The effect of the amount of working hours on job crafting and level of stress

The effect of the amount of working hours on job

crafting and level of stress

Date: 29-06-2016

University of Amsterdam Faculty Economic & Business Specialization: Business Studies Supervisor: E. Federici

Second supervisor: C. Boon

Statement of Originality

This document is written by Laura van Dokkum 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.

Table of Content

Abstract ... 4

1. Introduction ... 5

2. Literature review ... 7

2.1 Conceptualizations of job crafting... 7

2.2 Job crafting measurement ... 8

2.3 Job crafting outcomes ... 9

2.3.1 Individual outcomes ... 9

2.3.2 Organizational outcomes ... 10

2.4 Part-time job crafting ... 12

2.5 Hypotheses ... 13

3. Methodology ... 14

3.1 Procedure ... 14

3.2 Sample ... 15

3.3 Measures... 16

3.3.1 Control variables and Working hours ... 16

3.3.2 Task crafting ... 17 3.3.3 Relational crafting ... 17 3.3.4 Cognitive crafting ... 18 3.3.5 Stress ... 18 4. Results ... 19 4.1 Descriptive statistics ... 19 4.2 Regressions ... 20

4.2.1 Predicting stress by part timers and job crafting ... 21

4.2.2 Predicting stress by part time employees, task, relational, and cognitive crafting ... 22

4.3 Moderation effect ... 23

4.3.1 Effect of part timers on the relation between job crafting and stress ... 23

4.3.2 Effect of part timers on the relation between task crafting and stress ... 24

4.3.3 Effect of part timers on the relation between relational crafting and stress ... 25

4.3.4 Effect of part timers on the relation between cognitive crafting and stress ... 26

5. Discussion... 27 5.1 Theoretical contributions ... 28 5.2 Limitations ... 29 5.3 Practical implications ... 30 6. Conclusion ... 30 Appendices ... 34 Appendix A ... 34 Appendix B ... 40

Abstract

This study examined whether there was a positive relationship between (one of the types of) job crafting and the stress level of employees. These types included task crafting, relational crafting and cognitive crafting. Thereafter, there was examined how this positive relation was effected by working part time. In this study, the theoretical framework of Wrzesniewski and Dutton (2001) was used and therefore job crafting was defined as the psychical and cognitive changes an employee makes in the task or relational boundaries of their work. Respondents were Dutch employees working in different sectors and organizations (N = 118). The results of hierarchical multiple regressions showed that the relationship between job crafting or one of the types of it, and the stress level of an employee, was very small and moreover, not significant. Results of the moderation tests showed that the interaction of working part time and job crafting or one type of it, predicting stress, was positive and furthermore not significant. Although, a significant finding that was found indicated that when two employees craft their job (or one type of it) on an average level, or did not craft their job at all, full timers would experience more stress than part timers.

1. Introduction

As sociologists called constructivism, everyone can construct their own reality and perceive it as the truth even if it is not the reality (Goffman, 1956). Individuals construct their own reality, just as their own job. This means that the same job could have different meanings to different people. Nevertheless, to maintain employees, companies need to provide employees the opportunity to fill in the job as how individuals want it (Wrzesniewski & Dutton, 2001). Simultaneously with this filling in, changes are made in jobs of employees which are important, because it influences not only the performance and feelings of themselves, but also the performances of the organization (Slemp & Vella-Brodrick, 2013; Wrzesniewski & Dutton, 2001).

Job crafting is defined as the psychical and cognitive changes an employee makes in the task or relational boundaries of their work (Wrzesniewski & Dutton, 2001). From this point of view changes are made in task, relational or cognitive boundaries. Another theory stated that job crafting is more being focused on job demands and job resources (Petrou, Demerouti, Peeters, Schaufeli, & Hetland, 2012). They defined job crafting as proactive behaviour that employees showed consisting of seeking challenges, seeking resources and reducing demands and is based on the so called Job Demands-Resources theory (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001). Both theories were focused on job crafting, but the J-DR theory focused only on the behavioural part of it and excluded the cognitive part in contrast to the theory of Wrzesniewski and Dutton (2001). That is why job crafting, in this research, was seen as the way described in the theory of Wrzesniewski and Dutton (2001). Within this view, job crafting consisted of three different types: task crafting, relational crafting and cognitive crafting.

Job crafting could have different consequences for individuals as well as for organizations. One consequence of job crafting included reducing an employee’s perceived level of stress by enlarging the job resources (Crawford, LePine, & Rich, 2010). This was beneficial for an employee as well as the company itself, because it reduced absenteeism among employees and lowers marginal labour costs (Westman & Etzion, 2001).On the other hand, job crafting could not work out for individuals which, in turn, could lead to unpleasant feelings as stress. Concerning the outcomes for organizations, job crafting could increase job satisfaction as well as employee well-being (Tims, Bakker, & Derks, 2013). But here is a downside too, when job crafting for an employee could not work out, it could influence other employees with the negative outcome (Ghitulescu, 2006).

Job crafting is thus a double-edged sword. It could have positive outcomes but negative consequences simultaneously. Logically, when job crafting did not work out and it created

unpleasant feelings as stress and maybe even burnout, it could, on the long term, increase the marginal labour costs (Westman & Etzion, 2001). So, organizations need to be careful with providing the opportunity to employees to craft their job, because the different views of employees might not turn into positive outcomes (Goffman, 1956).

In organizations, there are often two types of labour contract: part time jobs and full time jobs. In the Netherlands, employees working full time work at least 36 hours per week (Hupkens, 2012). This, in turn, means that when employees work 35 hours per week or less, it is called part time employees. According to Dutch statistics, part time jobs have increased since the economic crisis of 2007 (Centraal Bureau voor de Statistiek, 2016). Apparently, according to previous research this is always the case in economic recessions (Kalleberg, 2000). From 2007, when the crisis began, there have been an increase of 10% of the part timers in the labour force. This is important, because these employees might view their jobs differently than full timers (Goffman, 1956). Also, they might feel like they are more able to craft their job than full timers or on the other hand, they experience more stress due to job crafting that did not work out (Wang, Demerouti, & Bakker, 2016; Wrzesniewski & Dutton, 2001).

The distinction between full timers and part timers has never been made in a research of job crafting. The positive relationship between job crafting and stress might be moderated by the amount of working hours of employees. The focus in this research was on the perceived stress of part time employees when crafting their job. According to the findings this research was going to present, organizations know if they have to provide different opportunities to part timers for crafting their job compared to the opportunities for full timers who wanted to craft their job. This in turn, could lead to more efficiency and commitment of the employees (Petrou et al., 2012). Also for individuals, this research gave part timers the knowledge of whether job crafting could give them more stress.

So, this research examined the perceived stress of part timers when crafting their job or at least one type of it. The research question that was going to be answered was: “How is working part time effecting the relationship between job crafting, task crafting, relational crafting, and cognitive crafting and the perceived level of stress of employees?”. Is there a positive relation between (one of the types of) job crafting and stress? And is this relationship moderated by working part time? This study provided answer to these questions.

In the next paragraph conceptualizations, measures and outcomes of job crafting were discussed. Also, job crafting was linked to part time jobs and full time jobs. The paragraph after that consisted of the methodology of this research including the procedure, sample and the measurements. Thereafter, in the result section the descriptive statistics were discussed following

by the analysis of the hierarchical multiple regression and the moderation tests. According to these result the hypotheses were tested and the most important findings were presented. This was followed by the discussion and, in turn, the conclusion consisting of the answer to the research question.

2. Literature review

In this paragraph, literature that were relevant for this research were discussed. The hypotheses were set up from these findings. First of all, there was an explanation of job crafting and the different types of it. Then the measures of job crafting were discussed. After that, the outcomes of where job crafting could lead to for individuals and organizations, were set out. This led to the part of job crafting for part time jobs which was followed by the hypotheses.

2.1 Conceptualizations of job crafting

An individual’s life, the job, the work tasks, and interactions were the raw materials employees used to construct their jobs (Wrzesniewski & Dutton, 2001). The other way around, the job has influenced an employee’s life since work leads to material benefits as money as well as needs for fulfilling autonomy, competence and relatedness (Wang et al., 2016). The job could even influence an individual’s well-being (Tims et al., 2013). A job plays thus a significant role in an employee’s life. That was why employee’s actively changed their work environment when their psychological needs were not met (Wang et al., 2016). Then, the employees are motivated to change their job tasks and characteristics which is also called job crafting.

Job crafting is defined as the psychical and cognitive changes an employee makes in the task or relational boundaries of their work (Wrzesniewski & Dutton, 2001). According to this theory, there are three types of job crafting: task crafting, relational crafting and cognitive crafting. Here, task crafting included alterations in the number of activities that employees made while doing a job. Besides, an employee could also change their cognitive task boundaries. Cognitive crafting involved the way employees looked at their job and how they approached it. The third type of job crafting was changing relational boundaries. Relational crafting referred to changing social interactions at work involving changing the quality and/or the amount of interactions with others at work (Wang et al., 2016; Wrzesniewski & Dutton, 2001). According to the research of Wrzesniewski and Dutton (2001), “job crafters” were thus individuals who framed their job physically, cognitively and relationally. These changes were being made all the time according to Wrzesniewski and Dutton (2001) which means that job crafting occurred on a daily basis. This

theory emphasizes the proactive changes that employees could make in the boundaries of their work that contribute to change the meaning of their work.

Another theory of job crafting involved employee proactive behaviour that is focused on making changes in job characteristics (Tims et al., 2013). This theory was based on the framework of Job Demands-Resources (JD-R) theory (Demerouti et al., 2001). This theory defined job crafting as proactive behaviour that employees showed consisting of seeking challenges, seeking resources and reducing demands (Petrou et al., 2012). According to this theory, job crafters changed their job characteristics including job demands and job resources (Tims et al., 2013). The changes in the level of job demands, the job resources, or both, are being made with their personal abilities and needs. Job demands were defined as job characteristics that required sustained physical or mental effort and are associated with certain physiological and psychological costs. Job resources were contributions towards achieving work-related goals, reducing the effects of job demands and associated costs, and stimulation of personal growth and development (Demerouti et al., 2001).

Comparing the types of job crafting of these theories, there were some differences. Job crafting based on the JD-R model is focused on only two targets: job demands and job resources. Job demands, speaking in terms of the theory of Wrzesniewski and Dutton (2001) could be seen as task crafting. In turn, job resources were seen as relational crafting. The last type of job crafting, cognitive crafting, is being excluded from the JD-R perspective, because this theory was focused only on the behavioural part of job crafting (Petrou et al., 2012). Since the research of Wrzesniewski and Dutton (2001) was focused on not only the behavioural part of job crafting, but cognitive crating as well, the definition according to this research was used in this research.

2.2 Job crafting measurement

Job crafters were employees who changed the task and relational boundaries of work, physically and cognitively (Wrzesniewski & Dutton, 2001). On the daily basis, there were three types of job crafting they could use to make these alterations. These were changes in task boundaries, cognitive boundaries and relational boundaries.

There were different scales to use for measuring job crafting (Wang et al., 2016). Some of them were based on the JD-R model and included seeking challenges, seeking resources and reducing demands (Petrou et al., 2012). This scale, in turn, was based on the scale Tims et al. (2013) provided and was reliable and valid. It consisted of four dimensions: increasing structural job resources, increasing social job resources, increasing challenging job demands and decreasing hindering job demands (Tims, Bakker, & Derks, 2012). Other measuring scales were based on the

theory of Wrzesniewski and Dutton (2001), such as the job crafting measuring scale consisting of physical, relational and cognitive crafting (Laurence, 2010; Slemp & Vella-Brodrick, 2013). Another measuring scale of job crafting was based on the latter scale and consisted not only of cognitive crafting, but also of relationship, task and skill crafting, although this scale had not been validated yet (Bindl, Unsworth, & Gibson, 2014). The difference of this scale was that it comprised both enhancing and limiting job crafting activities in the workplace. Enhancing forms of job crafting could be considered as seeking resources and seeking challenges, while reducing demands could be a form of limiting activities of job crafting (Wang et al., 2016). In this research, only task, relational and cognitive crafting as types of job crafting were measured and therefore, the scale based on the research of Bindl et al. (2014) was used.

2.3 Job crafting outcomes

Enhancing or limiting the different forms of job crafting is realized when employees experience proactive behaviours (Tims et al., 2012; Wrzesniewski & Dutton, 2001). Also, employees could be personally motivated to craft their job. This motivation arises from three different individual needs: the need to control their job in order to avoid alienation from their work, the need for positive self-image at work, and need for human connection with other employees (Wrzesniewski & Dutton, 2001). According to Slemp and Vella-Brodrick (2013) aligns this motivation closely with the needs for autonomy, competence, and relatedness. These three needs are part of the self-determination theory (SDT) what suggests that when psychological needs are met and gives satisfaction, this lead to optimal functioning and psychological adjustment (Slemp & Vella-Brodrick, 2013). Although, not every employee feels like they are motivated to fulfil these needs, because they might find that those needs are met somewhere else than at work (Wrzesniewski & Dutton, 2001). Nevertheless, for the employees who craft their job, it could lead to satisfaction on the short-term, but it also has individual consequences on the long-term (Wang et al., 2016). Not only job crafting could be beneficial for the employee, it also affects the company on the short-term as well as on the long-term. Besides the positive outcomes of job crafting, there is also a downside to job crafting which are discussed below.

2.3.1 Individual outcomes

Individuals could fulfil their (psychological) needs by crafting their job. Employees who crafted their job felt not only satisfied on the short term, but they also might feel more work engagement (Tims et al., 2012; Wrzesniewski & Dutton, 2001). Besides, the study of Slemp and Vella-Brodrick

(2013) provided empirical evidence for a positive relationship between on the one hand, job crafting and mental health on the other. This meant that job crafting influenced an employee’s well-being, though it depended on the extent to which employees engaged in job crafting (Slemp & Vella-Brodrick, 2013). Tims et al. (2013) also stated that engaging in job crafting was positively related with increased well-being. The results of their study suggested that employees could raise their own well-being by job crafting, but only when the organization allowed them to. On the long term, job crafting could change the meaning of work employees had as well as changing their identity and role in organization (Slemp & Vella-Brodrick, 2013; Wrzesniewski & Dutton, 2001). The meaning of work concluded the understanding of the purpose of individual’s work or the achievement the individual believed they got by doing the work (Wrzesniewski & Dutton, 2001). The work identity was according to Wrzesniewski and Dutton (2001), the definition employees gave themselves at work. In the end, job crafting provided employees the ability to turn their own job into an occupation (Slemp & Vella-Brodrick, 2013). In this case, an occupation meant a job an individual was drawn to, enjoyed intrinsically and found meaningful, and became a central part of one owns identity.

But, job crafting also could have more negatively consequences for individuals. Since job crafting started individually, it was important to focus on the negative outcomes of job crafting. First of all, employees could not craft their jobs in every situation, because sometimes in strong situations it was impossible or too difficult to undertake job crafting (Wrzesniewski & Dutton, 2001). In these strong situations, which could occur in every organization, individuals experienced a high degree of pressure to behave in a manner that is prescribed (Berg, Grant, & Johnson, 2010). This feeling of high pressure is associated with a higher level of burnout (Tims et al., 2013). Besides, when employees worked extra hours, which could happen during strong situations, it required additional responsibilities that could diminish individual’s job resources (Wang et al., 2016). Also, job crafting might not work out well, which could result in feelings of frustration among the employee. Other research stated also that employees with work overload experienced unpleasant states such as stress (Berg et al., 2010). Job crafting thus could lead to feelings as stress, but on the other side, it could also increase employees well-being (Tims et al., 2013).

2.3.2 Organizational outcomes

For organizations, the immediate effect of job crafting was that it could lead to job satisfaction. The study of Tims et al. (2013) found that job crafters who increased their job resources by changing them, feel more satisfied. Nevertheless, crafting job demands did not lead to an increase in job satisfaction. This feeling of satisfaction, which depended on the extent to which employees

were engaged in job crafting, predicted employee well-being too (Slemp & Vella-Brodrick, 2013). For organizations, satisfaction and employee well-being could lead to the expectation of improved job performance (Wang et al., 2016). Other studies stated this as well and argued that job crafting was positively associated with commitment to the organization and job effectiveness if job crafting was conceptualized as task, relational and cognitive crafting (Petrou et al., 2012; Slemp & Vella-Brodrick, 2013; Tims et al., 2012). Another consequence of job crafting was that it was negatively related to absenteeism which was beneficial for organizations (Petrou et al., 2012). Although, another study found that only task crafting had a marginal negative effect on it, while relational and cognitive crafting had no significant effect on absenteeism (Ghitulescu, 2006). Despite, job crafting could function as a means to increase attachment among employees and retain valuable employees, because it enhanced organizational commitment. And besides, it increased individual effectiveness and boosted team productivity which led to a direct impact on organizational performance (Ghitulescu, 2006).

Again, besides the positive consequences of job crafting, it could also lead to outcomes that were more detrimental for organizations. Job crafting of one employee could influence the employee’s colleagues (Ghitulescu, 2006). For example, when an employee crafted his own job in such a way that it created discrepancy with others from the work group, then job crafting could have negative effects on the effectiveness of the group. Notwithstanding, the effects of job crafting for organizations all depended on the situation (Wrzesniewski & Dutton, 2001). Besides, since job crafting also could lead to feeling such as stress, it could, in turn, lead to additional costs for organizations. Namely, stress that could follow from high workload or frustration, could in turn, lead to burnout which was not only detrimental for individuals themselves, but also for the organization (Crawford et al., 2010; Iacovides, Fountoulakis, Kaprinis, & Kaprinis, 2003). Previous findings showed that burnout, that could be a consequence of having stress, was positively related to absenteeism and higher marginal labour costs (Westman & Etzion, 2001; Yaniv, 1995). Despite, there was also significant evidence for the relationship between job crafting and decreased burnout (Tims et al., 2013). Job crafting, thus, could not only lead to negative outcomes for the individual or organization, it might also result in pleasant feelings or outcomes for employees and organizations as well.

This research examined the relationship between job crafting and stress of employees. The focus was on the positive relation between job crafting and stress, because the negative individual outcomes of stress were associated with high costs for the organization. Insight in this relationship, could prevent organizations from additional costs and losses of employees.

2.4 Part-time job crafting

Because previous researches already found a positive relationship between job crafting and stress of employees, this study was done to research this relationship among part timers. Previous findings were based on employees that worked full time (Tims et al., 2012). The previous studies concerning job crafting have never focused on only employees that work part time. For job crafting this could give an interesting outcome, because these employees could experience the feeling that the perceived opportunities could be more accentuated. This meant that they could be better able to craft their job (Wrzesniewski & Dutton, 2001). But, since job crafting was a double-edged sword, it also could not work out for part timers and could result in unpleasant feelings such as stress (Wang et al., 2016).

Part time employees were seen as all individuals of labour force that work less than full time. In other words, employees that were having a labour contract with 35 hours per week or less than that, were part timers (Hupkens, 2012). The reason for the focus on part timers was because of the increasing part they took among the labour force (Centraal Bureau voor de Statistiek, 2016). Almost 50% of all the Dutch working individuals between the age of 15 and 75 worked part time according to their labour contract. This also could be interesting for other countries, because part time jobs increased during economic recessions (Kalleberg, 2000). Concerning the statistics in the Netherlands, the amount of part timers increased with 10% since 2007 when the economic crisis began (Centraal Bureau voor de Statistiek, 2016).

Besides this increase in the amount of part time employees, the differences between the labour contracts of full timers and part timers could had an influence on job crafting. It seemed logical that because part timers were less present at work than full time employees, they could think job crafting is too difficult to realize. Another suggestion could be that part timers did not get the opportunity to craft their job, since they felt less connected to the organization. When these employees had intrinsic motivation to change their job, but they did not succeed in changing it, it could gave them feelings of stress (Berg et al., 2010). For organizations, it could be a disclosure to find out if part timers were more inclined to stress just because their job crafting did not work out. On the other hand, job crafters that work 35 hours per week or less, could also experience less stress. This research investigated the effect of working part time on the positive relation between job crafting and stress.

2.5 Hypotheses

For this research, job crafting was defined as the psychical and cognitive changes an employee makes in the task or relational boundaries of their work (Wrzesniewski & Dutton, 2001). There were three types of job crafting, these are task crafting, relational crafting and cognitive crafting. According to previous researches, there was a relationship between job crafting and well-being (Tims et al., 2013). This relationship could be positive and negative as well, because it could enhance work engagement, but, moreover, it also could give employees stress if job crafting did not work out (Berg et al., 2010; Ghitulescu, 2006). This positive relationship between job crafting and stress was examined in this research. Because working part time have increased among the Dutch employees, being a part time employee could affect the positive relationship between job crafting and stress. In this case, the amount of working hours, meaning working part time, was a moderating variable, job crafting (conceptualized as task crafting, cognitive crafting and relational crafting) was the independent variable and stress was the outcome, or the dependent variable. This was built into a conceptual model, finding below.

Fig. 1

A Model for Job Crafting

Part time/ full time Stress Task crafting Relational crafting Moderating variable

Job crafting General _effect

Cogntive crafting

According to this conceptual model, the next research question followed: “How is working part time effecting the relationship between job crafting, task crafting, relational crafting, and cognitive crafting and the perceived level of stress of employees?”. To answer this research question, hypotheses were set up and tested with a survey-based study:

Hypothesis 1: There is a positive relation between job crafting and stress. Hypothesis 1a: There is a positive relation between task crafting and stress. Hypothesis 1b: There is a positive relation between relational crafting and stress. Hypothesis 1c: There is a positive relation between cognitive crafting and stress.

Hypothesis 2: The positive relation between job crafting and stress is moderated by the amount of working hours, so that this positive relation is stronger for part timers compared to full timers.

Hypothesis 2a: The positive relation between task crafting and stress is moderated by the amount of working hours, so that this positive relation is stronger for part timers compared to full timers.

Hypothesis 2b: The positive relation between relational crafting and stress is moderated by the amount of working hours, so that this positive relation is stronger for part timers compared to full timers.

Hypothesis 2c: The positive relation between cognitive crafting and stress is moderated by the amount of working hours, so that this positive relation is stronger for part timers compared to full timers.

3. Methodology

This paragraph contained the research design including the procedure of the research, some information about the sample and the measures that were used to assess the data.

3.1 Procedure

In this research the effect of part timers on the relationship between job crafting and stress was studied. By doing a quantitative research, with a combination of a deductive and inductive approach, this was realized. This meant that this study was built on already existing theories, which, in turn, led to a new point of focus in the research field (Saunders, Lewis, & Thornhill, 2012). The literature review showed that job crafting could result in positive and negative outcomes as well.

Moreover, the existing literature showed that there was a relation between job crafting and stress (Berg et al., 2010). But these previous findings did not examine what the effect was of working part time on this positive relation between job crafting and stress. That was where this research was focused on, since none of the previous studies have investigated this yet.

The data for this study was gathered by standardized, self-completed questionnaires that were sent by e-mail to employees and employers. This e-mail contained the aims of the study, the emphasized confidentiality and the link to one of the questionnaire. These Web-based questionnaires were conducted by Qualtrics. There were two different questionnaires, one questionnaire for employees and the other for the employee’s manager. The questionnaire for employees consisted of 26 questions in Dutch consisting of statements about the influence of human resource management, their jobs and their work environment on themselves (Appendix A). Managers who wanted to participate in the research, had to answer 16 questions including rating the participating employee’s performance. This meant that data was collected in dyads what included one or two employees and their manager. To link the questionnaires that employees filled in with those of managers, every respondent needed to fill in their personal code in one of the questions of the questionnaire. When responses were still missing, a reminder was sent by e-mail to the respondents who did not fill in the questionnaire yet, or did not finish it. No confidential data of the employees was asked in the questionnaire, because the results were saved by the personal code respondents had to fill in. Collecting the respondents was done at one particular time what indicated that the time horizon of the research was cross-sectional (Saunders et al., 2012).

The research group, that consisted of five researchers, all collected dyads. Because this research was focused on the relation between job crafting and the perceived stress of employees, only data of respondents that were employees were usable. The data from the questionnaire filled in by managers were therefore excluded from the dataset. With this dataset the positive relation between job crafting and the perceived stress of employees was tested by doing hierarchical multiple regressions in SPSS. These regressions gave insight in the prediction of stress, based on its relationship with job crafting or one of the types of it. Besides, moderation tests were also executed with the dataset of employees. These tests are executed by Process, an external macro developed by A.F. Hayes. The output of these moderation tests aimed to understand the effect of part timers on the positive relationship between job crafting and stress.

3.2 Sample

convenience sampling technique. This technique meant that respondents were selected haphazardly, because they were easily available (Saunders et al., 2012). The ease to obtain respondents in this way, could have influenced the results that were beyond the control. This, result in that the results only applies to the respondents who filled in the questionnaire. On the other hand, the respondents worked in different sectors and organizations what could mean that the results could be applied to different types of employees.

Part of this diversity of respondents, was the sector where employees work in, ranging from hospitality to bank to supermarket. Also, the respondents were all Dutch since the researchers and the questionnaire were in Dutch, despite the possibility to get the questionnaire in English. Besides, there were some other variations among the respondents, for example the age ranging from 17 years till 63 years. The sample consisted of 118 employees (N = 118) what is a reliable sample size (N > 30). The response rate was 88% indicating that 134 employees were asked to fill in the questionnaire. From these 118 employees, about 55% were women and about 45% were men. The average age among the employees were 32.02 years (SD = 12.09). The mean tenure were 6.32 years (SD = 8.35), while the maximum of tenure were 40 years. Outstandingly, half of the respondents worked 2.5 years or less for their current organization. This also could have influenced the results. The average amount of working hours among the respondents were 28.43 hours per week (SD = 12.37). This was less than working full time what meant 35 working hours per week or less than that (Hupkens, 2012). No one in the sample worked 35 hours per week according to their labour contract, so among the part timers the limit were 32 hours per week and included 53% of the respondents. The other 47% work 36 hours or more per week and were considered as full timers.

In the end, the sample consisted of relative young employees who have not worked a lot of years at their current organization. Also, more than half of the respondents worked part time what confirmed the increase in part time jobs.

3.3 Measures

To measure how part timers affect the relationship between job crafting and stress, several measurements were used in this study. The control variables, the amount of working hours, job crafting disentangled as well as in total and stress were discussed below.

3.3.1 Control variables and Working hours

In this research, the control variables were gender, age and tenure when doing hierarchical multiple regression by SPSS. Gender was a dummy variable meaning that a male respondent was coded

with a zero and a female respondent with a one. Open questions were used to measure the age and tenure of the respondents. Also for the amount of working hours of the respondents, an open question was used to gather it. In this way the amount of working hours was a numerical variable, but since the focus in this research was on part timers, the variable was made categorical. This meant that every respondent working from 0 till 35 hours per week according to their labour contract, were coded with a zero meaning that is was considered as a part time employee. Respondents working more than 35 hours per week, or full timers, were coded with a one. From that point, the variable was considered as a dummy variable labelled as: part time versus full time. This variable was used for doing hierarchical multiple regression as one of the predictor variable as well as the moderator in the moderation test done by Process.

3.3.2 Task crafting

The measurement of job crafting is the combination of the measurements of task, relational and cognitive crafting. The scales of these three types of job crafting were based on a previous study, although it had not been validated yet (Bindl et al., 2014). From this scale skill crafting was excluded from the measurements, because the research of Wrzesniewski and Dutton (2001), where this research is based on, also did not focus on skill crafting. The scales for task, relational and cognitive crafting were used for doing hierarchical multiple regression to investigate whether they had a linear relation with stress. Besides, they were also used for doing the moderation tests as independent variables predicting stress.

Task crafting, as one part of job crafting in total, was measured by seven items having a Cronbach’s Alpha of .76. No item of the scale was removed, because it would not benefit the reliability of the scale when removing one item. Also, all items were strongly enough correlated to the total score from the questionnaire, because all of the corrected item-total correlation were above .30 (Appendix B, Table 1). Task crafting included four items of enhancing task crafting and three items of limiting task crafting. Enhancing task crafting consisted of statements such as “I changed my tasks so that they were more challenging” that was translated in Dutch in the questionnaire. A translated example of a statement of limiting task crafting was “I actively reduced the scope of tasks I worked on”. Employees indicated to what extent they agreed or disagreed with the seven statements with a seven-point Likert scale ranging from 1 (totally disagree) to 7 (totally agree).

3.3.3 Relational crafting

Relational crafting that formed the second type of job crafting, was measured by the scale given by Bindl et al. (2014). This scale included four items that were measuring enhancing relational

crafting. An example that was English from the beginning, but translated in Dutch in the questionnaire, was “I tried to spend more time with a wide variety of people at work”. Again, employees indicated themselves to what extent they agreed with the statements. Responses were given on a seven-point Likert scale that ranged from 1 (totally disagree) to 7 (totally agree). The Cronbach’s Alpha of this scale was .73 indicating that the scale was a reliable measurement. However, removing the first item of the scale would increase the Cronbach’s Alpha to .80 (Appendix B, Table 2). This item included the following statement: “I actively took on more tasks in my work”. Nevertheless, this item was not deleted from the scale, because that concluded that data would be removed while the corrected item-total correlation of this item was above .30 just as the other items (Appendix B, Table 2). This meant that the items correlated to the total score from the questionnaire.

3.3.4 Cognitive crafting

Also cognitive crafting that belonged to job crafting in total, was measured by the scale that was introduced by Bindl et al. (2014). This scale included six statements about enhancing and limiting cognitive crafting and had a Cronbach’s Alpha of .79. This meant that the measurement was reliable. Despite that Cronbach’s Alpha could increase by removing the last item of the scale, it remained in the scale because of the corrected item-total correlation which was above .30 and because otherwise data would be removed (Appendix B, Table 3). The three items about enhancing cognitive crafting consisted of statements such as “I thought about new ways of viewing my overall job”. An example of one of the three limiting cognitive crafting statements was “I focused my mind on the best parts of my job, while trying to ignore those parts I didn’t like”. Again, employees had to indicate whether they agreed with these statements. A seven-point Likert scale was used with responses ranging from “totally disagree” (1) to “totally agree” (7).

These items of the scales of task, relational and cognitive crafting together, were used to measure job crafting in overall. This 17-item scale based on the research of Bindl et al. (2014) included thus enhancing and limiting statements. This scale that measured job crafting was reliable just as the types of job crafting separately, because its Cronbach’s Alpha was .85. Just as for the three types of job crafting, no item was removed from the scale, because it would remove data and it would not increase the reliability of the scale (Appendix B, Table 4).

3.3.5 Stress

In this research stress was measured by the four-item scale based on an earlier study (Motowidlo, Packard, & Manning, 1986). This scale included statements as for example, “my job is extremely

stressful,” and “I almost feel never stressed at work”. Its internal consistency reliability (alpha) was .86 according to the research of Motowidlo, Packard and Manning (1986). The Cronbach’s Alpha of this scale was .75. This included that the scale was reliable and moreover, it would not increase when removing an item from the scale (Appendix B, Table 5). The employees had to indicate to what extent they agreed to these statements again. Also, a seven-point Likert scale with responses ranging from 1 (totally disagree) to 7 (totally agree) was used.

4. Results

In this paragraph the results from the multiple test that were executed, were interpreted. First, some descriptive statistics of the sample were discussed. This was followed by analyses of the output of the correlations, hierarchical multiple regressions and moderation tests. By analysing these outputs, the hypotheses were tested and supported or rejected.

4.1 Descriptive statistics

After collecting the data from the employees, the data of the sample were exported to run several tests by SPSS. Before testing the hypothesized model, the means and standard deviations were executed for all variables. These descriptive statistics of the sample (N = 118) were displayed in Table 1. This table included also the reliability of the scales that were used in this research. Only, the control variables (gender, age and tenure) and the moderating variable including working part time versus full time, were not tested on reliability since these variables were measured by one measurement. Concerning the Cronbach’s Alpha of the other variables, all of them are above .70 what included that all measures are reliable just as stated before. Therefore, no item was removed from any of the scales.

Before computing the correlations of the variables, the scale means were computed per variable. Since the amount of working hours was changed into a categorical variable, the scale mean could not be computed. Concerning the other variables, job crafting in total, the three types of job crafting separately and stress, the scale means were computed. This meant for example for task crafting, both enhancing and limiting items of the scale were taking into account when averaging out the scale. This applies also to the other variables, because no item was excluded from any scale. This resulted in new variables that were used for the other tests.

With the scale means of the variables, the correlations were computed and are shown in Table 1. Interpreting these correlations, there was a significant tendency to relation between relational crafting and age, r(116) = -.25, p < .01. This meant that when an employee craft their job

more relationally, they probably be a relatively younger employee since the correlation is negative. Another significant relationship was the relation between stress and the working part time versus full time. The Pearson correlation coefficient of this relation was r(116) = .34, p < .01 and meant that there was a tendency to relation. This finding concluded that when an employee’s stress level increased, the amount of working hours increased too, which indicated that it is inclined to being a full timer. Moreover, stress had a significant tendency to relate with gender, r(118) = .25, p < .01. Again this meant that when an employee felt more stressful, they rate higher at the measurement of gender that inclined to being female. Concerning job crafting or task or cognitive crafting, there is no significant correlation with stress, working part time or one of the control variables (p > .05). Furthermore, the Pearson correlation coefficient of stress and job crafting or one of the types of it, were not only non-significant (p > .05), but they were also very small which could indicate the absence of the relation.

4.2 Regressions

After getting a first insight in the data by computing the correlations, the first hypotheses were tested by doing hierarchical multiple regressions. Hierarchical multiple regressions were performed to investigate the ability of job crafting, the three types of job crafting separately, and part time employees to predict the level of perceived stress among employees, after controlling for gender, age and tenure.

Table 1

Means, Standard Deviations, Correlations

Variables M SD 1 2 3 4 5 6 7 8 9 1. Gender 1.55 0.50 - 2. Age 32.03 12.09 -.11 - 3. Tenure 6.32 8.35 .04 .64** _- 4. Part-time vs. full-time 0.47 0.50 -.26** _.23* _{.16 -} 5. Task Crafting 2.55 0.59 -.08 -.07 -.09 .12 (.76) 6. Relational Crafting 3.29 0.61 .17 -.25** _{-.16 -.10 .37}** _(.73) 7. Cognitive Crafting 2.92 0.68 .13 .05 -.03 .01 .37** _.42** _(.79) 8. Job Crafting 2.86 0.49 .07 -.09 -.10 .03 .80** _.69** _.80** _(.84) 9. Stress 3.79 1.20 .25** _{.08 .14 .34}** _{.08 .12 .09 .12 (.75)}

*. Correlation is significant at the .05 level (2-tailed). **. Correlation is significant at the .01 level (2-tailed).

4.2.1 Predicting stress by part timers and job crafting

In hypothesis 1, it was proposed that there was a positive relation between job crafting and stress. The first step of the hierarchical multiple regression included adding three predictors that were gender, age and tenure (see Table 2a). The model was statistically significant F(3, 107) = 2.85, p < .05 and explained 7% of the variance in the perceived stress of employees (Appendix B, Table 7). In the second step, the variables of working part time versus full time and job crafting in total, were added to the regression model. Task crafting, relational crafting and cognitive crafting were not added to this regression model yet, because it would lead to multicollinearity between the types of job crafting and job crafting in total. The results of this hierarchical multiple regression were displayed in Table 2a. This model was statistically significant F(5, 105) = 6.54; p < .001 and explained 24% of the total variance (Appendix B, Table 6). In this first model of the regression, only gender was found significant related to stress (β = .23, p < .05). The adjustment of adding the variable of working part time versus full time and job crafting in total, explained additional 20% variance in stress, after controlling for gender, age and tenure (R² Change = .201; F(2, 105) = 11.25; p < .001).

Analysing Table 2a, only two out of the five predictor variables were statistically significant, these included gender (β = .32, p < .01) and working part time versus full time (β = .41, p < .001). Working part time versus full time had a higher Beta value compared to gender what meant that working part time versus full time had got more influence on stress than the influence gender had on stress. This concluded that when the variable of working part time versus full time increased with one standard deviation, indicating that the labour contract is inclined to working full time, stress also increased with .41-unit with the other variables held constant. Concerning hypothesis 1, job crafting was not statistically significant related to stress (p > .05) what could indicate that there is no relationship between job crafting and stress, just as what appeared from the correlation. Regarding to this finding, hypothesis 1 was rejected.

Table 2a

Hierarchical Regression Model of Stress Predicted By Working Part time and Job Crafting

R R² R² Change B SE β t Step 1 .27 .07* Gender .54 .22 .23* _2.41 Age .01 .01 .06 .47 Tenure .02 .02 .11 .89 Step 2 .49 .24*** _.20*** Gender .76 .21 .32** _3.58 Age .00 .01 -.00 -.02 Tenure .01 .02 .09 .82

Part time vs. full time .96 .21 .41*** _4.54

Job Crafting .21 .21 .09 .99

Note: Statistical significance: *_p<.05; **_p<.01; ***_p<.001.

4.2.2 Predicting stress by part time employees, task, relational, and cognitive crafting

After this regression, another hierarchical multiple regression was done to test hypotheses 1a, 1b and 1c. These hypotheses proposed that task crafting (1a), relational crafting (1b) and/or cognitive crafting (1c) were positive related to stress. The regression to test these hypotheses, included the variables of task, relational and cognitive crafting instead of the variable of job crafting in total. Just as the hierarchical multiple regression before, the first model was statistically significant (F(3, 107) = 2.85; p < .05), because it included the same control variables (Appendix B, Table 7). And again, gender was found significant related to stress (β = .23, p < .05) It explained 7% of the variance in the perceived stress of employees. When adding the three types of job crafting to the regression model, it was again statistically significant F(7, 103) = 4.66; p < .001 and explained 24% of the total variance (Appendix B, Table 7). The adjustment of adding the variable of working part time versus full time and the three types of job crafting, explained additional 19% variance in stress, after controlling for gender, age and tenure (R² Change = .189; F(4, 103) = 5.65; p < .001).

Analysing Table 2b, only two out of the seven predictor variables were statistically significant, these were gender (β = .31, p < .01) and working part time (β = .42, p < .001). Just as the hierarchical multiple regression before, the same significant relations were found as well as the higher Beta value of working part time versus full time compared to gender. This again meant that working part time had got more influence on stress than the influence gender had on stress. In other words, an increase of one standard deviation of the variable of working part time versus full time increased with one standard deviation, which meant a labour contract more inclined to working full time, leads to .42 standard deviation increase in predicting stress. Besides, just as the regression

before, not one type of job crafting was found statistically significant to stress (p > .05). Moreover, the Beta values of the types of job crafting were very small what could indicate that there is no relationship between (the types of) job crafting and stress. Just as hypothesis 1, also hypotheses 1a, 1b and 1c are rejected meaning that there is no positive relation between (one of the types of) job crating and stress.

4.3 Moderation effect

To test what the effect was of working part time on the relationship between job crafting and stress, Process, an external macro developed by A.F. Hayes, was used. To run Process, some of the variables were standardized. These were the variables of task crafting, relational crafting, cognitive crafting and job crafting in total.

4.3.1 Effect of part timers on the relation between job crafting and stress

In hypothesis 2 was proposed that the positive relation between job crafting and stress was moderated by the amount of working hours, so that this positive relation was stronger for part timers compared to full timers. To decide whether this hypothesis could be supported, the regression model between job crafting and stress moderated by working part time was executed by Process. The results are displayed in Table 3a. This regression model explained 16% of variance in the perceived level of stress of employees.

Table 2b

Hierarchical Regression Model of Stress Predicted by Working Part time, Task, Relational and Cognitive Crafting R R² R² Change B SE β t Step 1 .27 .07* Gender .54 .22 .23* _2.41 Age .01 .01 .06 .47 Tenure .02 .02 .11 .89 Step 2 .49 .24*** _.19*** Gender .74 .22 .31** _3.39 Age .00 .01 .02 .13 Tenure .01 .02 .09 .78 Part-time vs. full-time .97 .22 .42*** _4.52 Task Crafting .05 .19 .02 .24 Relational Crafting .189 .21 .01 .92 Cognitive Crafting .00 .17 .00 .01

Based on these results, the regression coefficient of XM is c3 = -.39, but is not statistically

different from zero t(107) = -1.83, p > .05. Thus, when job crafting increased with one unit, the difference in stress between part timers and full timers decreased by .39 units. In other words, this finding could indicate there is no difference in the level of stress between part timers and full timers when crafting their job, although it is not significant.

A conditional effect that was found in Table 3a, was the relation between job crafting and stress when working part time versus full time was zero, meaning that employees work part time. The conditional effect was c1 = .29 what included the estimated increase of stress between two part

timers who differ by one unit of job crafting. So, when a part timer crafted their job one more unit than a part time colleague, the employee experienced more stress than the colleague. Only, this result was not found statistically different from zero (SE = .16, CI: -.02 to .60). An effect that was found significant, was the conditional effect of the moderating variable (c2 = .82, p < .001). This

contained that when job crafting would be average among part timers and full timers, the estimated difference in stress among them would increase by .82 units (SE = .29, CI: .40 to 1.25). So, when two employees crafted their job on an average level and the variable of working part time versus full time increased with one unit, indicating that the employee worked full time, stress increased with .82 units. Based on the previous findings, the interaction between job crafting and working part time was not found significant and that is why hypothesis 2 was rejected.

Table 3a

First Regression Model of Moderator

Coefficient SE t p

Intercept i1 3.39 .14 23.92 .0000

Job Crafting (X) c1 .29 .16 1.85 .0678

Part time vs. full time (M) c2 .82 .22 3.83 .0002

Job Crafting*Part time vs full time (XM) c3 -.39 .21 -1.83 .0707

R² = .16, p < .001 F (3, 107) = 5.92

4.3.2 Effect of part timers on the relation between task crafting and stress

After doing the moderation tests for the relation between job crafting in total and stress, the same was done for the three types of job crafting as well. In hypothesis 2a was suggested that the positive relation between task crafting and stress was moderated by the amount of working hours, so that this positive relation was stronger for part timers compared to full timers. Again, for testing this hypothesis Process was used with task crafting as independent variable, stress as dependent

variable and part time versus full time as the moderator. The results are shown in Table 3b below. This regression model explained 14% of the variance in stress among employees.

According to these results, the interaction regression coefficient XM is c3 = -.29, but this

interaction term was not found statistically different from zero, t(107) = -1.45, p > .05. The negative coefficient indicated that the stress among part timers who craft their job, was not different than the stress of full timers who craft their job. But, this relationship was not significant what meant that working part time or full time did not affect the relationship between task crafting and stress.

Also the conditional effect of the relation between task crafting and stress when working part time versus full time counted zero, meaning working part time, was not found significant (effect c1 = .19, SE = .16, CI: -.13 to .50). However, this finding would meant that when one part

time employee craft their job on tasks one more than a part time colleague, he or she experienced .19 more stress than the colleague who did not craft their tasks more. Concerning the conditional effect of working part time on stress when task crafting was average, the difference between a part timer and full timer was c2 = .80 and is significant, p < .001 (SE = .22, CI: .37 to 1.25). In other

words, when task crafting of two employees would be average and the variable of working part time versus full time increased with one unit, meaning that the employee worked full time, this led to an increase in stress of .80 units. Despite this findings, there is no significant interaction between task crafting and working part time and that is why hypothesis 2a was rejected.

Table 3b

Second Regression Model of Moderator

Coefficient SE t p

Intercept i1 3.41 .14 23.74 .0000

Task Crafting (X) c1 .19 .16 1.18 .2422

Part time vs. full time (M) c2 .80 .22 3.69 .0004

Task Crafting*Part time vs full time (XM) c3 -.29 .20 -1.45 .1497

R² = .14, p < .01 F (3, 107) = 4.96

4.3.3 Effect of part timers on the relation between relational crafting and stress

In hypothesis 2b was proposed that the positive relation between relational crafting and stress was moderated by the amount of working hours, so that this positive relation was stronger for part timers compared to full timers. The results that were based on the output of Process, were displayed in Table 3c. This moderation model explained 14% of the variance in the stress of employees.

Regarding Table 3c, the regression coefficient of the interaction, XM, was c3 = -.11, but it

negative indicating that there is no difference in stress between part timers and full timers who craft their job, although it was not significant.

Just as the other moderations done earlier, there was one significant and one non-significant conditional effect. The conditional effect that was not significant, was the relation between relational crafting and stress when part time versus full time was zero, meaning only for employees that work part time (effect c1 = .22, SE = .13, CI: -.05 to .48). This finding suggested that a part

timer who increased relational crafting with one more unit, perceived .22 units of stress more compared to a colleague who did not increase relational crafting.

The significant conditional effect included the relation between working part time and stress when relational crafting is average. According to the data, the difference in stress between part timers and full timers under the circumstance that relational crafting was average, was c2 = .84 (SE

= .22, CI: .41 to 1.27). This finding suggested that when two employees craft their job relationally on an average level, a full timer perceived .84 more units of stress than a part time employee would perceive. In the end, these findings resulted in no support for hypothesis 2b, because there is no significant moderating effect of working part time on the relationship between relational crafting and stress.

Table 3c

Third Regression Model of Moderator

Coefficient SE t p

Intercept i1 3.37 .14 23.37 .0000

Relational Crafting (X) c1 .22 .13 1.62 .1086

Part time vs. full time (M) c2 .84 .22 3.91 .0002

Relational Crafting*Part time vs. full time (XM) c3 -.11 .22 -.52 .6059

R² = .14, p < .01 F (3, 107) = 5.69

4.3.4 Effect of part timers on the relation between cognitive crafting and stress

In the last hypothesis, hypothesis 2c, it was suggested that the positive relation between relational crafting and stress was moderated by the amount of working hours, so that this positive relation was stronger for part timers compared to full timers. This hypothesis was tested by Process whose results are shown in Table 3d. This model explained 15% of the variance in stress of employees.

Concerning the moderating effect of working part time on the relation between cognitive crafting and stress, the interaction term XM = c3 = -.57, but just as (the other types of) job crafting

the effect was not statistically different from zero, t(107) = -1.82, p > .05. This negative coefficient of the regression suggested that there is no difference in stress between part timers and full timers

when crafting their job cognitively. Obviously, besides that it non-significant, all interaction coefficients from the moderation tests were negative indicating that the relationship between (one of the types of) job crafting and stress are not effected by working part time. These findings contributed to the rejection of hypothesis 2c.

Nevertheless, there were also two conditional effects found in the output. The first conditional effect included the relation between cognitive crafting and stress among part timers, because the variable of working part time versus full time counted as zero. The regression coefficient of this conditional effect is c1 = .23, but was not significant (SE = .13, CI: -.03 to .49).

However, this would mean that when one part timer increased their cognitive crafting with one more unit than a colleague, this led to an increase of .23 units of the perceived stress of the part time who increased cognitive crafting.

The other conditional effect of the relation between working part time and stress when cognitive crafting was average, had got a regression coefficient of c2 = .83 and was significant (SE

= .22, CI: .40 to 1.25). This meant that when the variable of working part time versus full time increased with one unit, so the employee worked full time, this employee would experience .83 more units of stress than a part timer, provided that cognitive crafting was average. Comparing all the significant conditional effects, when task crafting was done on average, stress increased the most, although the differences are very small. In overall, full timers experienced more stress than part timers when (one of the types of) job crafting was average. Also from the earlier analysis, the relation between working part time and stress were found negative. This indicated that employees experienced more stress when the amount of working hours increased, or working full time. Table 3d

Fourth Regression Model of Moderator

Coefficient SE t p

Intercept i1 3.38 .14 23.46 .0000

Cognitive Crafting (X) c1 .23 .13 1.78 .0778

Part time vs. full time (M) c2 .83 .22 3.86 .0002

Cognitive Crafting*Part time vs. full time (XM) c3 -.40 .22 -1.82 .0712

R² = .15, p < .01 F (3, 107) = 5.51

5. Discussion

This present study examined the relationship between job crafting and the perceived level of stress of employees. Furthermore, this study investigates what the effect is of working part time on this

relationship. The hypothesized model included that when crafting a job, or at least one of its types (task, relational and cognitive crafting) the perceived level of stress of employees increases. Moreover, this positive relationship will be stronger among part timers compared to full timers. However, the results of this study are not consistent with the hypotheses. Whether an employee work part time or full time, there is no significant relation found between (one the types of) job crafting and stress. In the following section the most important contributions of the present study will be discussed.

5.1 Theoretical contributions

The first contribution of this study is that it provides more insight into the negative outcome of stress that job crafting, or one of the three types of it, can lead to for individuals. This study has found that job crafting is not significantly positive related to stress, neither is task crafting, relational crafting and cognitive crafting. This means that there is no evidence for experiencing more stress when crafting a job, or crafting one type of it. This result is contradicting the findings of existing literature that stated that when job crafting do not work out, because of strong situations or high work pressure, it could lead to stress or, in turn, a higher level of burnout (Berg et al., 2010; Tims et al., 2013; Wrzesniewski & Dutton, 2001). However, this study did not investigate the strong work situations or high work pressure of respondents, meaning that the respondents perhaps did not find themselves in such circumstances, so it could not lead to stress as the literature stated.

The second contribution of this study is that it compares part time employees with full time employees regarding their job crafting and their level of stress. This study has found that when an employee craft their job, no matter what type of it, the level of stress did not depend on working part time or full time, although this finding was not significant. Hypothesized is that it would matter and that part timers will experience a higher level of stress when crafting their job compared to full timers. When researchers will be aware of the work situation of the respondents, it might give an insight in the effect of part timers who craft their job and their level of stress.

A third contribution of this study is that it provides insight in the relationship between working part time and the perceived level of stress of employees, under the circumstances that job crafting was done on an average level. Concerning the literature, employee well-being and, in turn, mental health could increase by crafting the job, although it depend on the extent of engagement in job crafting (Slemp & Vella-Brodrick, 2013). This study has found a significant conditional effect meaning that when two employees who crafted their job on tasks, relationally or cognitively, full timers experience more stress than part timers do. Besides, there is found a significant relation

between working more hours, so that it is more inclined to working full time, and higher perceived level of stress. Comparing the types of job crafting and job crafting in total, task crafting will increase stress the most, although the differences are very small. These findings could indicate that the respondents in this research do not craft their job fully, or they craft their job not at all. Future research need to be aware of that respondents craft their job on a high level meaning that they fully craft their job, or at least of one the types of it, to study whether stress is higher among part timers than full timers.

5.2 Limitations

Concerning the limitations of this study, there are some that has to be acknowledged. First, respondents were collected with the non-probability convenience sampling technique. This sampling technique consists of selecting cases that are easily available. It limits this study, because the sample in this study is easily chosen meaning that the results only applies to this sample. The findings are therefore not generalizable (i.e., external validity) to other employees. However, concerning the sample, it consists of respondents from all kinds of sectors which suggests that the results are valid for organizations from multiple sectors.

A second limitation is the time horizon of this study which is cross-sectional. Task crafting, relational crafting, cognitive crating and so job crafting was measures at one particular time, because respondents have filled in the questionnaire at one point of time. Consequently, it does not allow the inference of causal relationship and therefore, the possibility of having reverse causality cannot be ruled out.

The third limitation includes the measurements of task crafting, relational crafting, cognitive crafting and so, job crafting. These scales, that were based on the research of Bindl et al. (2014) has not been validated yet what could be seen as a limitation. However, concerning the Cronbach’s Alpha’s of these scales, there is no reason to question their quality. All of the variables have a Cronbach’s Alpha above .70 which includes that they are reliable. However, the reliability could be improved when items are added that are correlated with the total score of the questionnaire what would benefit the quality of the scales.

Also the last limitation of the study concerns the measurements of the variables. In the questionnaire, respondents need to indicate to what extent they agree with a given statement. These measures are called self-reported measures which concludes that the predictor variables as well as the criterion variable are provided with measures from the same person. So, an employee that rated their level of job crafting, rated their level of stress too. This limits the study, because it might