“Bevlogenheid in het middelbaar beroepsonderwijs” : how is work engagement affected by the challenging work situation in the MBO education sector : what mediating effect has job crafting on this relationship and how doe

(1)

“Bevlogenheid in het middelbaar beroepsonderwijs”

How is work engagement affected by the challenging work situation in the MBO education sector? What mediating effect has job crafting on this relationship and how does locus of control moderate this mediating effect?

MSc in Business Administration: Leadership and Management track MASTERS THESIS

Student: J. (Jurrien) Doelens Student number: 11156392

Supervisor: dr. A.H.B. (Annebel) de Hoogh

(2)

Statement of originality

This document is written by student Jurrien Doelens 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.

(3)

List of tables and figures

Table 1. Correlation matrix 27

Table 2. Hierarchical regression model of work engagement 28

Table 3. Mediation total, direct and indirect effects 30

Figure 1. Research conceptual model 20

Figure 2. Heat map participants 21

Figure 3. Research mediation model 29

(6)

Acknowledgements

I would like to thank my supervisor Annebel de Hoogh for all the support, and constructive feedback throughout the research process. I would also like to thank the students in my study group for their support and advice throughout the writing process. Last but not least, I would like to thank my wife Sandra Linders for her support not only during the thesis process but throughout the entire study period.

(7)

Abstract

Engagement and Retention of employees is difficult in all educational sectors in the

Netherlands. In the secondary vocational educational sector “middelbaar beroepsonderwijs” the high degree of complexity and diversity of the working environment adds to this difficulty. The dynamic work environment and high pace of change make this sector a challenging work situation for its employees. Work engagement is a construct known for its positive

organizational outcomes. Work engagement could contribute to employee retention and have other positive organizational benefits for educational institutions. Despite these potential positive outcomes little research has been done on the antecedents of work engagement on the employee level. This research focusses on the employees’ perceptions of these antecedents of work engagement in the educational sector. Previous research suggests that job crafting could be a mediation factor and locus of control could be a moderating factor in the relationship between the challenging work situation and work engagement in the educational sector. If so, this knowledge could be utilized by educational institutions in human resource instruments to engage and retain employees and benefit from the other positive outcomes from work

engagement. Current research adopted a quantitative approach, using a survey with 226 employees of different middelbaar beroepsonderwijs institutions. The results indicated that challenging work situation, locus of control and job crafting are statistically significant predictors of the perceived work engagement. No statistical support was found for the moderating role of locus of control on the relationship between challenging work situation and job crafting.

Support was found for the partial mediating role of job crafting in the relationship between challenging work situation and work engagement. Support was also found for the indirect effect of job crafting, meaning the stronger the perception of challenging work situation the higher work engagement will be due to their increased job crafting.

(8)

1. Introduction

In this chapter current research and the research question will be introduced.

1.1 Introduction

The MBO education sector delivers secondary vocational education and is a substantial educational sector in the Netherlands. Its (social) assignment is to provide employers ample qualified employees through offering vocational training in education programs. The MBO sector is characterized by a high degree of diversity and complexity (Waslander, 2017). The sector can also be described as dynamic (De Hoogh et al, 2005; Duncan, 1972) with a high pace of change (Ayyagari et al, 2011) this makes the MBO sector a challenging work situation. In this work situation, it is difficult to engage employees and retain employees (Rinnooy Kan et al, 2007; Vermeulen, van der Aa, 2008). Also, in international perspective the engagement and retention of public school teachers is of continuing concern (Chapman, 1984; Weiss, 1999; Rots et al, 2007).

Work engagement is seen as a critical antecedent of various positive organizational outcomes such as citizenship behavior and employee productivity (Akhtar, Boustani, Tsivrikos, Chamorro-Premuzic, 2015). There are studies that show that work engagement is positively related to task performance, contextual performance, and active learning (Bakker, Demerouti, Brummelhuis, 2012). Work engagement also influences productivity and performance (Saks, 2006). Evidence suggests that in the Netherlands up to 80% of employees are not engaged (Gallup Employee Engagement survey, 2011). So, there is a need to better understand how and why individuals become engaged with their work in order to improve employee well-being and organizational performance (Robertson & Cooper, 2010). This need for further understanding is also present with regard to employees in MBO institutions, so teachers, managers and support staff. From previous research (Schaufeli, 2004; Hakanen, Bakker, Schaufeli, 2006) we know that work engagement is an important construct in satisfaction and retention of employees.

We also know that job crafting is an important antecedent of engagement (Bakker, Tims, Derks, 2010; Petrou et al, 2012; Bakker, Tims, Derks, 2012). Job crafting involves changing the

cognitive-, task- and/or relational boundaries of the job to shape interactions and relationships with others at work. These altered tasks and relational configurations change the design and social environment of the job, which in turn, alters work meanings and work identity

(Wrzesniewski, Dutton, 2001). Research suggest that job crafting is especially relevant to work engagement in the context of a challenging work situation (Tims, Bakker, 2010; Duncan, 1972; Wrzesniewski, Dutton, 2001; Shamir and Howell, 1999). Many changes give employees

opportunities for change and if employees use this through job crafting their work engagement will increase.

Earlier research also suggests that there are personality characteristics (Strümpfer, 1990) that can influence employees in the extent to which they take use of opportunities. Locus of control

(9)

is a personality characteristic that influences a person’s perceptions of the effect of their actions. Internal locus of control is a conviction of positive results of his/her actions. And external locus of control the negative opposite. In this case a person is convinced that they are dependent on others and their own action are less relevant. So, internal locus of control could positively influence an employee’s engagement in job crafting in reaction to challenging work situations and external locus of control could do the opposite.

Thus, work engagement is an important outcome in the satisfaction, performance and retention of employees for the MBO sector (Schaufeli, 2004; Saks, 2006, Huang, Waxman, 2008). This research will therefore focus on the antecedents of work engagement. The research will give MBO organizations valuable insights in how work engagement is affected by the context of the challenging work situation and the mediating effect of employee behavior job crafting in this relationship. Finally, the moderating effect of the personality characteristic locus of control on relationship between job crafting and the context of challenging work situation will be explored. This research could give MBO organizations further insights to focus the management attention. The following examples illustrate this. What personality characteristics (locus of control) do employers need to look for in hiring new employees? Or what positive employee behaviors (job crafting) do employers need to stimulate and facilitate further? Employers could use

instruments and interventions likes coaching or adjustment of organizational forms or even adjustment of governance philosophy. Despite the possible importance of the above-mentioned research suggestions, this has not been researched in de MBO sector.

1.2 Research question

How is work engagement affected by the challenging work situation in the MBO education sector? What mediating effect has job crafting on this relationship and how does locus of control moderate this mediating effect?

(10)

2. Theoretical framework

In this chapter the theoretical framework will be discussed.

2.1 Challenging work situation

In this section first the MBO sector will be introduced. Then the changes in the MBO sector will be discussed. After that the dynamic aspects of the MBO sector will be elaborated on. Also, the pace of change will be explored. Then the challenging wok situation will be addressed. And finally, the difficulty to engage and retain employees in the MBO sector will be discussed.

2.1.1 The MBO sector

The MBO education sector delivers secondary vocational education and is a substantial

educational sector in the Netherlands. Its (social) assignment is to contribute to the provision of ample qualified workforce for employers. The MBO sector executes this assignment through offering vocational training in education programs. As a secondary education sector, it receives its students from the primary education sector, but also from the current workforce (second change education) that seeks to retain or expand their knowledge and skills. After completing their education students either enter the workforce or resume their education in the higher education sector.

2.1.2 Changes in the MBO sector

Change is not new to the MBO sector. This educational sector is faced with a lot of (educational) changes. One of the actors in change is the Dutch government, but there are numerous other stakeholders on national, regional and institutional level that influence and initiate changes in this educational sector (Westerhuis, 2017). Some examples of changes are the following. In a major reform in 1996 the different ROC (“regional opleidingscentrum”) institutions where formed in the MBO sector. This was executed through forced mergers of smaller vocational schools through the Dutch governmental legislation “Wet Educatie Beroepsonderwijs (1996)”. Another major educational reform that was due to be implemented in 2011 was “Competentie Gericht Onderwijs (2011)” where the MBO sector was urged to connect more with the

occupational sectors through adaptation of the course curriculum that has to be aligned with competencies that are needed in the future jobs. Waslander (2017) concludes that the MBO sector is characterized by high degree of diversity and complexity. Waslander (2017) also states that the MBO sector is currently involved in a number of big changes, one of the most far-reaching current changes that employees in the MBO sector is the change in diploma end-terms. In this change the goal is to reduce the number of end-terms, but also the structure of the courses is changed. This entails that the courses will consist of 3 parts, a base part, a profile part and an elective part. These examples show that change is a constant in the MBO sector. And due to the impact and frequency these changes, employees (teachers, managers and support staff) in this sector can perceive the work situation in the MBO as challenging. In the next paragraphs we will explore the characteristics of the MBO sector, especially the dynamic work environment and the pace of change in the MBO sector.

(11)

2.1.3 MBO sector as a dynamic work environment

Duncan (1972) defined the characteristics of organizational environments and the relationship with perceived environmental uncertainty. Specifically, he focused on the characteristics of the environment that make that members of the organization experience uncertainty in decision making (Duncan, 1972). In his research, Duncan (1972) introduces two dimensions:

static-dynamic and simple-complex environment dimension. The static-static-dynamic dimension is analyzed by asking respondents how often important factors changed, that they identified as being important in decision making in their internal and/or external environment. Also, the frequency in which decision makers take new internal and external factors into consideration was

explored. Duncan (1972) concluded that the highest score in perceived environmental

uncertainty was perceived in relation to the dynamic-complex environment. In this context the highest amount of uncertainty in decision making was experienced. Duncan (1972) also

concluded that the nature of the organization’s environment (static-dynamic dimension) rather than the kind of organization (simple-complex dimension) is most important in explaining the degree of uncertainty experienced in decision making. With the number of important changes in the MBO sector as established in the previous paragraph the MBO sector can be perceived as dynamic. In current research the aspect of perceived dynamic work environment is an

important aspect to explore further along with pace of change.

2.1.4 Pace of change in the MBO sector

Ayyagari et al (2011) defined dynamism as pace of change and the impact this has on changes in work processes and procedures and daily routine of employees. Meaning the changes as a result of the changing demands of the stakeholders, like the Dutch government. And the imposition of changes on the educational sector as this could have an effect on the employees. Especially the effects with regard to their work processes and procedures and daily routine. Shamir and Howell (1999) argue that these changing environments are characterized by few situational cues, few constraints and few re-enforcers to guide behavior. These dynamic and uncertain environments require new interpretations, novel responses, and different levels of effort and investment. They provide high latitude of decision discretion and ample opportunities to demonstrate leadership. This makes pace of change relevant to include in current research.

2.1.5 Challenging work situation

With the relevance of the previous paragraphs to the MBO sector in mind. The challenging work situation in the MBO can be viewed as (employee) perceived dynamic work environment or environmental dynamism. The challenging environment can be further operationalized as a dynamic work environment with a high pace of change. A dynamic work environment gives a high degree in uncertainty with regard to decision making and this degree of uncertainty gives urgency for employees in the MBO sector to use coping strategies like actively restructure their job design, through job crafting. Otherwise employees are faced with uncertainty and

(12)

2.1.6 Difficulty to engage and retain employees in the MBO sector

Based on the above paragraphs the MBO sector is characterized by a high degree of diversity and complexity (Waslander, 2017). The sector can also be described as dynamic (De Hoogh et al, 2005; Duncan, 1972) with a high pace of change (Ayyagari et al, 2011) this makes the MBO sector a challenging work situation for the employees. In this work situation, it is difficult to engage employees and retain employees (Rinnooy Kan et al, 2007; Vermeulen, van der Aa, 2008). Also, in international perspective the engagement and retention of public school teachers is of continuing concern (Chapman, 1984; Weiss, 1999; Rots et al, 2007). Thus, interventions are needed to solve the issues of engagement and retention.

2.2 Work engagement

In this section first, the emergence of work engagement as broadening of burn-out research is discussed. Then the perception of work engagement as a different concept is explored. Also, the operationalization of work engagement in de dimensions vigor, dedication and absorption is elaborated on. Finally, the positive aspects of work engagement are discussed.

2.2.1 Work engagement as broadening of burn-out research

Schaufeli et al (2002) formulates the emerging of the focus on job engagement research as broadening of the traditional concept and scope of burn-out research. Work or job engagement at its emergence was seen as the opposite or antipode of burn-out. Alternatively burn-out was seen as an erosion of engagement with the job (Schaufeli et al, 2002). Earlier attempts to broaden the scope of burn-out research and formulation of antipodes where made by Kahn (1992) and Maslach, Leiter (1997). Kahn (1992) presented a theoretical model of ‘psychological presence’ or ‘to be fully there’, a concept that emerged from role theory and is defined as an experiential state that accompanies ‘personally engaging behaviors’ that involve the channeling of personal energies into physical, cognitive, and emotional labors. Kahn (1992) however, did not present an operationalization of the construct. As cited by Schaufeli et al (2002) Maslach and Jackson developed an instrument, the Maslach-Burnout Inventory in 1981, to measure the three factors of burn out: emotional exhaustion, depersonalization/cynicism and reduced efficacy. Schaufeli et al (2002) cited that Maslach, Leiter (1997) defined three factors of

engagement as the direct opposites of the burn-out factors: energy, involvement and efficacy. So, they considered it the same concept: burnout, with factors on opposite sides of the

continuum. The perception of engagement as the antipode of burn-out still influences research with regard to the positive outcomes of engagement, as we will discuss in the last paragraph of this section, therefore it was relevant to discuss its emergence and the perception of the construct work engagement at that time. But after its emergence the perception of the construct developed further, as we will discuss in the next paragraph.

2.2.2 Work engagement as a different concept

Schaufeli et al (2002) considered burn-out and engagement to be different concepts. This called for an operationalization of the concept engagement. In this operationalization, vigor,

(13)

engagement that are direct opposites (exhaustion vs. vigor and cynicism vs. dedication), reduced efficacy and absorption are not each other’s direct opposites, they are conceptually distinct aspects that are not the end points of some underlying continuum. So, Schaufeli et al (2002) feel that conceptually speaking engagement is the positive antithesis of burnout. They acknowledge that the measurement of both concepts, and therefore its structure, differs. As a consequence, engagement is operationalized in its own right. In the next paragraph we will explore the dimension of work engagement further.

2.2.3 Work engagement dimensions: vigor, dedication and absorption

Schaufeli et al (2002) operationalize the definition of engagement in the elements vigor,

dedication and absorption as follows. Vigor is characterized by high levels of energy and mental resilience while working, the willingness to invest effort in one’s work, and persistence even in the face of difficulties. Dedication is characterized by a sense of significance, enthusiasm, inspiration, pride, and challenge. Absorption is characterized by being fully concentrated and deeply engrossed in one’s work, whereby time passes quickly and one has difficulties with detaching oneself from work. These dimensions will be included in current research. In short, work engagement can be defined as “a fulfilling work-related state of mind that is characterized by vigor, dedication and absorption” (Schaufeli, Bakker & Salanova, 2006). In the next paragraph we will discuss the positive aspects of work engagement.

2.2.4 Positive aspects of engagement

In paragraph we will discuss the positive aspects of work engagement and reason why it is worthwhile to pursue it for an organization. This will also explain why it is worthwhile to explore its antecedents. Two lines of reasoning will be explored, first from the perspective of (avoidance of) negative aspects derived from the opposite situation of work engagement. And the second line of reasoning from the perspective of positive aspects of work engagement.

The first line of reasoning is that the opposite of the positive state of engagement of employees, is in itself a negative state with negative aspects and outcome. And thus, needs to be avoided. We refer to the first paragraph of this section where the emergence of the concept work

engagement was originally seen as the opposite of burn-out (Kahn, 1992; Maslach, Leiter, 1997). And part of the operationalize of burn-out is still the direct opposite of work engagement

(exhaustion vs. vigor and cynicism vs. dedication) (Schaufeli et al, 2002; Maslach, Leiter, 1997). And avoiding the aspects of burn-out (emotional exhaustion, depersonalization/cynicism and reduced efficacy) is in itself a positive situation, and therefore a positive aspect of engagement. Below examples with regard to stress and personal efficacy are explored.

An example with regard to stress is shown in research that states that work-related stress is a large problem related to the working environment in de the European union (Skakon, Nielsen, Borg & Guzman, 2010). The link between working conditions and employee stress and affective well-being has been established in a wide variety of research (Cox, Griffiths, & Rial-Gonza´lez, 2000; De Jonge, Bosma, Peter, & Siegrist, 2000; De Lange, Taris, Kompier, Houtman, & Bongers, 2004; Grawitch, Gottschalk, & Munz, 2007). Work-related stress may be prevented or

(14)

underload, by improving social support (Levi & Levi, 2002). Studies suggest that between 50% and 60% of lost working days have some link with work-related stress. This represents a huge cost in terms of both human distress and impaired economic performance (Levi & Levi, 2002). Evidence suggests that in the Netherlands up to 80% of employees are not engaged (Gallup Employee Engagement survey, 2011). A conservative estimate of the fiscal costs caused by work-related stress indicates that they amount to some EUR 20 billion annually in the European Union (Levi & Levi, 2002). So, there is a need to better understand how and why individuals become engaged with their work in order to improve employee well-being and organizational performance (Robertson & Cooper, 2010).

An example with regard to personal efficacy is made by Judge, Bono (2001) who conclude that increasing one’s self-efficacy, and self-esteem could lead to higher overall job performance. Personal efficacy is a relevant predictor in teaching commitment and teacher retention (Coladarci, 1992; Chestnut, Burley 2015). Also, the definition of work engagement consists of activation and energy, certain personality aspects can support engagement due to their

characteristics (Xanthopoulou, Bakker, Demerouti, & Schaufeli, 2009). Especially characteristics like high extraversion, agreeableness, openness and conscientiousness as well as low

neuroticism, relate to high levels of work engagement.

The second line of reasoning explicates the positive aspects and outcomes of work engagement. In its own definition of work engagement as defined by Schaufeli et al (2002) the dimension vigor, dedication and absorption can be viewed as positive aspects. These can be summarized as follows. Vigor: high levels of energy, mental resilience, willingness to invest effort in one’s work, persistence in the face of difficulties. Dedication: sense of significance, enthusiasm, inspiration, pride and challenge. Absorption: full concentration, deeply engrossed in one’s work, whereby time passes quickly and difficulties with detaching oneself from work. So, in itself these are aspects that are worthwhile to pursue for any organization and therefore justify a further research of its antecedents.

Earlier research shows that work engagement is seen as a critical antecedent of various positive organizational outcomes such as citizenship behavior and employee productivity (Akhtar, Boustani, Tsivrikos, Chamorro-Premuzic, 2015). There are studies that show that work engagement is positively related to task performance, contextual performance, and active learning (Bakker, Demerouti, Brummelhuis, 2012). Work engagement also influences

productivity and performance (Saks, 2006). From previous research (Schaufeli, 2004; Hakanen, Bakker, Schaufeli, 2006) we know that work engagement is an important construct in

satisfaction and retention of employees. Hakanen, Bakker, Schaufeli (2006) used the Schaufeli ea (2004) scales of work engagement and especially the aspects vigor and dedication to establish the relationship between engagement and organizational commitment. Teacher satisfaction is an important factor in their commitment and retention to the profession (Huang, Waxman, 2008). So, from this line of reasoning there is also a need to better understand how and why individuals become engaged with their work in order to improve employee well-being and organizational performance (Robertson & Cooper, 2010).

(15)

Despite the possible importance that can be derived from both lines of reasoning and the above-mentioned previous research, the antecedents of work engagement have not been researched in de MBO sector. Current research will therefore focus on these antecedents of work engagement in the MBO sector.

2.3 Job crafting

In this section first, the construct job crafting will be introduced. Then we will explore existing research. And finally, we will discuss the relationship between job crafting and work

engagement.

2.3.1 Job crafting

Wrzesniewski, Dutton (2001) proposed that employees craft their jobs by changing cognitive-, task- and/or relational boundaries to shape interactions and relationships with others at work. These altered tasks and relational configurations change the design and social environment of the job, which in turn, alters work meanings and work identity (Wrzesniewski, Dutton, 2001). They envisioned employees as active crafters of their work. Wrzesniewski, Dutton (2001) assume that employees alter task and relational boundaries of their jobs to create work with which they are more satisfied.

2.3.2 Existing research on job crafting

Wrzesniewski, Dutton (2001) state that job crafting complements the theories of job design (Hackman & Oldham, 1980), as it addresses the processes by which employees change elements of their job. The idea that employees proactively shape jobs, skills and work environments is found in different studies of job design (e.g. Parker, Wall, Jackson, 1997). Oldham, Fried (2016) study the past, present and future of job design and describe the future direction of job design with an emphasis on job crafting. Wrzesniewski, Dutton (2001) argue that the motivation to craft a job is moderated by the perceived opportunity to do so, as well as by the individuals’ work and motivational orientations. Job crafting may be facilitated by job and individual characteristics and may enable employees to fit their jobs to their personal knowledge, skills and abilities on the one hand and to their preferences and needs on the other hand (Tims, Bakker, 2010). Strümpfer (1990) argues that employees who are resilient, who possess a high sense of coherence, personality hardiness and an internal locus of control, develop

‘occupational self-direction’ That is, they use initiative, thought and independent judgement in their work. This may help them to remain healthy and to be able to deal with stressors as challenges instead of high pressures. Tims, Bakker (2010) find that job crafting can be seen as a specific form of proactive behavior in which the employee initiates changes in the level of job demands and job resources. Grant (2000) discusses four constructs related to proactive

behavior: proactive personality, personal initiative, role breadth self-efficacy and taking charge. Tims, Bakker (2010) also state that job crafting is important because we can focus on the changes that employees make in their jobs, to enhance or benefit their own goals. Such as a better person–job fit, more enjoyable work, better well-being, less work–home conflict, etc. So, job crafting can be seen as a coping mechanism for employees to cope with aspects of their job

(16)

and adjust the design of their job. Employees could be encouraged to exert more influence on their job characteristics (Tims, Bakker, 2010). As practical and managerial implications, Tims, Bakker (2010) suggest that job crafting may be a good way for employees to improve their work motivation and other positive work outcomes. Work engagement can be seen as one of these positive outcomes.

2.3.3 Job crafting and work engagement

From existing research, we know that job crafting is an important antecedent of engagement (Bakker, Tims, Derks, 2010; Petrou et al, 2012; Bakker, Tims, Derks, 2012). This leads to our first hypothesis. H1: Job crafting is positively related to their work engagement.

2.4 Challenging work situation

In this section first, we will discuss the relationship between challenging work situation and job crafting. We will also discuss the mediating effect of job crafting on the relationship between challenging work situation and work engagement.

2.4.1 Challenging work situation and job crafting

Research suggest that job crafting is especially relevant to work engagement in the context of a challenging work situation (Tims, Bakker, 2010; Duncan, 1972; Wrzesniewski, Dutton, 2001; Shamir and Howell, 1999). This is supported by claims that work-related stress (caused by challenging work situation) may be prevented or counteracted by job redesign by empowering the employees, and avoiding both over- and underload, by improving social support (Levi & Levi, 2002). Job resources like job control were seen as a way for employees to deal with job

demands based on the job demands resources model (Karasek, 1979). The discussed job resources by Hakanen, Bakker, Schaufeli (2006) correspond with the Tims, Bakker (2010) job resources in the definition of job crafting where job resources are used to reshape the job design. Employees are in a position to make a lot of decisions to complete their tasks, so in this sense a challenging (dynamic) work situation will also give employees a lot of opportunities and perhaps even attract employees that are willing to work and act in these conditions. Because changing environments are characterized by few situational cues, few constraints and few re-enforcers to guide behavior (Shamir and Howell, 1999). These dynamic and uncertain

environments require new interpretations, novel responses, and different levels of effort and investment (Shamir and Howell, 1999). Because employees make these decisions with regard to their own job design, they are in the position to restructure their own jobs. This phenomenon is known as job crafting (Wrzesniewski & Dutton, 2001). The challenging work situation gives them the opportunity for job crafting. Petrou et al (2012) found support for Wrzesnieuwski, Dutton (2001) proposition that job crafting can be of particular importance during organizational change. When employees are not in phase of resistance, they may facilitate their adaption to the change proactively. The way to achieve this can be by adjusting their work environment. A high degree of uncertainty and fast pace change of a challenging work situation gives urgency for employees to actively restructure their job design, through job crafting. In this sense, challenging work situations create opportunity, motivation and urgency for employees to

(17)

engage in job crafting to become more satisfied and engaged in their work. This leads to our second hypothesis. H2: The challenging work situations is positively related to job crafting.

2.4.2 Job crafting and the mediating effect on the relationship between challenging work situation and work engagement

Hakanen, Bakker, Schaufeli (2006) studied the mediating role of work engagement between job resources and organizational commitment. Job resources that are influenced by the context of the job, for example challenging work situation. We’ve already established that there is

research that suggest that job crafting is especially relevant to work engagement in the context of a challenging work situation (Tims, Bakker, 2010; Duncan, 1972; Wrzesniewski, Dutton, 2001; Shamir and Howell, 1999). Bakker, Tims, Derks (2010) and Petrou et al (2012) have established the relationship between job crafting and work engagement. Petrou et al (2012) found a link between job crafting and work engagement that illustrates a potential motivational role for job crafting. Many changes give employees opportunities for change and if employees use this through job crafting their work engagement will increase. This leads to our third hypothesis. H3:

Job crafting mediates the relationship between the challenging work situation and work

engagement. Moreover, the stronger is the perception of challenging work situation the higher will be work engagement due to their increased job crafting.

2.5 Locus of control

In this section we will first introduce locus of control. Then we will discuss the moderating effect locus of control can have on the relationship between challenging work situation and job

crafting.

2.5.1 Locus of control

The concept of locus of control was introduced by Rotter in 1954. locus of control is the degree to which people believe that they have control over the outcome of events in their lives, as opposed to external forces beyond their control. Rotter (1966) differentiates internal and external locus of control. A person's locus is conceptualized as internal: ‘a belief that one's life can be controlled’ or external: ‘a belief that life is controlled by outside factors which they cannot influence, or that chance or fate controls their lives’ (Rotter, 1966). Internals are those who believe that they are the masters of their fate and, therefore, often are confident, alert, and directive in attempting to control their external environments. Further, they often perceive a strong link between their actions and consequences. Externals, on the other hand, are those who believe that they do not have direct control of their fate and perceive themselves in a passive role with regard to the external environment. They, therefore, tend to attribute personal outcomes to external factors or luck.

2.5.2 Locus of control and moderating effect on the relationship between challenging work situation and job crafting

In the next paragraphs we will first discuss the relevance of appropriate context for individual personality characteristics to be expressed. Then we will discuss individual personality

(18)

characteristics in comparable research. And finally, we will explore existing research that suggests locus of control can be a moderator in the relationship between challenging work situation and job crafting.

Context may be relevant for the expression of individual personality characteristics. The moderating role locus of control plays in the relationship between challenging work situation and job crafting is supported by trait activation theory (Tett and Burnett, 2003). Tett and Guterman (2000) showed that there is a relationship between self-report trait measures and trait-relevant behavioral intentions in situations that provide appropriate cues for trait expression. Tett and Burnett (2003) found support for this trait activation theory: personality traits require trait-relevant situations for their expression. So, individuals behave in trait-like ways in those situations that are relevant to the given trait. So, job crafting in a challenging work situation could attract employees with trait relevant personality characteristics. This is also found in comparable leaderhip and management research. Shamir and Howell (1999) state with regard to charismatic leadership: ‘Charismatic leaders are more likely to emerge ... in

environments characterized by a high degree of change and by great opportunities for change than in stable environments that offer few inducements for change or opportunities for change’. The same could be true for all employees. Employees with individual personality characteristics (like internal locus of control) that thrive in environments that are challenging may emerge in these environments and take full advantage of the opportunities for change (through job crafting). So, these individual personality characteristics may moderate their behavior.

Comparable research showed the importance of personality characteristics. NG et al (2006) found that internal locus was positively associated with favorable work outcomes, such as positive task and social experiences, and greater job motivation. Bakker, Tims, Derks (2012) have shown there is a relationship between proactive personality, job crafting and work engagement. They concluded that employees who were characterized by a proactive

personality were most likely to craft their jobs (increase their structural and social job resources and increase their job challenges). Job crafting, in turn, was predictive of work engagement (vigor, dedication, and absorption) and colleague-ratings of in-role performance. These findings suggest that, the personality aspect supports employees to engage in proactively adjusting their work environment, and as a result they manage to stay engaged and perform well.

Xanthopoulou, Bakker, Demerouti, & Schaufeli (2009) found that the definition of work engagement consists of activation and energy, certain personality aspects that can support engagement due to their characteristics. Especially characteristics like high extraversion,

agreeableness, openness and conscientiousness as well as low neuroticism, relate to high levels of work engagement. The role of emotional intelligence as an antecedent of work engagement was explored by Akhtar, Boustani, Tsivrikos, Chamorro-Premuzic (2015); Petrides & Furnham (2006). In line with this Grant (2000) discusses four constructs related to proactive behavior: proactive personality, personal initiative, role breadth self-efficacy and taking charge. This comparable research shows the growing interest in exploring the relationship between personality and work engagement and supports the possible importance of personality characteristics like locus of control.

(19)

Previous research suggests locus of control can be a moderator in the relationship between a challenging work situation and job crafting. Wrzesniewski, Dutton (2001) argue that the motivation to craft a job is moderated by the perceived opportunity to do so, as well as by the individuals’ work and motivational orientations. The challenging work situation could provide the perceived opportunity to do so. Tims, Bakker (2010) state that job crafting may also be facilitated by job and individual characteristics and may enable employees to fit their jobs to their personal knowledge, skills and abilities on the one hand and to their preferences and needs on the other hand. Locus of control could be an individual characteristics Tims, Bakker (2010) refer to. Far before the idea of Job crafting (by Wrzesniewski and Dutton) was proposed in 2001 locus of control was suggested as a predictor of ‘occupational self-direction’. Strümpfer (1990) argues that employees who are resilient, who possess a high sense of coherence,

personality hardiness and an internal locus of control, develop ‘occupational self-direction’ That is, they use initiative, thought and independent judgement in their work. In line with this exiting knowledge, current study proposes that locus of control could be a moderator to the

relationship between challenging work situation and job crafting, by changing the boundaries of the employees’ job itself in addition to the self-direction within the job that Strümpfer

proposes.

Overall, it appears that the personality trait of locus of control provides a useful theoretical perspective to enhance the explanation and prediction of employees’ workplace attitudes and behaviors. Whether, and to what extent employees actually engage in the opportunity for job crafting could be related to their own personal perceptions that are based on the individual characteristics (moderating effect). When employees have a high internal locus of control, they have the conviction that they have control over their situation and they can achieve positive results through their actions. Thus, I expect that employees with a high internal locus of control will perceive the challenging work situation as an opportunity to engage in job crafting (positive moderating effect). When employees have high external locus of control I expect the opposite effect. In this case employees do not have the confidence that their own actions will influence the outcome. Thus, they will perceive the challenging work situation as a threat, which they cannot influence, so they will not engage in job crafting (negative moderating effect). This leads to the fourth hypothesis. H4: Locus of control moderates the relationship between challenging

work situations and job crafting, such that the positive relationship will be stronger when locus of control is higher.

2.6 The overall research model: work engagement and its antecedents

Current research aims to contribute by exploring the relationship between de context of the challenging work situation and state of mind work engagement. Within this relationship the mediation effect of employee behavior job crafting is researched. Also, the moderating effect of the personality characteristic locus of control on the relationship between challenging work situation and job crafting is explored. The previous sections provided the building blocks that leads to the fifth hypothesis. H5: The perception of challenging work situations is related to work

(20)

situations and locus of control is related to job crafting, which in turn is related to work engagement.

2.7 Conceptual model

The conceptual model of current research is illustrated in the figure that is enclosed below.

Figure 1. Research conceptual model.

2.8 Summary of hypotheses

The proposed hypotheses from the sections above are summarized below.

H1: Job crafting is positively related to their work engagement. H2: The challenging work situation is positively related to job crafting

H3: Job crafting mediates the relationship between the challenging work situation and work engagement. Moreover, the stronger is the perception of challenging work situation the higher will be work engagement due to their increased job crafting.

H4: Locus of control moderates the relationship between challenging work situation and job crafting, such that the positive relationship will be stronger when locus of control is higher. H5: The perception of challenging work situation is related to work engagement via conditional indirect effects, such that the interaction between challenging work situation and locus of control is related to job crafting, which in turn is related to work engagement.

(21)

3. Method

In this chapter the adopted method of research will be discussed.

3.1 Participants

The total sample consisted of 226 employees in de the MBO sector. Participants are employed in different MBO institutions (ROC TOP, ROC van Amsterdam, Regio College, ROC Nijmegen). The employees’ age ranged from 22 to 64 (M = 46.32 SD = 11.63), they consisted of mostly females (female 60%, male 35% and undisclosed 5%). The employee tenure at the current employer ranged from 0 to 40 rounded years (M = 9.99 SD = 9.32) The employee tenure in the MBO sector ranged from 0 to 41 rounded years (M = 11.04 SD = 9.69). The function type of the participants is spread over the different function types: 55% onderwijzend, 34% ondersteunend and 11% management/directie/bestuur.

A heat map of the participants is enclosed below, most participants are from the greater Amsterdam area (MBO institutions: ROC TOP, ROC van Amsterdam, Regio College) and the Nijmegen area (MBO institution: ROC Nijmegen):

Figure 2. Heat map participants.

3.2 Procedure

The data was gathered by administrating one survey. The survey was formulated in Dutch which is the language that is common in the participating organizations. In order to maximize the number of responses and due to the limited time available for the data collection, non-probability convenience sampling was used in approaching the participants. The survey was distributed through an email containing an invitation to participate and all necessary

(22)

information and the anonymous survey link. To increase the response rate, participants were offered a small incentive. This incentive is attendance of a workshop where the results of the study will be presented and discussed. Twelve participants expressed interest in attending the workshop.

The survey was sent to 610 email addresses, and 226 participants completed the survey, thus resulting in a response rate of 37%. No reminder was sent.

3.3 Measures

Below we will discuss the measures that were used. All measures used scales that were existing scales from previous research. The description and summary of the measures per variable of the conceptual model and the control items are elaborated on below. Also, the references are included. If applicable, the possible dimensions, and the items and type of scale is included. When applicable a translation and back-translation was executed to check and ensure the correct translation from English to Dutch.

3.3.1 Challenging work situation

The challenging work situation is operationalized into two dimensions: Perceived dynamic work environment (De Hoogh et al, 2005) and Pace of change (Ayyagari et al, 2011).

Dimension perceived dynamic work environment (De Hoogh et al, 2005). This is measured through a scale developed by De Hoogh et al (2005) and will be used by author providing a measure of perceived dynamic work environment based on three items: ‘Mijn werk is uitdagend'. 'Mijn werk is dynamisch'. And: 'Mijn werk geeft grote kansen om veranderingen door te voeren'. The items were rated on a 7-point response scale, ranging from 1 (helemaal nooit/not at all) to 7 (heel vaak of altijd/very much so).

Dimension pace of change (Ayyagari et al, 2011). A scale used by author providing a measure of perceived pace of change based on four items: ‘Er vinden veranderingen plaats in

werkprocessen en procedures’. ‘De dagelijkse routine op het werk verandert’. ‘De eisen die worden gesteld aan mijn werk veranderen’. ‘De manier waarop we werken verandert’. These scales were formulated in English. Translation was used to convert the questions to Dutch and back-translation was used to ensure the correct translation of the scales. The items were rated on a 7-point response scale, ranging from 1 (helemaal nooit/not at all) to 7 (heel vaak of altijd/very much so).

The scale showed a good reliability (Cronbach’s alpha .806) satisfying the acceptable standard value of .70. The answers were averaged first on dimension and finally into one score.

(23)

3.3.2 Work engagement

With regard to work engagement Schaufeli, Bakker (2004) provide a measure of work engagement consisting of three dimensions: Vigor, Dedication and Absorption and a total of nine items. The items were rated on a 7-point response scale, ranging from 1 (helemaal nooit/not at all) to 7 (heel vaak of altijd/very much so).

The dimension vigor was based on three items on which the following questions were asked: ‘Op mijn werk bruis ik van energie’, ‘Als ik werk voel ik me fit en sterk’, ‘Als ik ‘s morgens opsta heb ik zin om aan het werk te gaan’.

The dimension dedication was based on three items on which the following questions were asked: ‘Ik ben enthousiast over mijn baan’, ‘Mijn werk inspireert mij’, ‘Ik ben trots op het werk dat ik doe’.

The dimension absorption three items on which the following questions were asked: ‘Wanneer ik heel intensief aan het werk ben, voel ik mij gelukkig’, ‘Ik ga helemaal op in mijn werk’, ‘Mijn werk brengt mij in vervoering’.

The initial scale showed a good reliability (Cronbach’s alpha .937) satisfying the acceptable standard value of .70. The answers were averaged first on dimension and finally into one score.

3.3.3 Job crafting

Tims, Bakker, Derks (2012) provides a measure of job crafting in 4 dimensions existing of 15 items on a 7-point response scale ranging from 1 (zeer mee oneens/not at all) to 7 (zeer mee eens/very much so). The dimensions are: increasing structural job resources, decreasing hindering job demands, increasing social job resources, increasing challenging job demands. Dimension increasing structural job resources is a measure based on four items on which the following questions were asked: ‘Probeert zijn/haar vaardigheden te ontwikkelen’, ‘Probeert zichzelf op werkgebied te ontwikkelen’, ‘Probeert nieuwe dingen te leren op het werk’, ‘Zorgt ervoor dat hij/zij volledig gebruik maakt van zijn/haar vaardigheden’.

Dimension Decreasing hindering job demands is a measure based on four items on which the following questions were asked: ‘Zorgt er voor dat zijn/haar werk mentaal minder zwaar is’, ‘Zorgt ervoor dat zijn/haar werk op emotioneel gebied minder intens is’, ‘Richt zijn/haar werk dusdanig in dat hij/zij zo min mogelijk te maken krijgt met problemen van anderen die

hem/haar emotioneel zouden kunnen raken’, ‘Richt zijn/haar werk dusdanig in dat hij/zij zo min mogelijk contact heft met mensen met onrealistische verwachtingen’.

Dimension Increasing social job resources is a measure based on three items on which the following questions were asked: ‘Vraagt anderen om hem/haar te coachen’, ‘Vraagt anderen of zij tevreden zijn met zijn/haar werk’, ‘Richt zich op anderen om inspiratie op te doen’.

(24)

Dimension Increasing challenging job demands is a measure based on four items on which the following questions were asked: ‘Biedt zichzelf actief aan bij interessante projecten’, ‘Staat vooraan bij het leren en toepassen van nieuwe ontwikkelingen’, ‘Ziet rustige perioden op het werk als een kans om nieuwe projecten op te starten’, ‘Neemt regelmatig extra taken op zich ook als dit niet extra beloond wordt’.

The initial scale showed a good reliability (Cronbach’s alpha .815) satisfying the acceptable standard value of .70. Improvement was possible (to Cronbach’s alpha .830). Due to the fact that this improvement is minor and only just above the normal correction threshold of .10 no dimensions where excluded, because this would decrease the dimension sample size. The answers were averaged first on dimension and finally into one score.

3.3.4 Locus of control

With regard to locus of control De Hoogh, Den Hartog (2009) provide a measure in Dutch, based on the English measure of Spector (1988) consisting of sixteen items, on a 7-point response scale ranging from 1 (Zeer mee oneens/not at all) to 7 (Zeer mee eens/very much so). The formulation of some questions was such that it produced reverse coded responses. The following questions were asked: ‘Een baan is wat je er zelf van maakt’, ‘Bij de meeste banen kunnen mensen in feite bereiken, wat ze zich voornemen te bereiken’, ‘Als je weet wat je in een baan zoekt, dan kun je de baan vinden die dat biedt’, ‘Als werknemers ontevreden zijn over een beslissing van hun baas, dan moeten ze daar wat aan doen’, ‘De baan krijgen die je zoekt, is meestal een kwestie van geluk’, ‘Geld verdienen is hoofdzakelijk een kwestie van geluk’, ‘De meeste mensen kunnen hun werk goed doen als ze hun best doen’, ‘Om een echte goede baan te krijgen, heb je familie of vrienden in hoge posities nodig’, ‘Promoties zijn meestal een kwestie van geluk’, ‘ Een goede baan krijgen is meer een kwestie van wie je kent dan van wat je weet’, ‘Promotie op het werk krijg je als je je werk goed doet’, ‘Om veel geld te verdienen, moet je de juiste mensen kennen’, ‘In de meeste banen moet je geluk hebben om als werknemer op te vallen’, ‘Mensen die goed presteren, worden daar over het algemeen voor beloond’, ‘De meeste werknemers hebben meer invloed op hun leidinggevenden dan ze zelf denken’, ‘Het grootste verschil tussen mensen die veel geld verdienen en mensen die weinig geld verdienen is geluk’. The initial scale showed a good reliability (Cronbach’s alpha .838) satisfying the acceptable standard value of .70. Improvement was possible (to Cronbach’s alpha .841). Due to the fact that this improvement is minor and below the normal correction threshold of .10 no dimensions where excluded, because this would decrease the dimension sample size. The answers were averaged into one score.

3.3.5 Control items

Based on comparable research the following control variables are used: gender, age, tenure current employer and tenure in the MBO sector, currently working in MBO, function type (onderwijzend, ondersteunend and management/directie/bestuur).

(25)

Gender, age and tenure are common control variables used in similar research (Sonnentag, 2003). Gender, as the gender stereotypes have shown to influence perceptions. Age and tenure are used as control variables as they could influence the perceptions of participants locus of control, job crafting and work engagement. The question whether the participant is currently working in the MBO sector is a control question to establish whether the participant is in the right target group. Participants not working in the MBO sector are excluded from the results. Function type was introduced as a control variable due to the fact that work engagement could be influenced by the function that an employee fulfills in the organization. The job demands and resources as well as personal resources could differ for teachers, support staff and board and management and this could influence work engagement (Mauno, 2007).

Per control item, the following control questions were asked at the beginning of the survey. Gender: ‘Wat is uw geslacht: man/vrouw’, it was also possible not to respond to this question. Age: ‘Wat is uw leeftijd in jaren’, participants were asked to round their answers in years. Currently working in MBO sector: ‘Bent u op dit moment werkzaam in het MBO: Ja/Nee’. Tenure: ‘Hoeveel jaren werkt u bij uw huidige werkgever’ and: ‘Hoeveel jaren werkt u in het MBO’, participants were asked to round their answers in years.

Function type: ‘Wat is uw huidige functietype: onderwijzend personeel/ondersteunend personeel/management directie bestuur’.

(26)

4. Results

This chapter will present the results of data analysis. The data is based on the attained sample. First the raw data handling is described. Then the correlations matrix will be discussed. Then, the hypotheses will be tested using regression analysis. If significant, Process will be used to further probe the interaction effects (Hayes, 2013). Finally, a summary of the results per hypothesis is included.

4.1 Raw data handling

In this section the handling of the raw data, and the steps and actions take are taken are discussed.

Data cleaning, missing data and frequencies

The data collection was executed through a survey. The survey was distributed through emailing an invitation to participate in the research. This invitation included a link to the online survey in Qualtrics Survey Software. The data was imported in SPSS from Qualtrics Survey Software. Imported were 226 recorded responses in SPSS. On the data a missing variable analysis was run on all items. The incomplete responses with missing data were deleted, after this action 197 complete responses remained. On the data, frequencies per variable on the item level were run. Examination of frequencies gave no errors in the data entry in the remaining recorded

responses.

Recoding counter-indicative items

Several items in the survey were reverse coded. To be able to run analysis these items were recoded in counter-indicatives, these were eight items in variable locus of control. Recoding was executed into new items. These items are 7-point Likert scale. Values were adjusted according to original scale items and measure to scale in variable view.

Skewness and Kurtosis

All variables were analyzed on the item level for skewness and kurtosis through running descriptives. This showed that all variables show skewness: moderate negative skewness. This was corrected for all items (through using the formula X*=Sqrt(K-X)). After this transformation skewness and kurtosis are no longer an issue.

Outliers check

Outliers were identified based on the standardized values of variables and frequency tables. Excluded Zscore > |3|. Two outliers were identified in variable job crafting (one item) and locus of control (one item). After this, 195 items remain.

Dummy coding

Dummy coded the control variables gender (man, vrouw, undisclosed) and functiontype (onderwijzend, ondersteunend, management/directie/bestuur) into new items.

(27)

Computing reliability and scale means

The Cronbach’s alpha’s are: challenging work situation .806 (best), job crafting .815

(improvement possible to .830), locus of control .838 (improvement possible to .841), work engagement .937 (best). Due to little improvement (only one > .10) no dimensions where excluded, because this would decrease dimension sample size. All scales are above .70, so they satisfy the minimum reliability criteria. Scale means where calculated, first items then

dimensions and final on the variable level.

4.2 Descriptive statistics

In this section the correlation will be discussed.

Correlation

Correlation allows us to quantify the intensity and meaning of the relationship between two variables. The APA style correlation matrix is included below.

Table 1.

Correlation matrix

Variables one through five are control variables. There are two significant linear relations between tenure current employer and job crafting (,16 that can be interpreted as: ‘absence of relation’) and between function type and locus of control (-,21 that can be interpreted as: ‘tendency to negative relation’). Between the variables of the hypotheses, variables six through to nine all show significant linear relations (from ,29 to ,45 that all can be interpreted as:

‘tendency to positive relation’)

Analysis of variance (ANOVA)

The control variables gender, age and tenure sector do not give a reason to further analyze the variances between different populations. The other control variable gives a significant linear relation that can be quantified as absence of relation (tenure current employer with job crafting) which, because of its size, is not relevant to pursue with regard to the hypotheses at this stage. Moreover, because tenure sector does not show a significant linear relation with job crafting. Current research focusses on the challenging work situation in de MBO sector, not

Table Means, Standard deviations, Correlations

Variables M SD 1 2 3 4 5 6 7 8 9

1. Gender 1,70 0,57

-2. Age 46,54 11,78 ,01

-3. Tenure current employer 9,85 9,28 ,01 ,55**

-4. Tenure sector 10,94 9,71 ,01 ,57** ,83**

-5. Functiontype 1,54 0,69 -,06 ,20** -,10 -,02

-6. Challenging work situation 5,39 0,77 -,04 -,13 -,11 -,13 -,062 (.81)

7. Job Crafting 5,24 0,61 -,07 ,09 ,16* ,11 ,01 ,37** (.82)

8. Locus of control 4,94 0,72 ,03 -,06 ,09 -,02 -,21** ,24** ,29** (.84)

9. Work engagement 5,37 1,01 -,09 ,01 ,08 ,02 -,05 ,42** ,45** ,43** (.94) ** Correlation is significant at the 0.01 level (2-tailed).

* Correlation is significant at the 0.05 level (2-tailed). Note 1. Gender: 1=male, 2=female, 3=undisclosed

(28)

small possible effects that derive from the length of the relationship between the employee with any one employer. There was a statistically significant effect of function type on locus of control, F(1, 193) = 15.88, p <.00. Tukey post-hoc tests revealed that the perceived locus of control was significantly higher in the management/directie/bestuur function type group compared to the onderwijzend function type group (p= .00). There was no statistically

significant difference of the ondersteunend function type group with the onderwijzend function type group (p=.41).

4.3 Exploratory analysis

In this section exploratory analysis will be conducted using regression and Process.

Regression

Regression allows us to predict the behavior of a variable based on its relationship with another variable. The table showing the hierarchical regression model of work engagement is included below.

Table 2.

Hierarchical regression model of work engagement

Hierarchical multiple regression was performed to investigate the ability of challenging work situation, job crafting and locus of control to predict levels of work engagement, after controlling for gender, age, tenure and function type.

Table Hierarchical Regression Model of Work Engagement R R2 R2 change B SE ! t Step 1 0,17 0,29 Gender -0,02 0,01 -0,10 -1,31 Age 0,00 0,00 -0,20 -0,22 Tenure current employer 0,01 0,00 0,18 1,38 Tenure sector 0,00 0,00 -0,12 -0,88 Functiontype -0,02 0,03 -0,34 -0,45 Step 2 0,7 ,48*** ,34*** Gender -0,01 0,01 -0,07 -1,22 Age 0,00 0,00 0,01 0,08 Tenure current employer 0,00 0,00 0,04 0,36 Tenure sector 0,00 0,00 0,00 -0,02 Functiontype 0,01 0,03 0,03 0,42 Challenging work situation 0,33 0,08 ,26*** 3,91 Job Crafting 0,42 0,11 ,26*** 3,89 Locus of control 0,43 0,10 ,29*** 4,55 Note: statistical significance:*p <.05, **p <.01, ***p <.001

(29)

In the first step of hierarchical multiple regression, five predictors were entered: gender, age, tenure current employer, tenure sector and function type. This model wasn’t statistically

significant F (5, 189) = 0,841; p = 0,522 and explained 2.2% of the variance in work engagement. After entry of challenging work situation, job crafting and locus of control at step 2 the total variance explained by the model as a whole was 35,8% F (8, 186) = 12,985; p < .001. The

introduction of challenging work situation, job crafting and locus of control explained additional 33,7% variance in work engagement, after controlling for gender, age, tenure and function type (R2 change = ,337; F(3, 186) =32,52; p < 0,001). In the final model three out of eight predictor variables were statistically significant, with locus of control recording a higher Beta value (b = .29, p < 0,001) than job crafting (b = .26, p < 0,001) and challenging work situation (b = .26, p < 0,001). In other words, if a person’s locus of control increases by one, their work engagement increases by 0,29. If a person’s job crafting increases by one, their work engagement increases by 0,26. And if a person’s perception of challenging work situation increases by one, their work engagement increases by 0,26. The predictors gender, age, tenure current employer, tenure sector and function type do not have statistically significant effect on the variance in work engagement.

Analysis of interaction effects using Process

Because the research model is a moderated mediation model in Process this means that first model 4 simple mediation is executed and then model 7 that gives the results for the

conditional effect through moderation.

Mediation

In this part of the model challenging work situation is the independent variable (X), work engagement is de dependent variable (Y) and job crafting is the mediating variable (M). We will explore the mechanisms by which challenging work situation causally influences work

engagement. The direct effect, the indirect effect and the total effect are tested.

Figure 3. Research mediation model

The results of the analysis are shown in the table of the total, direct and indirect effects is enclosed below.

(30)

Table 3.

Mediation total, direct and indirect effects

Indirect effect (a1b1 both significant p < 0,001)

a1. Two employees that differ by one unit on challenging work situation are estimated to differ by a= ,29 units on job crafting. The sign is positive, meaning that those relatively higher in challenging work situation are estimated to be higher in job crafting. This effect is statistically different from zero, t=5,6008, p=.000, with a 95% confidence interval from 0,1883 to 0,3929. b1. Two employees who experience the same level of challenging work situation (who are equal on challenging work situation) but differ one unit on their level of job crafting are estimated to differ by b= ,56 units in work engagement. The sign of b is positive so this means that those relatively higher in job crafting are estimated to be higher in work engagement. This effect is statistically different from zero, t=5,1693, p=,0000 with a 95% confidence interval from 0,3446 to 0,7698.

Direct effect (c1’ significant p < 0,001)

c1’. Two employees that differ one unit on challenging work situation and who experience the same level of job crafting (who are equal in Job crafting) are estimated to differ by 0,371 units in their work engagement (to be 0,371 unit more work engaged). This effect is statistically

different from zero, t=4,4338 p=,0000 with a 95% confidence interval from 0,2062 to 0,5366. The total variance of work engagement explained by the overall model is 28% (R2), which is statistically significant (P<0,01)

Total effect (C1 significant p < 0,01)

Total effect of challenging work situation on work engagement is c = 0,5333. Two employees who differ by one unit in challenging work situation are estimated to differ by 0,53 units in their perceived work engagement. The sign is positive meaning the greater perceived challenging

Consequent

Job crafting (M) Exit (Y)

Antecedent Coeff SE p Coeff SE p

Challenging work situation (X) a1 0,29 0,052 <0,001 c1' 0,3714 0,83 <0,001

Job Crafting (M) --- --- --- b1 0,557 0,108 <0,001

constant i1 11,5522 0,0824 <0,001 i2 0,7358 0,083 <0,001

R2 = ,14 R2 = ,2778

F(1, 193) = 31,3695, p <0,001 F(2, 192) = 36,9220, p < 0,001

Effect SE p LLCI ULCI

Direct effect c1' 0,3714 0,83 <0,001 0,2062 0,5366 Total effect C1 0,5333 0,0827 <0,001 0,3702 0,6965

Boot SE Boot LLCI Boot ULCI

(31)

work situation means higher work engagement. This effect is statistically different from zero,

t=6,4483, p=0,0000 or between 0,3702 and 0,6965 with 95% confidence.

Indirect effect (a1b1 significant P < 0,001)

Two employees who differ by one unit in their challenging work situation are estimated to differ 0,16 units in their perceived work engagement as a result of those who perceive a higher

challenging work situation they are more likely to engage in job crafting, which in turn translates to higher work engagement. This indirect effect is statistically different from zero, as revealed by a 95% BC bootstrap confidence interval that is entirely above zero (0,0922 to 0,2605). The conclusion with regard to mediation is that, because both the direct and indirect effect are significant, there is no full mediation but partial mediation.

Moderation

In moderation the conditional effect is tested. In our hypothesis the condition under which locus of control (W) has a moderating effect on the mediation job crafting (M). Challenging work situation is the independent variable (X), work engagement is de dependent variable (Y).

Figure 4. Research moderation model

The p value of the interaction term is p = 0,5005 so the moderation effect is not taking place (p is not <,05). So, this means that there is no statistical evidence that locus of control moderates the (partial) mediation that job crafting has in the relationship between challenging work situation and work engagement.

4.4 Test of hypotheses

In this section the results per hypothesis are discussed.

Hypothesis 1: Job crafting is positively related to their work engagement.

The table of means, standard deviations, correlations shows a significant linear relation between challenging work situation and job crafting (of ,45 significant at the 0.01 level (2-tailed)) that can be interpreted as ‘tendency to positive relation’.

Hierarchical multiple regression shows a statistically relevant model with challenging work situation, job crafting and locus of control explained additional 33,7% variance in work

“Bevlogenheid in het middelbaar beroepsonderwijs” : how is work engagement affected by the challenging work situation in the MBO education sector : what mediating effect has job crafting on this relationship and how doe