life Satisfaction
On the side of satisfaction concerned with the wellbeing outside of work one finds Life Satisfaction. This positive emotion or attitude towards life is a cognitive and judgmental process that is defined by Shin and Johnson (1978, p. 478) as "a global assessment of a person's quality of life according to his chosen criteria".
Recent research in the field of Life Satisfaction has been growing and has been motivated by the growth of the new field of Positive Psychology (Diener et al. 1999;
Vaillant, 2000; Fergusson et al, 2015). The idea behind Life Satisfaction is that it is a desirable psychological state of positive emotions and research focusing on its different predictors is abundant. Predictors such as socio-economic factors (Bellis et al. 2012), social connectedness (Frey et al. 2004; Lucas, 2005; Gardner & Oswald, 2006; Lucas & Clark, 2006; Dolan et al. 2008; Mellor et al. 2008), (un)employment (van Praag et al. 2001; Dolan et al. 2008), physical health (van Praag et al. 2001; Ryan & Deci, 2001; Oswald & Powdthavee, 2006; Dolan et al. 2008; Bellis et al. 2012), personality traits (DeNeve & Cooper, 1998; Ozer & Benet-Martinez, 2006; Steel et al. 2008) were analyzed among many others (Fergusson et al, 2015). Taking the example of the physical health predictor, a physically healthy person will be more satisfied as a result of one’s good health than a person with lower physical health. The same happens with Employability, as a more physically healthy person score higher in Employability than one with lower physical health. Employabilty itself has an impact in Satisfaction too, as we will see.
Referring to Life Satisfaction as a positive outcome, it is easy to understand Simpson et al (1996) and Lam et al’s (1997) research results that report the negative association between Life Satisfaction and depression.
Employability
When diving in the literature regarding Employability, one easily perceives that there has been a big effort to clarify the concept over the past two decades, as it was an “obscure concept” before (McQuaid & Lindsay, 2005) that has not been clarified yet either (Gazier, 1998; Philpott, 1999; Thijssen et al, 2008) as we will see.
The definitions of Employability range from the simplistic “an individual's chance of a job in the internal and/or external labor market” (Forrier & Sels, page 106, 2003) to more elaborate ones such as the one from Hillage and Pollard (1998) which suggests that employability deals with the competence of capturing and retaining fulfilling job/work. Hillage and Pollard add that employability is “the capability to move self-sufficiently within the labor market to realize potential through sustainable employment” (1998, page 12). Harvey (2001) agrees but McQuaid, Green and Danson (2005) have a diverging perspective as they propose that “employability remains a contested concept in terms of its use in both theory and policy, and throughout the past century has been used as both a predominantly labor supply and a labor demand concept” (page 191).
Aligned with Hillage and Pollard (1998), so do Wickramasinghe and Perera (2010) refer to Employability as the ability to gain, maintain and manage employment where more and more commonly the job requirements change from one position to the next, as working place dynamics are constantly evolving. This can happen within the same organization, moving to different jobs or roles, but also between organizations. With the capability to keep up-to-date applicable knowledge in their working domain, workers are seen as scoring higher in employability than if they would not be able to do that. In addition to knowledge, also openness and resilience to change are identified as
employability capabilities (Froehlich, Beausaert, Segers & Gerken, 2014), that is, also openness and resilience contribute for higher scores of Employability.
In contemporary society, the ability to stay employable is a critical skill that needs to be present across the whole lifespan of the individual (Hess et al., 2012; Del Corso, 2013)
The term Employability is adopted with such broad amplitude and it is applied in a variety of contexts where it can lose accuracy and clarity as a concept that can be operationalized (Hillage & Pollard, 1998). As pointed out by Philpott (1999), “employability” is a concept used mainly used as a “buzzword”. Thijssen et al (2008) agree by saying it is “an attractive but confusing professional ‘buzzword’” or, as stated by Gazier (1998), it is even just a “fuzzy notion” that is not defined at all.
According to de Grip et al (2004), there are 3 levels of the definition of Employability: the core definition that deals with the potential for success of the individual in labor market situations; the broader definition that combines the willingness with the capacity of the individual to accomplish a variety of jobs and it includes the ability to learn, connecting the characteristics of the individual with the worker’s future position on the market; and the all-embracing definition that adds to the previous all the contextual and individual contextual conditions. The three definitions are focused in the worker’s future position on the labor market.
The interrelated problems with the operationalization of employability have often been lamented (Harvey, 2001; Forrier, Verbruggen & Cuyperc, 2015). To resolve this issue and create an applicable framework that could incorporate the different interpretations of what employability is, Forrier and Sels (2003) suggested a conceptual model that would help empirical researchers to identify the main factors influencing
labor market transitions for individuals. In their article the authors reiterate the importance of further research to identify a set of indicators of this “ability”.
In a world where organizational careers are being replaced by expectations of a more self-managed attitude towards individual careers we see that old fashioned paternalistic kind of contract is no longer the mainstream (Hall & Mirvis, 1995). It is up to the employee to adapt, and as a result careers have started to include positions in multiple organizations and even across industries (Arthur, 1994; Arthur & Rousseau, 1996; Hall, 2002; Leana & Rousseau, 2000). Putting a lot of emphasis on the importance of the individual, the article by Fugate, Kinicki and Ashforth (2004) introduces an employability construct that is defined as a “multidimensional aggregate of career identity, personal adaptability, and social and human capital” where employability is person-centered.
In opposition to this individual perspective, this “capability to move self-sufficiently within the labor market to realize potential through sustainable employment” (Jim Hillage & Emma Pollard, second page of Brief, 1998) is of great importance at an institutional level too as it can be used to influence policies. For instance, signed by the European Union (EU) on the 7th of February of 1993, the Maastricht Treaty has a social protocol where the Union expresses the importance of the focus on the “the development of human resources to ensure a high and sustainable level of employment” and “the integration of persons excluded from the labor market”. Although the term Employability is no longer included in the pillars of its approach as it was in the beginning, the EU is still focused on the “investment in human capital and strategies for lifelong learning”, which is a reflex of the nowadays present concern with employability (CEC, 2003; Ronald W. McQuaid & Colin Lindsay, 2005).
As we can see, the Employability field is very fragmented and studies point in many different directions. Some studies focus on the individual (Fugate et al, 2014), some on the institution (Estiénne, 1997); some studies give focus on the predictors (Gokuladas, 2011) and some on the outcomes (Chan et al., 2015). This confusion of studies and their outcomes, with some studies reporting higher impact than others as we will see, means that a Meta-Analytic of the Employability literature is long overdue. Although not aiming to a clarification or solution of the definitional issues, this Meta-analysis intends to better understand the impact of Employability on Satisfaction, how big its impact is and how can it be operationalized by individuals and organizations.
The focus on Employability has been influencing studies in many different ways. Literature shows that among the predictors of Employability we can find personality traits, particularly the “Big 5”(Patel and Thatcher, 2012; Zacher, 2013), education’s influence (Zacher, 2013; Rodrigues et al, 2015; Ye, 2015), different temporal focus of the individual (Zacher, 2013 and 2014; Froehlich et al, 2014), years of work experience (Patel and Thatcher, 2012; Stumpf, 2014; Bozionelos et al,2015), one’s self-esteem (Waters et al, 2014; Rusu et al, 2015), age (Enache et al, 2013; Johnson et al, 2013; Rodrigues et al, 2015)and focus on opportunities or limitations (Froehlich et al, 2014), just to name a few.
The outcomes of Employability are also very diverse. Salary or other benefits/rewards (Stumpf, 2014), depressive symptoms or distress (Silla et all, 2008; Stengård et al, 2014), individual’s performance (De Cuyper et al, 2011; Zacher, 2013; Rodrigues et al, 2015; Bozionelos et al,2015) or turnover intentions (De Cuyper et al, 2011; Rodrigues et al, 2015), among many other.
Research Question
Employability is then, a desirable state, allowing one to move from job to job, change company or even to change industry more easily. Satisfaction is also something desirable as it is a positive state that gets exposed by both emotions and attitudes.
When the time one spends in today’s society working takes a considerable part of the day, it is only natural that some satisfaction arises from what keeps us busy, our occupation.
As seen before, some studies were conducted to better understand both Employability and Satisfaction but both concepts are complex and studies tend to focus on a specific dimension of the concepts in particular.
To shed some light over this subject, one hopes to answer the following question: Research Question: How does Employability relate to Career and Life Satisfaction and how does age moderate this relationship?
To answer the question, the following model (Figure 1 – Research Model) was built. It shows that Work Satisfaction and Life Satisfaction are influenced by Employability. It also shows that Work Satisfaction is an aggregation of both Career Satisfaction and Job Satisfaction. To that, the moderation of Age is to be investigated.
Research suggests that employability is linked with career resources such as effective coping with uncertainty (Baruch & Quick, 2007), proactivity (Porter, Woo, & Tak, 2015) and/or career adaptability (Chan et al., 2015). Those are generality agreed to be positive employee outcomes. Empowered employees with those resources easily have more control over their job and career and this ability brings confidence in finding a new job or influencing the career itself. Employees who find themselves in the position of confidence report higher levels of wellbeing (Briscoe et al, 2012). Gamboa et al (2009) presented a study that shows that there is an influence from Employability in predicting Satisfaction with different aspects of work.
Specifically for Career Satisfaction literature also shows more than one time that individuals with higher levels of employability report higher levels of Career Satisfaction (De Vos & Soens, 2008; Grimland, Vigoda-Gadot, & Baruch, 2012). The reason for that might be the before mentioned increase of confidence that results in satisfaction with the career. For that, one expects that to be shown across studies.
Hypothesis 1: Employability is positively related to Career Satisfaction
That said, one expects that not only the same positive relationship remains present for Career Satisfaction across studies but also for Job Satisfaction, as it is part of the same Work Satisfaction concept.
Moreover, people employed with a lower past job insecurity were reported to have higher job satisfaction (Maggiori et al, 2013). Considering that Employability is closely related to job security one can expect a positive influence of Employability on
Job Satisfaction. Besides, “results show that employability and personal initiative predict in a cumulative way the intrinsic, extrinsic and social job satisfaction" (Gamboa et al, page 632, 2009). We therefore expect that Employability has a positive impact in Job Satisfaction..
Hypothesis 2: Employability is positively related to Job Satisfaction
Regarding Work Satisfaction, because it is analyzed as a result of aggregation of both Career Satisfaction and Job Satisfaction, and because both are expected be a consequence of Employability, one hypothesis that Work Satisfaction is a predictor to which Employability is positively related.
Hypothesis 3: Employability is positively related to Work Satisfaction
Job insecurity, described as the employee’s perceptions and concerns with potential involuntary loss of job (Greenhalgh and Rosenblatt, 1984; Heaney et al, 1994; Van Vuuren, 1990; De Witte, 1999) has been shown to negatively affect employees’ wellbeing (Jacobson, 1991; De Witte 1999; Cheng et al, 2005). The opposite construction of this would be that Job Security has a positive impact on employees’ wellbeing, that is, job security has a positive impact on wellbeing. One also expects that Employability being strongly related to job security, impacts positively on wellbeing of the individual under the form of Life Satisfaction. This means that Job Security, as a proxy for Employability, has positive impacts on Satisfaction, which is align with Briscoe et al’s (2012) finding as we will see below.
Also Aronsson and Goransson (1999) studied the relationship between employability and employee wellbeing, where they found that workers with low job mobility reported lower levels of satisfaction, which was evident after more than 25% of the permanent employees framed by the study report that they did not regard themselves as being in their desired profession. Employees with more mobility would more easily land jobs of their personal interest than those without job mobility. Employees scoring higher in Employability reported higher wellbeing (Bernhard-Oettel, De Cuyper, Berntson and Isaksson 2008). Indeed, literature shows more than one time that individuals with higher levels of employability report higher levels of psychological well-being (Briscoe et al., 2012), which is a proxy for Life Satisfaction. Logically, employable, individuals will be more easily able to adapt to the work that better fits their preferences. According to Ryff (1989) wellbeing and life satisfaction are treated as synonyms in the extant literature. Because of that, it is hypothesized here that employability will have a positive impact on Life Satisfaction because the possibility of choosing, in a job specific context, will bring satisfaction to the individual.
Hypotheses 4: Employability has a positive impact in Life Satisfaction.
In addition, aligned with all that was previously said, and although it is not the focus of this study, one expects also expects a positive spill-over from Work Satisfaction to Life Satisfaction and vice-versa so that both concepts are identically influenced as outcomes of Employability, which supports both hypothesis.
Van der Heijden et al (2009) reported differences in the outcomes of Employability under the moderation of age. Also Cheng and Chan (2008) present
resilient evidence, that is, strongly supported argumentation, that age has an impact in negative health outcomes among employees in a positive way, meaning that as individuals age they are more likely to endure negative health outcome. In other words that is to say that as people get older and therefore they are more fragile to diseases and take longer to recover from health issues. It is logical to say then, that there is an influence of age and one expects the relationship between employability and satisfaction to vary as a function of age so that it is going to positively related to both constructs of Work Satisfaction and Life Satisfaction, as being employable is exponentially more important at an older age than at a younger one. That is, older people have more difficulties to adapt to new situations, and therefore, losing a job will impact more negatively their life than if they were younger and more adaptable.
Hypothesis 1.1: Age moderates the relationship between Employability and Career Satisfaction so that this relationship is stronger for higher values of age.
Hypothesis 2.1: Age is a moderator of the relationship between Employability and Job Satisfaction so that this relationship is stronger for higher values of age.
Hypothesis 3.1: Age is a moderator of the relationship between Employability and Work Satisfaction so that this relationship is stronger for higher values of age.
Hypothesis 4.1: Age is a moderator of the relationship between Employability and Life Satisfaction so that this relationship is stronger for higher values of age.
Method Employability proxies
Due to its complex construction, Employability is many times "unfolded" in just one of its many components and only then analyzed, that is, often Employability is analyzed with the focus on a specific aspect of the concept. This happens because, has shown before, it is a very complex construction. For the purpose of this Meta-analysis the focus is on the following faces of Employability: (Overall) Career Adaptability, Protean Career Orientation, Boundaryless Career Orientation, Organizational Mobility Preference and Perceived Employability.
Although very similar and with a big overlapping, each of this variables focus on specific differences in the aspects of Employability and because of that they were cataloged as proxies of the bigger concept - Employability.
(Overall) Career adaptability
As pointed out earlier on, and in line with Hillage and Pollard (1998), Employability is the capability of moving across the labor market. This capability to adapt is also called of Career Adaptability by Savickas (1997) when referring to the readiness to cope with change, being it either in the work or the working conditions. To this ability, certain competencies, attitudes and behaviors enable job search opportunities and allowing one to better fit in preferred work contexts and work roles (Savickas, 1997, 2013; Duffy, 2010; Klehe et al, 2012; Tolentino et al., 2013).
Protean Career Orientation & Boundaryless career orientation
The Boundaryless Career Orientation is described by Arthur and Rousseau (1996) as the type of career that opposes the traditional organizational career. The latter is characterized by careers that are tied to organizations (usually large stable
organizations) with long lasting arrangements/contracts that are achieved through vertical coordination (Gabriel & Nasina, 2012). This type of career was more common in the past when it would be the norm for people to spend most of their career in the same company. Traditional careers are still common in less developed countries but it is less common in the developed western civilization. On the contrary, Boundaryless careers are those defined by being independent of organizations, social bounds, vocations and regulatory mechanisms (Arthur, 1994). This implies that Boundaryless Career Orientation is “not tied to a single organization, not represented by an orderly sequence, and marked by less vertical coordination and stability” (Briscoe & Hall, page 6, 2006).
The Protean Career Orientation takes the individual, and not organizations, as career actors that build on their own psychological assets and decision making (Hall, 1976). Empowering the individual to pursue a career under “one’s own terms” brings to Hall’s concept of Protean orientation an emphasis on the relational and “whole-life” career qualities and the control over it.
Summarizing it is possible to say that boundaryless career orientation has into consideration the many options presented and how to recognize and take the best of such situations in order to achieve success (Arthur, Inkson, & Pringle, 1999; DeFillippi & Arthur, 1996). The Protean career orientation is more focused in valuing self-focused approach to the career and use one’s own ideologies and values to drive that career (Briscoe & Hall, 2002).
After looking to both concepts of Boundaryless and Protean orientation separately, it is now important to combine the two as they are, as previously shown, very similar. Both concepts are of great use for the revising of different career orientations (Briscoe & Hall, 2006). According to Briscoe and Hall (2005), there is value
in combining the two concepts and from the overlapping of that combination many different profiles within examination of possible combinations can be identified. The combination of the two concepts is also said to be helpful for the reduction in confusion on the application of the concepts, making the study of new career forms more accessible. Therefore, the variables used for those concepts were combined with the objective of simplification.
Organizational mobility preference (OMP)
The Organizational Mobility Preference is a scale designed to score the strength of one’s wish to remain (or not) with either a single employer or a multitude of them. This scale is a result of 5 questions proposed by Briscoe and Hall (2005) and it is composed by reversed score items to questions like “I like the predictability that comes with working continuously for the same organization” and “If my organization provided lifetime employment, I would never desire to seek work in other organizations”. The use of this variable is valuable for the present meta-analysis because it is used to assess Boundaryless Career Orientation where the likelihood of one crossing organization boundaries can be expected to score high in Organizational Mobility Preference (Briscoe et al, 2006).
Perceived Employability
Perceived Employability is positively associated with global health and mental wellbeing (Berntson & Marklund, 2007), engagement and life satisfaction (De Cuyper et al., 2008), explaining the inclusion of this Employability concept to the study. It not only is expected to influence in some way satisfaction but it also is closely related to Employability as the term self-explains.
Career Satisfaction (Career Satisfaction & Professional Satisfaction)
Stretching the meaning of Career Satisfaction, one finds a logical overlapping on the meaning of Professional with this concept of Career Satisfaction. Literature shows that "professional satisfaction" is used the same way as the previous construction, being an integral part of it too. Olsen et al (1995) refer to Professional Satisfaction as a disaggregate form of Career Satisfaction, that is, Professional Satisfaction is part of and cannot be separated from Career Satisfaction.
That being said, one expects the same positive relationship between Employability and Professional Satisfaction as encountered between Employability and Career Satisfaction and because of that both variables were analyzed under Career Satisfaction’s umbrella.
Age
Retrieved from the information available in the studies used, age was used as a moderator. The value of age used corresponds to the average age of the participants of the sample.
Procedures
The development of a coding scheme to catalogue and in/exclusion criteria of the studies will be the first step. Primarily, the inclusion criteria includes the year of publication (>1980), language (English, Dutch, Italian, Portuguese, Serbo-Croatian and Spanish as those were the languages for which the involved team had proficiency), studies should be quantitative and quantify the measure of the relation between employability (or at least one its dimensions) and at least one predictor or outcome.
These measures might be correlation coefficients, Cohen's d or other effect sizes. The exclusion criteria discharges non-empirical and/or qualitative studies.
The systematic collection of both published and unpublished (which in the end did not happen for unpublished studies as we will see) primary studies from the employability literature allowed us to code the studies found. As part of this study specific characteristics, the search and documentation of relevant literature used the software Zotero as reference database. Search engines included the following: PsycINFO, ERIC, ProQuest and Web of Science. If time permits the computer-based literature search will expand for one or more of the following: LIBRIS, Management Contents, MedLine, Mental Health Abstracts, NIOSHTIC, Sociological Abstracts, Social SciSearch, Uncover, Metabus, Ortis, ISI Web of knowledge, Pubmed, PEDro, Cochrane Controlled Trials register or Embase.
To complement this search for articles, a search in relevant journals was done. Enlisted journals include: Academy of Management Journal, Human Relations, Journal of Applied Psychology, Journal of Vocational Behavior, Journal of Occupational Health Psychology, Journal of Occupational and Organizational Psychology, Journal of Organizational Behavior, Psychological Bulletin, Human Decision Processes, Decision science, Journal of applied behavioral science, Journal of international business studies, Journal of Management Studies, Management science, Organizational Behavior and Human decision processes, Social Science & Medicine, Stress Medicine, Work & Stress, Administrative Science Quarterly, Strategic Management journal.
After the primary identification of the available records, records that are duplicate were removed and the remaining ones went through the screening process. This screening process filtered more thoroughly and from it one expects to have the full-text
articles assessed for eligibility. The selection of studies ends with the list of studies that have all the requirements and will be included in the meta-analysis (Moher D. et al, 2009).
The list with the studies that are elected for the meta-analysis will favor an overview of the materials available and will grant the possibility to code and catalog those same studies with the assistance of the already created Coding Manual (Appendix 1). The use of such coding scheme gave us the ability to better compare the studies with each other and analyze possible overall effect size. By the end of this coding process, one has at disposal all the relevant information structured and organized for an easier and faster examination of variables and relationship between variables.
Reflecting on the data available at this stage and its constitution, one is able to run Meta-Analyses to test the hypotheses formulated above with the purpose of enlighten the general community.
The possibility of bringing together the research already done in the field of Employability will allow the scientific community to combine different studies for a better overview of results. This statistical technique one is about to apply, merges the multitude of studies with the objective of structure material from the mass of fragmented studies. This aggregation brings an increase of the number of observations and statistical power to the results and reduce some risks, namely, marginal error and erroneous conclusions from small studies. A larger population sample helps one to more likely produce meaningful data (Ioannidis, J. P. & Lau J., 1999; Walker et al, 2008).
The use of a meta-analysis is also a more thorough and careful approach comparing to the more common literature review approach as the last is more narrative and prone to the influence of personal interpretations of the researchers (Garg et al,
2008).
One is also aware of possible downsides of conducting such systematic review. The main threat is the publication bias which is the bias concerned with the problem on focus on only publishing studies with significant results, being the rest of the studies victims of “file drawer effect” (Jeffrey D. Scargle, 2000). To fight the problem of publication bias and the “file drawer effect”, the manual search of conference programs is also to be conduct with the objective of finding unpublished studies applicable to this Meta-Analysis. The conferences targeted are: Society for Industrial and Organizational Psychology (SIOP), Academy of Management Conference (AOM), The European Association of work and Organization Psychology (EAWOP), WAOP Conference, HRM Network Conference, ICAP Conference and The European Academy of Management (EURAM).
Study selection
PsycINFO, Business Source Premier, ERIC, Medline, Web of Science, Sociological Abstracts and Scopus were used to identify primary studies. From the aforementioned sources, on the 21st of March of 2016, 12.764 studies were retrieved with the use of the
query in Appendix 2 applied to the different search engines. 3.823 studies were duplicates which gave us 8.941 studies to start with. Two rounds of selection followed, bringing the total number of relevant studies back to 3.981 in the first round and 1360 in the end of the second round, from which 159 studies were duplicates not identified in the initial scan.
# Raw References 1st Round 2nd Round
PsycINFO 2.779 Business Source Premier 1.619 ERIC 1.179 Medline 649 Web of Science 2.571 Sociological abstracts 622 Scopus 3.345 Total 12.764 3981 1360 Total Deduplicated 8.941 1201
Table 1 - Search Engines Results.
Initially, the inclusion criteria used in the initial search were significantly looser than in the first and second of selection with the objective of not excluding any relevant article and to provide some additional training to those involved in the process of selection.
From the database created and after the coding of the articles, only 9 articles were selected. The small number of studies used in this research paper was very small compared to the poll of available studies because only studies clearly focusing in the relationship between Employability and Satisfaction were included. In addition, only studies with pertinent information for a positive participation in the testing the hypothesis presented earlier were included. This means that, as a matter of example, studies focus on Employability that had work or life satisfaction as precedent were not included. The final selection prior to the meta-analysis was done having into consideration the focus of analysis of each article. Common to the articles listed in the
Reference List with an asterisk is the analyze of the Employability outcome of work or life satisfaction, those being the ones included in this particular analysis.
Study # Sample size (N) Response Rate (R) Mean sample age SD of sample age Males in the sample Country 1 655 NA 36.4 10.3 56.1% Portugal 2 654 NA NA 55% China 3 639 NA 36 10.3 38.1% Spain 4 184 NA 21.1 2.5 54% USA 5 664 NA 20.23 2.56 NA Australia 6 202 73.19% 25.7 2.48 61.09% Ireland 7 1723 6.33% 46.67 11.31 44.2% Australia 8 2002 NA 41.99 8.61 48.40% Switzerland 9 463 NA 34 10 34.7 Belgium Total 7186
Table 2 – Background information of selected Studies.
The results reported in the different studies are diverse. Life Satisfaction has positive inter-correlations with Employability of 0.33 (p<0.01) and 0.13 (p<0.01) from study no.5 and study no. 9, respectively. As a proxy of Employability, Career Adaptability measures report positive inter-correlations between 0.24 and 0.37 (p<0.01) in study no. 4 and a bit more in study no.7 with an inter-correlation of 0.4 (p<0.01) but the
correlations in study no.2 are of only 0.63 (p<0.001) for the same relationship with Satisfaction. Study no.6 reports a negative effect of Employability in both Job Satisfaction and Career Satisfaction, while studies no.7 and no.8 report positive effects of Employability in Career and Job Satisfaction, respectively.
Meta-Analysis
In order to run the Meta-analysis in the SPSS, the information collected from the studies included in this analysis was extracted from the initial file created from during the Selection Process to and Excel file. In that Excel file, information was double checked and organized in order to be used in the SPSS sheet. The data organized included: number of study (StudyID), size of sample (N), correlation of Life Satisfaction with the proxy of Employability (r), reliability of Employability (RelEmp), reliability of Life Satisfaction (RelOth) and mean age of the sample (AGE). The correlation coefficients (r) represent unadjusted measures of the effect size.
After extracting the information mentioned above from the Excel Coding File, other variables (e.g. Minimum Effect Size, Maximum Effect Size and Weighted SD) needed to be computed using SPSS’ Syntax to make the data ready for further analysis. These variables are important because the data used was extracted from different studies. This studies have different effect sizes that might come from different features, sample sizes variation, different techniques applied and studies have different reliabilities.
Starting with the reliability adjustment, one expects to correct for unreliability and to do so one applies a correction on the raw correlation coefficients. Then, and because Correlation as a measure of effect size is not suitable in its raw form due to a problematic standard error formula, the use of Fisher’s Zr transformation allows us to stabilize the variance. To finish the preparation of all the variables we need, each effect size was put in its correct proportion regarding its weight on the overall meta-analysis. This is important because the studies have different sample sizes, therefore larger studies should have more weight in the analyses than smaller studies because bigger
studies are assumed to be more “precise” estimations of the population’s effect size than the smaller studies. In this analysis, each effect size is weight by the inverse variance.
After the previous transformations and adjustments, it is time to run the Meta-analysis with the help of Professor David B. Wilson’s macros for meta-Meta-analysis to test the hypotheses. In order to run the macro one needs to start by choosing the location of the file and then adjusting the formulas so that the formulas match the names in our dataset. This macro will perform basic central tendency statistics like mean effect size, confidence intervals, homogeneity test and both fixed and random effect models.
Results
This study aims to analyze the relationship between Employability and both Work and Life Satisfaction with age as a moderator. Variables and the correlations were extracted from the asterisked entries in the Reference List articles were Employability (or one of its proxies) were correlated with Work and/or Life Satisfaction. Within those proxies one could find Perceived Employability, Career Adaptability, Protean career orientation and Boundaryless career orientation. Both Work and Life Satisfaction variable was always found as an outcome of Employability.
Starting with Hypotheses 1 (Employability has a positive impact in Career
Satisfaction), data was introduced in SPSS as mentioned in the previous section and the
Table 3 – Employability and Career Satisfaction Results.
In the results for this first hypotheses one can see in the distribution description the number of effect sizes (K), one can see that 4 studies were used, with a minimum effect size of -0.262, a maximum effect size of 0.673 yielding a weighted standard deviation of 0.271.
The previous Table also shows the results of the application of the Fixed and Random Effects Model. One can find the results of the fixed effects model with the assumption that studies share a common “true” effect (Kelley & Preacher, 2012). Nevertheless, it does not make much sense to focus on the fixed-effects as in that case all the studies included in this analysis are assumed to be functionally identical, which
Meta-Analytic Results Distribution Description
K Min ES Max ES Wghtd SD
4.000 -.262 .673 .271
Fixed & Random Effects Model
Mean ES 95% Conf. Interval SE Z P Fixed .3664 .3319 .4009 .0176 20.7987 .0000 Random .2241 -.1111 .5593 .1710 1.3106 .1900
Random Effects Variance Component
v= 0.114845
Homogeneity Analysis
Q df p
might not be true. Also, the Fixed Effect Effects are not useful for generalizations to other population as it only gives the effect size for the identified population. Therefore, the Random Effects are of more use than Fixed Effects. Looking for them is better in our meta-analysis because they use the accumulated data from our poll of different studies and instead of assuming a common effect size as a result of exactly identical processes, Random Effects allow one to generalize the results to a broader set of situations/scenarios. Assuming that with different countries being the focus of different studies (only the Netherlands have two studies) and therefore zero overlapping, it is possible to say that the sample is not narrowly defined and the results can be extrapolated to other populations. In other words, in the random effects model one assumes that this “true” effect in the studies varies among studies and the summary effect is the weighted average of the effects reported in the different studies. It tends to be the result of a more conservative estimation but can be generalized beyond the studies/samples used in this model.
On the Homogeneity Analysis one can observe that Q is p<0.01 and because of that we reject the null hypothesis of homogeneity. We use random effects model because Q is significant and we can assume that the excess variability across effect sizes derives from random differences across studies as it exceeds what would be expected based on sampling error alone. This means that the differences between effect sizes have additional source rather than subject-level sampling error, which is an indication of possible existence of moderators. Next we will try to see if some of this excess variability is caused by the moderator effect of age.
Being non-significant, Hypotheses 1 (Employability has a positive impact in
Meta-Analytic Results Descriptives Mean ES R-Square k .2221 .0969 4.0000 Homogeneity Analysis Q df p Model .2175 1.0000 .6410 Residual 2.0261 2.0000 .3631 Total 2.2435 3.0000 .5234 Regression Coefficients
B SE 95% Conf. Interval Z P Beta Constant -.2515 1.0446 -2.2990 1.7959 -.2408 .8097 .0000 AGE .0138 .0297 -.0443 .0720 .4663 .6410 .3113
Restricted Maximum Likelihood Random Effects Variance Component
v = .23662 se(v) = .16884
Table 4 – Meta-analytic results for age as a moderator of the relationship between employability and career satisfaction.
Running the aforementioned Wilson’s macro we get the table above (Table 4 – Meta-analytic results for age as a moderator of the relationship between employability and career satisfaction). Observing under Homogeneity one can see that the model is not significant and that only 0.2175 of 2.2435 is explained by the model, not being significant which show the lack of relevance of the model. Age is not a significant moderator of the relationship between employability and career satisfaction, therefore
the Hypotheses 1.1 (Age moderates the relationship between Employability and Career
Satisfaction so that this relationship is stronger for higher values of age) is not
supported.
Moving to the test of hypotheses 2 (Employability has a positive impact in Job
Satisfaction), data was introduced in SPSS with the same previously mentioned process
and the results are as presented in the following Table 5 – Employability and Job Satisfaction Results:
Table 5 – Employability and Job Satisfaction Results.
In the results for this second hypotheses one can see in the distribution description the number of effect sizes (K), 3 studies were used, with a minimum effect
Meta-Analytic Results Distribution Description
K Min ES Max ES Wghtd SD
3.000 -.430 .056 .159
Fixed & Random Effects Model
Mean ES 95% Conf. Interval SE Z P Fixed -.0480 -.1002 .0042 .0267 -1.8013 .0717 Random -.1236 -.3550 .1078 .1181 -1.0470 .2951
Random Effects Variance Component
v = .039096
Homogeneity Analysis
Q df p
size of -0.430, a maximum effect size of 0.056 and a weighted standard deviation of 0.159.
As explained earlier, only Random Effects are of relevance and in this care one can observe a negative effect size of -0.1236 and that it is not significant (P=0.2951).
On the Homogeneity Analysis one can observe that the value of Q (35.5291) is p<0.01 and because of that we reject the null hypothesis of homogeneity. As before, we use random effects model because Q is significant and one can assume that the excess variability across effect sizes derives from random differences across as it exceeds what would be expected based on sampling error alone. As before, next one will try to see if some of this excess variability is caused by the moderator effect of age and the proposed Hypotheses 2.1 (Age is a moderator of the relationship between Employability and Job
Table 6 – Meta-analytic results for age as a moderator of the relationship between employability and Job Satisfaction.
Running Wilson’s macro we get the previous Table 6. Under Homogeneity one can see that the model is non-significant, therefore the hypothesis 2.1 is not supported. Testing hypotheses 3 (Employability has a positive impact in Work Satisfaction), data was introduced in SPSS the same way as for the test of the previous hypotheses and the results are as follows in Table7:
Meta-Analytic Results Descriptives Mean ES R-Square k -.1126 .8567 3.0000 Homogeneity Analysis Q df p Model 6.6460 1.0000 .0099 Residual 1.1118 1.0000 .2917 Total 7.7578 2.0000 .0207 Regression Coefficients
B SE 95% Conf. Interval Z P Beta Constant -1.5046 .5450 -2.5728 -.4364 -2.7608 .0058 .0000 AGE .0429 .0166 .0103 .0755 2.5780 .0099 .9256
Restricted Maximum Likelihood Random Effects Variance Component
v = .01371 se(v) = .01330
Meta-Analytic Results Distribution Description
K Min ES Max ES Wghtd SD
6.000 -.430 .673 .305
Fixed & Random Effects Model
Mean ES 95% Conf. Interval SE Z P Fixed .2918 .2607 .3228 .0159 18.4024 .0000 Random .0977 -.1883 .3837 .1459 .6696 .5031
Random Effects Variance Component
v = .125204
Homogeneity Analysis
Q df p
370.9723 5.0000 .0000
Table 7 – Employability and Job Satisfaction Results.
In the results for this third hypotheses one can see in the distribution description he number of effect sizes (K), one can see that 6 studies were used, with a minimum effect size of -0.430, a maximum effect size of 0.056 and a weighted standard deviation of 0.305.
As explained earlier, only Random Effects are of relevance and in this care one can observe a positive effect size of 0.0977 and that it is not significant (P=0.5031).
On the Homogeneity Analysis one can observe that the value of Q (370.9723) is p<0.01 and because of that we reject the null hypothesis of homogeneity. As before, we use random effects model because Q is significant and one can assume that the excess variability across effect sizes derives from random differences across studies as it
exceeds what would be expected based on sampling error alone. Again, next one will try to see if some of this excess variability is caused by the moderator effect of age and test hypotheses 3.1 (Age is a moderator of the relationship between Employability and Work
Satisfaction so that this relationship is stronger for higher values of age).
Table 8 – Meta-analytic results for age as a moderator of the relationship between employability and Work Satisfaction. Meta-Analytic Results Descriptives Mean ES R-Square k .0967 .2394 6.0000 Homogeneity Analysis Q df p Model 1.2653 1.0000 .2606 Residual 4.0199 4.0000 .4033 Total 5.2852 5.0000 .3821 Regression Coefficients
B SE 95% Conf. Interval Z P Beta Constant -.7247 .7488 -2.1923 .7429 -.9679 .3331 .0000 AGE .0250 .0223 -.0186 .0687 1.1249 .2606 .4893
Restricted Maximum Likelihood Random Effects Variance Component
v = .16178 se(v) = .09488
Running Wilson’s macro we get the previous table 8 -Meta-analytic results for age as a moderator of the relationship between employability and Work Satisfaction. Under Homogeneity one can see that the model is not significant and therefore hypothesis 3.1 is not supported.
The next table (Table 9) contains the first output of our SPSS analyzes for the Hypotheses 4 (Employability has a positive impact in Life Satisfaction) and, starting with the distribution description and the number of effect sizes (K), one can see that 6 studies were used, with a minimum effect size of -0.401, a maximum effect size of 0.437 and a weighted standard deviation of 0.314.
Table 9 – Employability and Work Satisfaction Results.
Meta-Analytic Results Distribution Description
K Min ES Max ES Wghtd SD
6.000 -.401 .437 .314
Fixed & Random Effects Model
Mean ES 95% Conf. Interval SE Z P Fixed -.0666 -.0952 -.0380 .0146 -4.5593 .0000 Random .0994 -.1923 .3911 .1488 .6679 .5042
Random Effects Variance Component
v = .130840
Homogeneity Analysis
Q df p
On the Homogeneity Analysis one can observe that Q is p<0.01 and because of that we reject the null hypothesis of homogeneity. This means that the differences between effect sizes have additional source rather than subject-level sampling error, which, like we previously analyzed, is an indication of possible existence of moderators. Next we will try to see if some of this excess variability is caused by the moderator effect of age and therefore test hypothesis 4.1 (Age is a moderator of the relationship between
Employability and Life Satisfaction so that this relationship is stronger for higher values of age).
Table 10 – Meta-analytic results for age as a moderator of the relationship between employability and Life Satisfaction.
Running Wilson’s macro we get the results presented above in Table 10. Observing under Homogeneity one can see that the model is significant and that with 23.7466 it represents a big part of the total 27.5742. This means that only the small part of 3.8276 of 27.5742 is not explained by the moderation of age. Under Descriptives one can see than 86 % of variance explained by the regression “model” and with it we see
Meta-Analytic Results Descriptives Mean ES R-Square k .0871 .8612 6.0000 Homogeneity Analysis Q df p Model 23.7466 1.0000 .0000 Residual 3.8276 4.0000 .4298 Total 27.5742 5.0000 .0000 Regression Coefficients
B SE 95% Conf. Interval Z P Beta Constant 1.2014 .2345 .7418 1.6610 5.1238 .0000 .0000 AGE -.0341 .0070 -.0479 -.0204 -4.8730 .0000 -.9280
Restricted Maximum Likelihood Random Effects Variance Component v = .01419
that the excess variability is caused by the moderator effect of age that is significant and of great importance for the model, confirming Hypothesis 4.1.
Discussion Summary of evidence
Summarizing the evidence, one can say that Hypothesis 1 (Employability is
positively related to Career Satisfaction) pertaining to the positive relationship between
Employability and Career Satisfaction, was not supported, nor was age as a moderator of this relationship proposed by hypothesis 1.1 (Age is a moderator of the relationship
between Employability and Job Satisfaction so that this relationship is stronger for higher values of age).
For hypothesis 2 (Employability is positively related to Job Satisfaction), one sees that is not supported because the relationship between Employability and Job Satisfaction is non-significant and happens to be negative, which is contrary to what was hypothesized. The moderation of Age helps to explain this relationship, although not being significant which refutes hypotheses 2.1 (Age is a moderator of the relationship
between Employability and Job Satisfaction so that this relationship is stronger for higher values of age).
The hypotheses 3 (Employability is positively related to Work Satisfaction) is also refused by not being significant. The same with the sub hypotheses 3.1 (Age is a
moderator of the relationship between Employability and Work Satisfaction so that this relationship is stronger for higher values of age) where indeed Age appears to have some
important impact explaining more than 20% of the variation of results in this relationship but in fact it is also non-significant.
Lastly, the relationship tested in hypotheses 4 (Employability has a positive
impact in Life Satisfaction) is not supported because it is non-significant. On the contrary, the sub hypotheses 4.1 (Age is a moderator of the relationship between
Employability and Life Satisfaction so that this relationship is stronger for higher values of age) is significant and therefore, supported.
Limitations
There are also possible downsides to conducting a systematic review such as the one reported here. The main threat is publication bias (Rothstein, Sutton and Borenstein, 2005) which is the bias concerned with the problem of only focusing on published studies with significant results, with the rest of studies being victim of the so called “file drawer effect” (Scargle, 2000). Unfortunately, time constraints did not allow our search to also target unpublished studies.
Like it happens with most meta-analysis, by using secondary data, one is not in full control of the quality of the processes used in the studies included, neither is one able to improve the quality of these studies.
The use of only cross-sectional studies does not allows any casual inferences to be made from our results. Although this is a limitation of this study, it also highlights the opportunity for future research to overcome this limitation by using longitudinal research designs with data collection at multiple time points, tracking participant’s career choices, motivations and consequences of those choices over time. The same should be done to assess the impact in the individual’s wellbeing. Only then causal interpretations can be established.
All studies included were conducted after the last Economic Crisis that started in 2007 that lead to the Great Recession of 2008–2012 (Eigner & Umlauft, 2015) which may have an influence on the data collection, especially in Portugal, Ireland and Spain where the context of crisis and high unemployment rates had deep social implications.
Regarding selection and during the initial gathering of studies, it was not possible to conduct an extensive review to the different selection results to make sure all the persons involved were applying the same processes and, therefore, get a consistent selection. Also during this initial phase, the counting of the studies in the end of each round of selection did not correspond with total accuracy what was expected to be the number of studies by the sum of the individual number of studies included of those involved. This might be because of the deduplications applied or for something related with the software used. On the negative side, the reasons that lead to that are not clear. On the bright side, the deviation was smaller than 10% and it assumed not to have a great impact in the forthcoming steps of this study. For future meta-analysis a more strict approach should be applied in order to track the number of studies from one round to the next.
During the next phase, the coding of the previously selected studies, also because of time constraints, a trial/test of performance was to be conducted but did not happen. The idea behind this trial was to align the different interpretations of what information should be retrieved from the studies and how to properly code it. At the same time of the final coding, an overlap system was planned but not formally checked for. The idea was to overlap some of the coding for a future analyze. In addition to the manual check overlapping check, also the computation for interrater reliability was planned but did not happen. Ideally the results of that analysis would show little or no difference among different coders, and allow for the resolution of differences in interpretation.
Although a big database of coded studies and information about Employability was created, not so many studies were completely fit for this particular analysis. Due to
insufficiencies in the time framework, from the studies selected to be coded, a portion was not coded and was set aside to be coded in posterior time. Then, if not all the studies selected where coded by the time this meta-analysis was conducted, as mentioned before, logically, some relevant studies might be in that pool of parked studies waiting to be coded. Furthermore, it means that there might be studies that will be coded in the future with potential to be part of this analysis, leaving space for further research not only for the inclusion of those missing studies but also for the development of the proposed model.
The main limitation of this study was, indeed, time. Any further time spent on the analysis of relationship between Employability and Work and Life Satisfaction is of great use as it is of such great importance in the wellbeing of the individual and as the sum of them, also the wellbeing of the community.
Some suggestions for future research can be made too. For example, the use of Job Success, Career Success, Professional Success or even Life Success was, sometimes, not that different from the meaning of the variables included in this meta-analysis: Career Satisfaction, Job Satisfaction, Professional Satisfaction and Life Satisfaction. Then, the use of Success might have been used with the same intention as Satisfaction across different studies. There is, therefore, the possibility to analyze the difference between Success and Satisfaction and to support or refute the possibility of those two names for the same thing after all.
In addition, aligned with all that was previously said, and although it is not the focus of this study, one also expects a positive spill-over from Work Satisfaction to Life Satisfaction and vice-versa so that both concepts are identically influenced as outcomes
of Employability. This spill-over has the potential to be the focus of new research so that it can be clarified which one is the cause of the other.
Conclusions
Employability will, in the future, continue to be in the spotlight regarding research as individuals seek to gain control over their career. Having a job still has a relevant central position in life. The same will happen with Satisfaction while people try to balance their wellbeing with the fulfillment of employer expectations.
After putting together a process of selection of studies that, one way or another, focused in employability, both in the outcomes or the predictors, and after coding the selected studies into a database, the consequences of Employability for various forms of Satisfaction were explored. By the end of this meta-analysis it was not possible to conclude that Career and Job Satisfaction are indeed part of the outcomes of Employability due to lack of evidence but they do represent a logical outcome that deserves further exploration.
The major contribution of this paper to research is related with the evidence that age moderated a large part of the relationship when talking about the impact of Employability in Life Satisfaction. With this results, it is safe to say that age should have a more prominent focus on the Employability field, if not in research in general. One should avoid doing research with age heterogeneous samples as it will result in misleading findings. Age should be more commonly tested as moderator and not simply regarded as sample characteristics.
