THE IMPACT OF A PROACTIVE PERSONALITY ON DAILY WORK
ENGAGEMENT, AND THE ROLE OF JOB CRAFTING AND I-DEALS.
University of Amsterdam Master of Business Administration Track Leadership & Management Master Thesis
Author: Kim van Beek
Student number: 11363851 Supervisor: E. Federici 2nd Supervisor: C.T. Boon
2
STATEMENT OF ORGINALITY
This document is written by Student Kim van Beek 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 TABLE OF CONTENT ACKNOLEDGEMENT 5 ABSTRACT 6 INTRODUCTION 7 THEORETICAL BACKGROUND 9
Proactive personality and work engagement 9
Job crafting vs. Idiosyncratic deals 12
The mediating role of daily job crafting 13
The mediating role of Idiosyncratic deals 17
METHOD 22
Research design 22
Sample 22
Measures 24
General questionnaire (between-level) 24
Dairy study (within-level) 24
RESULTS 27
Analytical strategy 27
Exploratory Factor Analysis 27
Descriptive statistics 31
Hypothesis testing 32
Additional analysis 35
DISCUSSION 39
Limitations 40
Practical implication and future research 42
CONCLUSION 43
REFERENCES 44
APPENDIX 51
Appendix I – Surveys 51
4
LIST OF TABLES
Table 1: Comparison of two concepts of work design and its dimensions 13 Table 2: Demographic characteristic profile of respondents 23 Table 3: Summary of exploratory factor analysis results for measurement scale 29 daily job crafting
Table 4: Summary of exploratory factor analysis results for measurement scale 31 I-deals
Table 5: Means, Standard Deviations, and Correlations among Variables 32 Table 6: Estimated Coefficients of the Parallel Mediation Model (day-level) 34 Table 7: The (In)direct Effects of the Parallel Mediation Model (day-level) 35 Table 8: Estimated Coefficients of the Parallel Mediation Model (general-level) 37 Table 9: The (In)direct Effects of the Parallel Mediation Model (general-level) 38
LIST OF FIGURES
Figure 1: Hypothesised research model 9
Figure 2: Results of the hypothesised research model (day-level) 35 Figure 3: Results of the hypothesised research model (general-level) 36
5
ACKNOWLEDGEMENT
This thesis could not be finished without the help and guidance of others. Therefore, I would like to take the opportunity to express my gratitude and deepest appreciation to everyone who have helped me during this research project.
First of all, I would like to express my appreciation to my supervisor at the University of Amsterdam, Eloisa Federici, for guiding me through the entire process of this research project. Apart from my own efforts, her useful remarks, comments, general support and caring guidance largely contributed to the successful completion of this research. Furthermore, the same accounts for my fellow students in the “job crafting team”, by helping each other out if needed and by collecting data together we made our research projects valuable.
Second of all, I would like to thank all survey participants who took the time to fill out the questionnaire and dairy survey for five consecutive work days. Their participation made it possible to conduct this research.
To round off, I would like to thank my family, friends and fellow Master students for their help, understanding and support throughout my entire studies and especially throughout this whole Master’s year. All in all, without my loved ones nothing of this would have happened and I would not be the proud and dedicated person I am today.
6
ABSTRACT
The aim of this study was to compare two parallel mechanisms of proactive behaviours at work, daily job crafting and I-deals, in explaining the relationship between proactive personality and daily work engagement. These two bottom-up concepts of work design can be found in recent literature, which both reflect the proactive behaviours of today’s employees. This quantitative study collected data by using a general questionnaire and a diary study of five consecutive work days. The hypotheses were tested with a sample of 111 Dutch employees, working more than three days a week, from wide range of sectors (e.g. services, health care, education, transport & logistics). The results found evidence of the relationship between employees with a proactive personality and daily work engagement. Thus, the present results enhance our understanding of the antecedents of daily work engagement, making the use of repeated real-time measures the main strength of this study. Unexpectedly, this study did not find evidence of the parallel mediation of daily job crafting and I-deals assuming the relationship between proactive personality and daily work engagement. However, one step is made in the exploration of the investigated variables. Further research, for example on a longitudinal basis, needs to be done to gain more knowledge on the links between proactive personality, daily work engagement and proactive work behaviours as daily job crafting and I-deals.
Key words: proactive personality, proactive work behaviours, job crafting, Idiosyncratic deals, work engagement, day-level job crafting, day-level work engagement, repeated measure design
7
INTRODUCTION
Over the years, researchers have used the design of jobs as a starting point to examine how employees experience their work (Berg, Wrzesniewski & Dutton, 2010). Traditionally, research focused on the top-down processes of managers designing jobs for their employees and the job design theory (Campion & McClelland, 1993; Grant, Fried, & Juillerat, 2011). Organisations were supposed to offer their employees sufficient job resources, including social support, feedback, and skill variety. Research indeed argued that managers can influence employees’ job demands and resources (Nielsen, Randall, Yarker & Brenner, 2008; Piccolo & Colquitt, 2006), and may indirectly influence employee’s work engagement and performance (Harter, Schmidt & Hayes, 2002). However, scholars have recognised the important role of employees in designing their own jobs (Black & Ashford, 1995; Miner, 1987), highlighting the proactivity part in their efforts (Grant & Ashford, 2008). Bakker and colleagues (2012) for example, have shown that employees with a proactive personality are most inclined to change their work environment in a proactive way, by mobilising job demands and job resources, which in turn facilitate work engagement. Therefore, it may be equally important that employees organise their own job, to respond to the complexity of contemporary jobs and deal with the needs of today’s work environment (Demerouti, 2014). While this environment changes quickly, it is particularly important for employees to show proactive behaviour and create their own work environment (Bakker, Tims & Derks, 2012).
Two bottom-up concepts of work design can be found in recent literature, which both reflect the proactive behaviours of today’s employees: job crafting and idiosyncratic deals. Both these behaviours increasingly occur due to the changing nature of work (e.g. flexible human resource practices, virtualisation) (Hornung, Rousseau & Glaser, 2009). It is relevant for organisations to know on which terms these behaviours are initiated and how they differ from each other. The question is to which extent job crafting and I-deals are any different, contributing to and enriching each other. More specifically, the simultaneous mechanisms of job crafting and I-deals, in explaining the link between a proactive personality and work engagement have not been investigated yet. This study aims to compare two parallel mechanisms of proactive behaviours at work, in explaining the relationship between proactive personality and work engagement.
On the one hand, Wrzesniewski and Dutton (2001) have complemented traditional top-down approaches of job design with the concept of job crafting, defined as “the physical and
8
cognitive changes individuals make in the task or relational boundaries of their work” (p. 179). Employees can take action to adjust their work to better match their personal needs, strengths, and interests (Berg, Dutton & Wrzesniewski, 2008; Wrzesniewski, Berg & Dutton, 2010), not specifically authorised by the manager (Ross, Greene & House, 1977). On the other hand, idiosyncratic deals (or I-deals) constitute a middle path between top-down work redesign and a single employee’s private efforts to craft a job (Hornung, Rousseau, Glaser, Angerer & Weigl, 2010). I-deals, in general, are employment terms that employees negotiate for themselves, taking multiple forms from flexible schedules to career development (Hornung, Rousseau & Glaser, 2008), and, as opposed to job crafting, always intended to benefit both the employee and the manager (Lai, Rousseau & Chang, 2009).
In addition, currently knowledge on the link between proactive personality and work engagement is predominantly based on research at the between-person level, which infer stability within a person. However, most behaviours vary over time and are dependent on situational or personal circumstances (Ohly, Sonnentag, Niessen & Zapf, 2010), which particularly holds true for job crafting. In other words, how employees perform on a particular day is probably dependent on what occurs on that day and how employees experience that day (Tims, Bakker & Derks, 2014). Daily job characteristics that affect mood, which translates into well-being, work engagement (Teuchmann, Totterdell, & Parker, 1999) and proactive behaviours (Ohly & Fritz, 2010) are examples of behaviours that are dependent on what occurs on a particular day. These everyday variations in individuals’ behaviours are mostly ignored or treated as a measurement-error in between-person studies. Therefore, this study will go beyond the enduring work engagement and job crafting level and investigate these variables on a daily level, making use of repeated real-time measures. As Fisher and To (2012) argue in their article these repeated real-time measures are relevant and preferable for creating accurate person-level summary measurements. The constructs of proactive personality and I-deals will still be studied on the between-level as they infer stability within a person.
All in all, this study describes the relation between proactive personality trait and work engagement on a daily level, and the role of two bottom-up concepts of work design as mediators between this relationship. To clarify, job crafting as a daily mediator and I-deals as a general mediator of the relationship between proactive personality and daily work engagement will be investigated. The proposed research model is visualised in Figure 1. When the proposed hypotheses fail to reject, this suggests that employees with a proactive personality and proactive behaviours in the workplace are relevant mechanisms for human resource managers who are
9
seeking to enhance work engagement in their organisations.
This study will be structured as follows: At first, the concepts of proactive personality, job crafting, I-deals and work engagement will be explained and the research model with its related hypotheses will be introduced. Subsequently, the way this research is conducted will be discussed in the methodology section. Next, the outcomes of the relations of the research model will be analysed. Finally, a conclusion, implications and limitations of this research will be given. Besides, suggestions for future research areas will be described.
FIGURE 1: Hypothesised research model
THEORETICAL BACKGROUND
The exploration of the relation between proactive personality and work engagement on a daily level, and the role of two bottom-up concepts of work design as mediators between this relationship, is guided by an extensive theoretical background of the relevant literature. The formulated hypotheses were set up from these findings. At first, the direct link between proactive personality and (daily) work engagement is discussed. Followed by a section in which the two bottom-up concepts of work design, job crafting and I-deals, are compared and distinguished. Subsequently, the role of job crafting and I-deals as mediators in the relationship between proactive personality and work engagement will be discussed.
PROACTIVE PERSONALITY AND WORK ENGAGEMENT
Bateman & Crant (1993) were the first to introduce a measure of the ‘proactive personality’ trait and discussed the proactive element of organisational behaviour. They define proactive personality as “the relatively stable tendency to effect environmental change”. According to Buss and Finn (1987) people are not “passive recipients of environmental presses”, however, they have a specific influence on their own environments. Identifying proactive personality as a determinant of proactive behaviour supposes that proactive employees are proactive across
Proactive Personality I-Deals Daily Work Engagement Daily Job Crafting H2a H2b + + + + H2 H3 H1 H3a H 3b +
10
different circumstances and over time, regardless of the contingencies of a work situation or in one’s career (Parker & Bindl, 2017). People who have a proactive personality identify opportunities and act on them, they show initiative, take action, and work hard until they bring about significant change (Bakker et al., 2012). Furthermore, they transform the mission of their organisation, find and solve problems, and accept responsibility themselves and take personal initiative to have an impact on the world around them (Crant & Bateman, 2000). In contrast, individuals who do not have a proactive personality show the opposite patterns: they fail to identify, let alone seize, opportunities to change things (Crant, 1996). In other words, these people are more reactive and passive and tend to adapt to circumstances rather than change them (Seibert, Crant & Kraimer, 1999). The proactive personality is “a tendency to initiate and maintain actions that directly alter the surrounding environment” (Bateman & Grant, 1993). According to Buss and Finn (1987), proactivity is an instrumental trait because it belongs to a class of behaviours that impact the environment.
As organisations are increasingly growing complex and unpredictable, the topic of proactivity at work has become of great importance for contemporary workplaces. Proactivity drives performance and innovation of teams and organisations, and boosts individuals’ well-being and careers. When employees are proactive, they use their initiative at work to create a better future for themselves and their employers. They scan for opportunities, persist until change is achieved, and take charge to prevent problems from reoccurring in the future (Parker & Bindl, 2017). Findings of four meta-analyses on proactive personality (Fuller & Marler, 2009; Thomas, Whitman & Viswesvaran, 2010; Tornau & Frese, 2013; Spitzmuller, Sin, Howe & Fatimah, 2015) indicate that employees with a proactive personality tend to perceive more autonomy, psychological empowerment, job control, role clarity at work and feel more self-assured to complete specific work tasks (self-efficacy) and pursue goals that are job or career related (flexibility role orientation, role breadth self-efficacy, career self-efficacy, and job search self-efficacy). It has also been demonstrated that individuals with a proactive personality are probably more satisfied with their work and more committed to the organisation (i.e. low turnover intention and high commitment) (Parker & Bindl, 2017). Furthermore, studies have found relationships between proactive personality and individual job performance (Crant, 1995), career outcomes (Seibert et al., 1999), entrepreneurship (Becherer & Maurer, 1999; Crant, 1996), charismatic leadership (Bateman & Crant, 1993; Crant & Bateman, 2000; Deluga, 1998), organisational citizenship behaviours (e.g. altruism, courtesy, and sportsmanship) (Greguras & Diefendorff, 2010), and team performance (Kirkman & Rosen, 1999).
11
Moreover, previous research has meta-analysed the relationships between proactive personality and two opposing states of work; work engagement and burnout. Christian, Garza and Slaughter (2011) argue that proactive individuals are more involved in their work environment and thus more engaged in their work. Furthermore, scholars found that proactive personality was negatively related to the three sub-dimensions of burnout: emotional exhaustion, depersonalisation, and reduced personal accomplishment (Alarcon, Eschleman & Bowling, 2009). The concept of work engagement is defined as a positive state of fulfilment that is indicated by vigor, dedication, and absorption at work (Schaufeli & Bakker, 2010) and is the opposite of burnout (Demerouti, Bakker, Nachreiner & Schaufeli, 2001). In essence, work engagement captures how employees experience their work. The vigor component reflects a job experience which is stimulating and energetic, and to which an employee really wants to devote time and effort. Dedication occurs when the employee experiences the job as a significant and meaningful pursuit. Lastly, an employees’ work experience as engrossing and something on which they are fully concentrated is the absorption component of work engagement (Bakker & Demerouti 2008; Bakker et al., 2012). Engaged employees are highly energetic, identify strongly with their jobs, and experience states of flow at work.
Most research on work engagement has focused on differences between individuals and has treated day-to-day fluctuations in work engagement as a measurement-error (Sonnentag, Dormann, & Demerouti, 2010). However, literature argued that it is also possible that there are daily fluctuations in the experience of work engagement within an individual (Kahn, 1990). Sonnentag (2003) was the first to question this prevailing belief that engagement refers to a persistent and extensive affective-cognitive state (Schaufeli, Salanova, González-Romá & Bakker, 2002). She proposed that work engagement should not only be seen as an enduring experience. Rather, Sonnentag demonstrated and showed that levels of work engagement may vary within the same employee on a daily level, in response to specific personal and situational conditions (Sonnentag et al., 2010). Thus, daily work engagement was introduced as complementary to enduring work engagement (Bakker, 2014). Whereas enduring work engagement refers to how engaged employees feel in relation to their work in general, over long periods of time, daily work engagement reflects a temporary state of mind that exists on a certain moment and fluctuates within the same employee over short periods of time (e.g., day-to-day or momentary from hour to hour) (Bakker, 2014; Sonnentag et al., 2010).
Proactive individuals demonstrate initiative and dedication (Bateman & Crant, 1993; Crant, 1995). Hence, proactive personality is probably related to enduring work engagement
12
because employees who are more involved in their work surroundings are also more likely to immerse themselves in their work. This study will go beyond the enduring work engagement level and investigate work engagement on a daily level, making use of repeated real-time measures. As Fisher and To (2012) argue in their article these repeated real-time measures are relevant to create accurate person-level summary variables or measurements. Computing the desired indicators from multiple real-time measures is preferable because for some variables, the amount of intra-individual variability shows trait-like stability. This means that some employees are typically less stable in other variables than others, like work engagement. It is possible that individuals do not describe this variability well when they reflect on it. Therefore, this study contributes to the research field by creating accurate measurements of daily work engagement. And so, the direct link between proactive personality and daily work engagement is included in the proposed research model (see Figure 1), which lead to the first hypothesis:
H1: There is a positive, direct relationship between proactive personality and daily
work engagement
JOB CRAFTING VS. IDIOSYNCRATIC DEALS
As been mentioned before, Bakker and colleagues (2012), have shown that employees with a proactive personality are most inclined to change their work environment in a proactive way, by mobilising job demands and job resources, which in turn facilitate work engagement. Therefore, it may be equally important that employees organise their own job, responding to the complexity of contemporary jobs, and dealing with the needs of today’s work environment (Demerouti, 2014). Organisations increasingly introduce flexible work arrangements and customisation of jobs in order to be able to motivate employees. At the same time, new economic realities demand organisations to be more flexible, to adapt rapidly to the changing conditions in the market. Hence, employees are also expected to be more flexible, proactive, and able to adapt to changing work circumstances (Bal & Rousseau, 2016; Grant & Parker, 2009).
Two bottom-up concepts of work design can be found in recent literature, which both reflect the proactive behaviours of today’s employees: job crafting and Idiosyncratic deals. Both these behaviours increasingly occur due to the changing nature of work (e.g. flexible human resource practices, virtualisation) (Hornung et al., 2009). It is relevant for organisations to know on which terms these behaviours are initiated and how they differ from each other. The question
13
is to which extent job crafting and I-deals are any different, contributing to and enriching each other. More specifically, the simultaneous mechanisms of job crafting and I-deals, in explaining the link between a proactive personality and work engagement has not been investigated yet. This study aims to compare two parallel mechanisms of proactive behaviours at work as mediators, to explain the relationship between proactive personality and work engagement.
Bal and Rousseau (2016) give the following central assumption of job crafting and I-deals: “...the attention to the individual experience of the employee, who no longer follows a standardised career trajectory but for whom everything at work is and individualised experience that might or might not be shared with others. Hence, job adaptations currently follow an individualised approach…” (p. 4). Although job crafting and I-deals are both proactive work behaviours and concepts of work design, they work on a separate note. Table 1 shows the most important comparisons and dimensions of the two concepts. The primary goal of job crafting is to fulfil personal needs of employees through recognising and restructuring job demands and resources, where employees are the actors of the job. I-deals, in contrast, aim to achieve mutual benefit between employers and employees through negotiating employment features, which makes employees both actors and recipients of the job (Hornung et al., 2010; Parker & Bindl, 2017).
TABLE 1
Comparison of two concepts of work design and its dimensions (Hornung et al., 2010)
Dimensions Job Crafting I-Deals
Initiation Bottom-up by employee Bottom-up typically by employee Implementation Employee discretion Employee-employer negotiation Authorisation Unauthorised or within zone
of acceptance
Authorised by agents or human resource approval
Employee’s role Actor Both actor and recipient
Focus Individual job or position Individual job or position Primary goal Personal needs Broad mutual benefit
Design content Tasks and interactions Any or all employment features
Process Ongoing Intermittent events
The next sections will provide a complete theoretical review on both proactive work behaviour mechanisms; job crafting and I-deals. Furthermore, the paragraphs will describe and explain
14
the mediating roles of daily job crafting and I-deals in the relationship between proactive personality and daily work engagement.
THE MEDIATING ROLE OF DAILY JOB CRAFTING
Traditionally, research focused on the top-down processes of managers designing jobs for their employees and the job design theory (Campion & McClelland, 1993; Grant, Fried, & Juillerat, 2011). More recently, Wrzesniewski & Dutton (2001) have complemented traditional top-down beliefs of job design with the concept of “job crafting” as a bottom-up process, defined as “the physical and cognitive changes individuals make in the task or relational boundaries of their work” (p. 179). Employees can take action to adjust their work to better match their personal needs, strengths, and interests, or in other words to “turn the job they have into the job they want” (Berg et al., 2008; Wrzesniewski et al., 2010). In literature, two different forms of job crafting conceptualisations can be found.
To begin with, Wrzesniewski & Dutton (2001) distinguished three forms of job crafting. The first form involves changing the job's task boundaries, where employees change the number, scope, or type of job tasks at work. By deciding to do less, more, or different tasks than prescribed in the formal job description, employees craft a different job. Secondly, job crafters can change their relationships at work by altering the nature or extent of their interactions with others (“e.g. a computer technician offering help to co-workers as a way to have more social connection and teach new technicians”). Finally, job crafting could occur as employees change the cognitive task boundaries of their work, which reflects the third form of job crafting. In other words, employees change their jobs by altering how they perceive tasks, either by viewing it as a set of separated tasks or as an integrated whole. Wrzesniewski and Dutton’s (2001) view on job crafting is restricted to the changes that employees may make in their specific tasks, work relationships, and cognitions about work. Some recent studies have suggested that job crafting may take other forms as well (Lyons, 2008).
The present study follows the conceptualisation of job crafting proposed by Tims, Bakker and Derks (2012). Based on the Job Demands-Resources (JD-R) model (Bakker & Demerouti, 2014), they define job crafting as “the changes that employees may make to balance their job demands and job resources with their personal abilities and needs” (p. 174). Job demands are the physical, social or organisational job characteristics that require sustained physical and/or psychological effort. Job resources are the physical, psychological, social or organisational elements of the job that make sure employees achieve their goals and facilitate
15
their personal development (Demerouti et al., 2001). The JD–R model incorporates demands and resources of particular interest in organisations without focusing on predefined job features (Petrou, Demerouti, Peeters, Schaufeli, & Hetland, 2012). By examining job crafting in terms of job resources and job demands, job characteristics and other aspects that employees may develop in their jobs could be captured. Tims et al. (2012) have shown that job crafting can take the form of three different types of behaviours: (a) increasing (structural or social) job resources (e.g. autonomy, social support and feedback); (b) increasing job demands/challenges (e.g. new projects); and (c) decreasing job demands (e.g. fewer cognitive demands). Job demands refer to job characteristics that need continuous effort from employees and are, therefore, linked to various costs. On the other hand, job characteristics that contribute towards accomplishing specific goals at work, reducing the effect of job demands and its related costs, and supporting personal development are referred to as job resources. The authors propose a key distinction between social job resources and structural job resources. Structural job resources are factors such as autonomy, responsibility, capabilities and knowledge of the job, whereas social job resources are factors such as support from colleagues, feedback and supervisory coaching (Tims, Bakker & Derks, 2013). This distinction is particularly important. On the one hand, it allows to focus on significantly different job crafting behaviours. On the other hand, it assists in our understanding of how employees proactively relate to different elements of their working life, such as task structure and social structure (Berdicchia, Nicolli & Masino, 2016).
Although Wrzesniewski and Dutton (2001) define job crafting as “everyday” behaviour, most conceptualisations and operationalisations in the literature do not tap this aspect (Demerouti, 2014). There is some evidence that employees also employ in job crafting on a weekly and daily basis. To capture the “everyday” changes in job characteristics that employees may go after, various researchers (Petrou et al., 2012; Tims & Bakker, 2010) theoretically frame daily job crafting by using the Job Demands-Resources (JD–R) model (Bakker & Demerouti, 2007; Demerouti et al., 2001). This conceptualisation of job crafting helps to examine the construct of job crafting in a different light. By using the JD-R model job crafting can be seen as a construct that can unfold on a daily level as well and as being directed towards the surroundings of the work environment of the employee, namely the specific job resources and job demands. Wrzesniewski and Dutton (2001) and Petrou and colleagues (2012) suggest that even in if the environment is stable with clear work procedures and detailed job descriptions, employees adjust the tasks they are doing and mobilise the resources they need to achieve their tasks successfully, on a daily basis (Demerouti, Bakker & Halbesleben, 2015). In this study, the
16
extent to which job crafting can be conceived on a daily level in addition to the general level (as been proved in previous studies), will be investigated. This because merging momentary states and individual habits is necessary to understand the dynamics of organisational behaviour as they unfold daily (Ilies, Schwind, & Heller, 2007). Furthermore, as been previously mentioned, also Fisher and To (2012) argue that repeated real-time measures are useful for creating accurate person-level summary measurements.
Applying Crant’s (2000) theory of proactive personality, literature argues that employees with a proactive personality create favourable conditions and opportunities for themselves at their work. Tims and Bakker (2010) stated that proactive employees strive for coherence with their work environment in terms of needs and abilities. In other words, these employees shape their work environment such that the job demands and job resources better fit their own needs and abilities. Based on these arguments, Tims, Bakker and Derks (2012) found that employees with a proactive personality are most likely to ask for advice and feedback from colleagues or managers (social job resources), and proactively improve their work environment, for instance ask for autonomy, create skill variety, and follow training (structural job resources). Furthermore, they found that proactive employees are most likely to search for challenges, for instance, ask for more tasks or work when they feel under-stimulated. Therefore, the following hypothesis is formulated:
H2a: There is a positive relationship between proactive personality and daily job crafting (increasing job resources, increasing job challenges, decreasing demands). Job crafters proactively change their work environment and align their job demands and resources with their own abilities and needs (Tims & Bakker, 2010). Bakker and colleagues (2012) have shown that individuals with a proactive personality are most inclined to change their work environment in a proactive way, by mobilising job demands and job resources, which in turn facilitate work engagement. Whether employees develop symptoms of burnout or work engagement is, at least, partly the role of job design (Hornung et al., 2010). Literature revealed that job resources have a positive effect on work engagement and that job demands (e.g., workload, time pressure, role conflicts), however, have a negative effect on work engagement. Job resources (e.g. feedback, social support, and skill variety) are assumed to play either an extrinsic motivational role by being instrumental in achieving work goals, or an intrinsic motivational role of fostering employees’ growth, training, and development (Bakker et al., 2012). In addition, the research of Wrzesniewski and colleagues (2010) indicates that
17
“employees, at all levels, in all kinds of occupations, who try job crafting often end up more engaged and satisfied with their work lives, achieve higher levels of performance in their organisations, and report greater personal resilience.” (p. 115)
Quantitative diary studies suggest that daily variations in job resources explain daily variations in work, partly through their influence on daily personal resources like daily self-efficacy and daily optimism (Bakker, 2014). Finally, examining job crafting and its correlates daily is in line with the affective events theory (Weiss & Cropanzano, 1996), which states that job features influence employee affect through specific work events. For example, daily job characteristics that affect mood, which translates into well-being (Teuchmann et al., 1999) or proactive behaviours on a daily basis (Ohly & Fritz, 2010). Following this argumentation daily variations in job crafting could explain daily variations in daily work engagement. Therefore, the following hypothesis is formulated:
H2b: There is a positive relationship between daily job crafting and daily work engagement.
To conclude, the theoretical arguments so far suggest that proactive personality has a direct relationship with daily work engagement. However, this direct link will be more positively stronger through the mediation of daily job crafting. As individuals with a proactive personality are most inclined to change their work environment in a proactive way, by mobilising job demands and job resources (being job crafters), this in turn facilitate work engagement on a daily basis even more. All in all, this leads to the following hypotheses: H2: Daily job crafting mediates the positive relationship between proactive personality
and daily work engagement.
THE MEDIATING ROLE OF IDIOSYNCRATIC DEALS
I-deals and its effects on several employee outcomes can be explained using the theories of social exchange (Blau, 1964), and besides the norm of reciprocity (Gouldner, 1960). According to the social exchange theory, when an employee and an manager or supervisor trust each other in an relationship of exchange, mutual agreements between the two drive the behaviours of both parties. I-deals serve as a basis for reciprocity between the employee and the organisation or employer, because the mutual agreements that they have jointly coordinated enhances the relationship between the two (Bal, De Jong, Jansen, Bakker, 2012).
18
redesign and a single employee’s private efforts to craft a job (Hornung et al., 2010). Human resource practices often evolve through exceptions made to standard arrangements. I-deals can establish new criteria that eventually form the basis for broader changes in job design, as a flexible response to the changing needs (Hornung et al., 2010). The physical foundation of I-deals is established within the exchange relationship between an employee and the organisation. I-deals, in general, are employment terms that employees negotiate for themselves, taking multiple forms from flexible schedules to career development (Bal & Rousseau, 2016; Hornung et al., 2008), and intended to benefit both the employee and the manager over time (Lai et al., 2009). Idiosyncratic deals are employment arrangements that are different in nature from those given to other employees and are crafted to meet the specific needs of individual employees. These I-deals offer employees additional resources (e.g., special promotion tracks or flexible scheduling) not readily available to their colleagues (Rousseau, Ho & Greenberg, 2006). Moreover, the content of I-deals may vary quite extensively across employees. For instance, some individuals may only have idiosyncratic arrangements regarding their work schedules, while others may have idiosyncratic deals which address career promotion opportunities and compensation packages (Ng & Feldman, 2010).
Previous research identified two commonly negotiated forms of idiosyncratic deals: (1) flexibility I-deals and (2) developmental I-deals. So, in contrast to the belief that I-deals are just one single concept, it can be divided into two kinds of I-deals, namely: ‘hard’ I-deals (i.e. flexible working hours) and ‘soft’ I-deals (i.e. development; Hornung et al., 2008). Hard I-deals are solid, factual arrangements which have a general shared meaning, and can be measured objectively (e.g. working hours). To implement and record these hard I-deals, definitive metrics (i.e. number of hours worked) can be used. This means that flexibility I-deals are arrangements that personalise the scheduling of work and allow employees to customise their working hours to better fit his or her personal needs and preferences (e.g. personal discretion over scheduling) (Hornung et al., 2008; Rousseau, 2005). On the other hand, developmental I-deals are more soft than flexibility I-deals as they are more different per individual and because the way employees see development, their perception of it, is different in meaning for each individual (Rousseau, Hornung & Kim, 2009). Thus, soft I-deals obtain their value from the relationship between the employer and the employee (Rousseau et al., 2009). These I-deals are in nature more subjective and therefore need supportive surroundings to be productive (Rousseau, 2005). This means that developmental I-deals refer to customised opportunities to develop employees’ skills and competencies and meet personal goals for professional or career advancement (e.g. challenging
19
work assignments, individual recognition of performance, special training, career opportunities) (Hornung et al., 2008; Rousseau, 2005).
Previous studies have found relationships between I-deals and employee commitment, job satisfaction and organisational citizenship behaviours (OCBs) and the motivation to continue working after retirement (Anand, Vidyarthi, Liden & Rousseau, 2010; Bal, De Jong, Jansen & Bakker, 2012; Hornung et al., 2008; Liu et al., 2013; Rosen 2013; Van der Meij & Bal, 2013). More specifically, the organisation or manager negotiates a particular arrangement with the employee, and in return, the employee get more connected to the organisation (Hornung et al., 2008; Ng and Feldman, 2009), contributes to a higher degree (Hornung et al., 2008), and has a better relationship with the organisation (Rousseau et al., 2009). Furthermore, literature revealed that developmental I‐deals are positively related to social exchange and can boost the relationship between the employee and the organisation by increasing connection and competence (Hornung et al., 2008). Finally, flexibility I‐deals can lead to a better work-life balance, since family responsibilities are the main reasons for employees to pursue flexibility (Bailey & Kurkland, 2002). However, possible negative effects of developmental and flexibility I‐deals are found in literature. For example, some employees choose not to negotiate developmental I‐deals. If an organisation would merely depend on I‐deals for career development, some employees will be excluded and therefore not develop themselves, taking a chance that the positive effects of an I‐deal will backfire (Rousseau et al., 2009). In addition, flexibility I‐deals may also have negative consequences. To clarify, because employees with flexibility I-deals have different agendas, they may signal a lack of organisational commitment or engagement because their job attendance harder to notice (Perlow, 1997). Besides, when organisations fail to control the gained flexibility, their employees can start to perform less, which can become a major problem if this is followed by decreasing connection (Hornung et al., 2008).
In line with the hypothesis of the relationship between proactive personality and daily job crafting, also the relation between proactive personality trait and I-deals is hypothesised in this study. This to compare, in a parallel mediation, the two different forms of proactive behaviours of employees at work. Applying Crant’s (2000) theory of proactive personality, this study argues that employees with a proactive personality create favourable conditions and opportunities for themselves in their work. People with proactive personalities identify opportunities and act on them, they show initiative and take action (Bakker et al., 2012). Hence, they will negotiate their employment terms with the managers for themselves, to create the
20
favourable conditions, by using (flexibility and developmental) I-deals. In other words, these employees shape their work environment such that the job demands and job resources better fit with their own needs and abilities. Therefore, this leads to the following hypothesis:
H3a: There is a positive relationship between proactive personality and I-deals.
As been mentioned before, research have found several positive outcomes of I-deals for employees, one positive outcome that is not discussed yet in this theoretical background is work engagement. Employees who negotiate I-deals with their managers, proactively change their work environment and align their job demands and resources with their own abilities and needs (Tims & Bakker, 2010). Bakker and colleagues (2012) have shown that individuals with a proactive personality are most inclined to change their work environment in a proactive way, by mobilising job demands and job resources, which in turn facilitate work engagement. Bal and Vink (2011) showed that flexibility I-deals are related to an increase in motivation to continue working, while the links of developmental I-deals only manifested under conditions of a favourable work environment. On the other hand, a quantitative research revealed that developmental I-deals are linked to higher work engagement, while flexibility I-deals are linked to lower work–family conflict (Hornung, Rouseau, Glaser, Angerer & Weigl, 2011). This means we could argue for both types of I-deals to relate with work engagement. Moreover, Serrano and Reichard (2011) examine the relationship between I-deals and work engagement from the manager or organisational point-of-view, instead of the employee. They argue that managers can collaborate with employees to negotiate work responsibilities and demands, by creating I-deals. This makes sure that the work of the employees is in line with their personal needs, which, they say, makes work engagement more likely. Besides, Hornung and colleagues (2010) found three mechanisms through which the creation of I-deals positively relates to work engagement. At first, creating I-deals increased job complexity or “the degree tasks allow the use of intellectual abilities, require collaboration, and support skill acquisition” (Hornung et al., 2010, p. 195). Subsequently, I-deals increase job control and employees’ ability to work alone and use judgement during a decision-making process. Finally, I-deals positively relate to work engagement by reducing job demands or obstacles to get the things done. Overall, creating I-deals supports employees to become vigorously and more involved in their jobs. This because employees get more engaged in completing complicated tasks, exercise decision-making judgement, and control for undesirable job demands (Hornung et al., 2010).
21
between individuals and has treated day-to-day fluctuations in work engagement as a measurement-error (Sonnentag et al., 2010). However, literature argued that it is also possible that there are daily fluctuations in the experience of work engagement within an individual (Kahn, 1990). The relation between I-deals (on a general level) and work engagement (on a day level) has not been investigated yet. And, as Ohly and colleagues (2010) argue, it is relevant to combine general questionnaire information with a daily diary study of what employees do during their workday. Therefore, the following hypothesis is formulated:
H3b: There is a positive relationship between I-deals and daily work engagement. To conclude, the theoretical arguments so far suggest that proactive personality has a direct relationship with daily work engagement. However, this direct link will be more positively stronger through the mediation of I-deals. As individuals with a proactive personality are most inclined to change their work environment in a proactive way, using I-deals to negotiate employment terms and arrangements with their managers, this in turn facilitate (daily) work engagement even more. All in all, this leads to the following hypotheses:
H3: I-deals mediate the positive relationship between proactive personality
22
METHOD
This chapter will explain how the hypotheses of this research will be tested. First of all, the method of research will be discussed. Secondly, a description of the sample will follow given a thorough understanding of the descriptive statistics of the final sample. Finally, the measurements of the independent variable, dependent variable, mediators, and control variables will be discussed.
RESEARCH DESIGN
This quantitative study collects data by using a general questionnaire and a diary study of five days. The data collection was done by three students from the Amsterdam Business School, but each student conducted its own research. Employees, working more than three days a week, from various organisations in the Netherlands (e.g. services, health care, education, transport & logistics) were asked to take part in this research. Participation was voluntary, however, an incentive of six vouchers, each worth €50 on Bol.com, was used to motivate participation. Participants were informed that upon agreement, they would receive an online diary questionnaire and that they were invited to fill out the survey for five consecutive working days at the end of each day. Before starting with the diaries, respondents first have to fill out a general online questionnaire in which they provide demographic data and information on the general level of the measured variables. Both surveys were administered online, using Qualtrics software. The complete version of both surveys can be found in Appendix I. With a self-completion questionnaire, respondents answer questions by completing the questionnaire themselves. The online version operates by inviting prospective respondents by e-mail to visit a website at which the questionnaire can be found and completed online. The cover letter used to explain the research to the respondents could be found in Appendix II. With these type of questionnaires, there is an absence of interviewer effects, there is no interviewer variability and there is more convenient for respondents because they can complete a questionnaire when they want and at the speed that they want to go (Bryman & Bell, 2011).
SAMPLE
The population of interest for this study is ‘Dutch workers’; working more than three days a week. As the population is large and the sampling frame is unknown, this study was conducted using a non-probability convenience sample. This is a sample that has not been selected using a random selection method and occurs when either the probability that every unit or respondent
23
included in the sample cannot be determined (Bryman & Bell, 2011). Respondents were reached through personal e-mail and social media channels among the personal network of the researchers. The between-level and within-level surveys were send to 132 participants. In total, 117 respondents filled out the questionnaire on the between-level, reaching a response rate of 88.64%. On the within-level, in total, 111 Dutch workers responded to minimal three days of the dairy study, reaching a response rate of 84.09%. All in all, 127 measurement points were collected on the within-level survey.
Table 2 shows the mean values and standard deviations adopted to illustrate the sample characteristics; employees in the Netherlands working 3 days a week or more. With an average age of 34 years (SD = 13.04), 60.4% of the respondents were female employees; 82% had an educational level of HBO or higher; and on average respondents had 7 years of organisational tenure (SD = 9.23). Furthermore, the respondents work in a wide range of sectors (e.g. services, health care, education, transport & logistics) with an average of 37.45 working hours per week (SD = 7.49). When comparing the final demographic results of the sample frame with the data of the Dutch working population (n = 8.474.000), a small amount of bias is shown. This especially accounts for the demographic variables gender and age (CBS, 2017). In the sample, woman are overrepresented and the average age was lower.
TABLE 2
Demographic characteristic profile of respondents (N = 108)
Variables N % Mean SD Gender .60 .49 Male Female Age Educational level Secondary School MBO HBO University (bachelor) University (master) PhD Other Organisational Tenure Working hours per week
42 64 3 14 42 5 38 2 2 39,6 60,4 2.8 13.2 39.6 4.7 35.8 1.9 1.9 33.85 3.71 7.16 37.45 13.04 1.31 9.23 7.49
Note. Men = 0, woman = 1. Age is coded in years. Educational level is based on Dutch system, secondary school = 1, MBO =
24
MEASURES
As respondents’ native language was Dutch, each measuring scale was translated. To validate the translation, the back translation method, a bilingual technique, and pretesting, as suggested by Brislin, Lonner and Thorndike (1973) was utilised.
General questionnaire (between-level)
Proactive personality. Proactive personality was assessed using the Dutch translation (Claes et al., 2005) of the six-item version of Bateman and Crant’s (1993) Proactive Personality Scale (PPS). Sample items include: ‘This employee is always looking for better ways to do things’, and ‘If this employee sees something s/he doesn’t like, s/he fixes it’. Respondents could respond on a five-point scale ranging from 1 = disagree through 5 = agree. The reliability of the scale was good; Cronbach’s α =.77.
I-Deals. Using a Dutch translation of Rousseau’s measures (e.g., Hornung et al. 2008), employees indicated the extent to which in their current jobs they had “asked for and successfully negotiated individual arrangements different from their peers” in terms of flexibility I-deals (three items) and development I-deals (four items). Flexibility items include: ‘Flexibility in starting and ending the workday’ and ‘Individually customised work schedule’. Developmental items include: ‘On-the-job activities’, ‘Training opportunities’ and ‘Career development’. Respondents could respond on a five-point scale ranging from 1 = not at all through 5 = to a very great extent. The reliability of the scale was very good; Cronbach’s α =.90.
Dairy study (within-level)
The diary study consisted of five identical questionnaires, one for each day. Similar to that of previous diary studies (Fritz & Sonnentag, 2009), respondents indicated how representative each statement was for the past day using a scale, ranging from 1 = strongly disagree to 5 = strongly agree. All respondents were instructed to answer the questionnaires right before leaving the office that day. Besides, by means of an online questionnaire the author was more capable of checking the actual day and time (i.e. more accuracy in the measurement).
Daily Job Crafting. Daily job crafting was assessed using the Dutch translation of Petrou and colleagues’ (2012) subscales. The day-level job crafting questionnaire consisted of three
25
subscales: Day-level seeking recourses included four items as “Today, I have asked colleagues for advice”. Day-level seeking challenges included three items as “Today, I have asked for more responsibilities”. Day-level reducing demands was measured by using three items as “Today, I have tried to ensure that my work is emotionally less intense”. Cronbach’s α for the total scale =.83.
Daily Work Engagement. This study adapted three items from Schaufeli, Bakker and Salanova’s (2006) short-form Utrecht Work Engagement Scale (UWES) to assess daily work engagement (Lanaj, Johnson & Barnes, 2014). Sample items translated to Dutch include: “Today, my job inspired me” and “Today, I was very enthusiastic about my job.” Respondents could respond on a five-point scale ranging from 1 = strongly disagree through 5 = strongly agree. Cronbach’s α =.76.
Control variables. The variables tenure, age and gender are included as a control variable in this study. At first, literature revealed that tenure may have an effect on job proactive behaviours as job crafting and I-deals. Several authors argue that tenure in the organisation could be one factor that is likely to influence how employees make sense of the possibilities for job crafting, and should be investigated in future research (Berg et al., 2010; Ghitulescu, 2006). Berg, Wrzesniewski and Dutton (2010) argue that it is possible that employees that are shorter tenured engage in more job crafting or job proactive behaviours. This because short tenured employees proactively shape their jobs to themselves and quickly examine where changes could be made, while longer-tenured employees are more used to the job and examine it more as a fixed entity. Two possible competing hypotheses could be that longer-tenured employees receive more authority and support from their managers and colleagues, in making changes to their job (Wrzesniewski, Bartel, & Wiesenfeld, 2009). Furthermore, it could be that longer-tenured employees could better take the opportunities for job proactive behaviours available to them due to their time spent in the organisation (Berdicchia et al., 2016). Thus, the employees’ tenure with the organisation (the length of the time the employee had worked for the organisation) is controlled. An open question was used to measure the tenure (in years) of employees working for the organisation: "How long do you work for this organisation (in years)?".
Next to that, Kooij, Tims and Kanfer (2015) discussed whether older employees may craft their jobs differently than younger employees, since motivation and competencies change with age and individuals use strategies to age successfully in life. Furthermore, they say that
26
older employees know their own competencies, abilities, roles, tasks and responsibilities due to their long work experience and job tenure. As such, the bottom-up approach of job crafting is the most appropriate approach for older employees to adapt the job to age-related changes they are confronted with, and to age successfully at work. Therefore, age was included as a control variable in this study.
Finally, the role of gender could have an influence on the proactive behaviours at work. There is a mixed picture found in literature. In terms of proactive job search, networking behaviours and voicing concerns about issues at work, men were found to be more proactive than woman. However, the influence still remains inconclusive and it is suggested that further research will be helpful to clarify the relevance of gender in job proactive behaviours (Berdicchia et al., 2016; Bindl & Parker, 2010). Therefore, gender was included as a control variable and measured as a dichotomous variable, i.e., 0 for male and 1 for female.
27
RESULTS
In this section the results of the different analyses will be presented. At first, they analystical strategy on the collected data will be discussed. Subsequently, the descriptive statistics for the most important variables are given. Moreover, this section provides a correlation matrix to examine the relation between the different variables. Besides, the reliability analyses of all scales are given. Finally, this section ends with testing the hypothesis formulated in this research, followed by an additional analysis of the variables on a general level instead of the daily level.
ANALYTICAL STRATEGY
At first, a check of the frequencies was computed to examine if there were any errors in the data; no errors were found. Besides, by excluding cases listwise missing values were dealt with. This means that only cases that had no missing data in any variable were analysed. Descriptive statistics (means, standard deviations and correlations between the study variables), skewness, kurtosis and normality tests were computed for all variables. All variables were normally distributed with skewness and kurtosis ranging between -1 and 1; Proactive Personality (skewness =.088, kurtosis =.195), Daily Job Crafting (skewness = -.087, kurtosis =.670), Idiosyncratic Deals (skewness = -.019, kurtosis = -.643) and Daily Work Engagement (skewness =.058, kurtosis -.272). Furthermore, to verify the underlying dimensions of the independent variable, the mediators and the dependent variable, an Exploratory Factor Analysis was conducted on the measurement scales of proactive personality, daily job crafting, Idiosyncratic deals, and daily work engagement. Finally, the PROCESS Macro written by Andrew F. Hayes for SPSS was used to test the hypotheses.
EXPLORATORY FACTOR ANALYSIS
Proactive Personality (Independent variable). A principal axis factoring analysis (PAF) was conducted on the six items with oblique rotation (direct oblimin) in order to investigate the reliability of the item list used to measure proactive personality.
At first, the correlation matrix was analysed to investigate multicollinearity and whether correlations were too small. As the determinant of the correlation matrix (.239) was not smaller than the necessary value of 0.00001, no multicollinearity existed. Moreover, Bartlett’s test of sphericity, which tests the null-hypothesis that the original correlation matrix is an identity matrix, was significant for this particular set of data (χ2 (15) = 149.250, p = <.001). This means
28
that the correlation matrix is not an identity matrix and that correlations between items were sufficiently large for PAF. Furthermore, the Kaiser-Meyer Olkin measure verified the sampling adequacy for the analysis, KMO =.77, as this is above the acceptable limit of.5 (Field, 2013). This means that all assumptions for the PAF are met.
An initial analysis was run to obtain eigenvalues for each factor in the data. One factor had eigenvalues over Kaiser’s criterion of 1 and explained 46,56% of the variance. The scree plot showed an inflexion that would justify retaining one factor. Therefore, one factor was retained because of the use of fewer than 30 variables, the convergence of the scree plot and Kaiser’s criterion on this value. All six items load strongly on the one underlying factor (all above.505), representing “Proactive Personality”.
In order to compare the independent variable to the other variables, one variable is computed that reflects the mean score of the six items that constitute proactive personality (ProTOT).
Daily Job Crafting (Mediator 1). A PAF was conducted on the ten items with oblique rotation (direct oblimin) in order to investigate the reliability of the item list used to measure job crafting on a day-level.
At first, the correlation matrix was analysed to investigate multicollinearity and whether correlations were too small. As the determinant of the correlation matrix (.001) was not smaller than the necessary value of 0.00001, no multicollinearity existed. However, Bartlett’s test of sphericity, which tests the null-hypothesis that the original correlation matrix is an identity matrix, was significant for this particular set of data (χ2 (45) = 781.662, p = <.001). This means that the correlation matrix is not an identity matrix and that correlations between items were sufficiently large for PAF. Furthermore, the Kaiser-Meyer Olkin measure verified the sampling adequacy for the analysis, KMO =.77, as this is above the acceptable limit of.5 (Field, 2013). This means that all assumptions for the PAF are met.
An initial analysis was run to obtain eigenvalues for each factor in the data. Three factors had eigenvalues over Kaiser’s criterion of 1 and explained 78.65% of the variance. The scree plot shows an inflexion that would justify retaining three factors. Therefore, three factors are retained because of the use of fewer than 30 variables, the convergence of the scree plot and Kaiser’s criterion on this value. Table 3 shows the factor loadings after rotation. The items that cluster on the same factor suggest that factor 1 represents “Day-level seeking resources”, factor 2 represents “Day-level seeking challenges” and factor 3 represents “Day-level reducing
29
demands”.
In order to compare the mediator to the other variables, one variable is computed that reflects the mean score of the ten items that constitute daily job crafting (DjcraTOT).
TABLE 3
Summary of exploratory factor analysis results for measurement scale Daily Job Crafting (N = 111)
Factor Loadings Item Seeking resources Seeking challenges Reducing demands I have asked others for feedback on my
job performance
.20 .02 .70
I have asked colleagues for advice -.19 -.08 .85
I have asked my supervisor for advice .11 .07 .60
I have tried to learn new things at work .29 .10 .53
I have asked for more tasks if I finish my work
.86 .01 -.02
I have asked for more responsibilities .96 -.04 .04
I have asked for more odd jobs .92 -.01 .05
I have tried to ensure that my work is emotionally less intense
-.05 .95 .12
I have made sure that my work is mentally less intense
.01 .94 -.12
I have tried to ensure that my work is physically less intense
-.02 .72 .01
Eigenvalues 4.07 2.36 1.4
% of variance 40.74 23.57 14.34
Note. Factor loadings over.40 appear in bold
Idiosyncratic Deals (Mediator 2). A PAF was conducted on the seven items with oblique rotation (direct oblimin) in order to investigate the reliability of the item list used to measure Idiosyncratic deals.
At first, the correlation matrix was analysed to investigate multicollinearity and whether correlations were too small. As the determinant of the correlation matrix (.004) was not smaller
30
than the necessary value of 0.00001, no multicollinearity existed. Moreover, Bartlett’s test of sphericity, which tests the null-hypothesis that the original correlation matrix is an identity matrix, was significant for this particular set of data (χ2 (21) = 545.973, p = <.001). This means that the correlation matrix is not an identity matrix and that correlations between items were sufficiently large for PAF. Furthermore, the Kaiser-Meyer Olkin measure verified the sampling adequacy for the analysis, KMO =.81, as this is above the acceptable limit of.5 (Field, 2013). This means that all assumptions for the PAF are met.
An initial analysis was run to obtain eigenvalues for each factor in the data. Two factors had eigenvalues over Kaiser’s criterion of 1 and explained 79.53% of the variance. The scree plot shows an inflexion that would justify retaining two factors. Therefore, two factors are retained because of the use of fewer than 30 variables, the convergence of the scree plot and Kaiser’s criterion on this value. Table 4 shows the factor loadings after rotation. The items that cluster on the same factor suggest that factor 1 represents “Flexibility I-deals” and factor 2 represents “Developmental I-deals”. Although this analysis showed that “on-the-job-activities” represents flexibility I-deals, literature suggests that this item represents developmental I-deals (e.g., Hornung et al. 2008). It is possible that on-the-job activities load strongly on flexibility I-deals because flexibility I-I-deals are arrangements that personalise the scheduling of work and allow employees to customise their working hours to better fit the individual needs and preferences, which also fits with on-the-job activities. Still, however, the ‘wrong’ loading is not a problem because in the end this study averages across all items.
In order to compare the mediator to the other variables, one variable is computed that reflects the mean score of the seven items that constitute I-deals (Ideal).
Daily Work Engagement (Dependent variable). A PAF was conducted on the three items with oblique rotation (direct oblimin) in order to investigate the reliability of the item list used to measure daily work engagement.
At first, the correlation matrix was analysed to investigate multicollinearity and whether correlations were too small. As the determinant of the correlation matrix (.466) was not smaller than the necessary value of 0.00001, no multicollinearity existed. Moreover, Bartlett’s test of sphericity, which tests the null-hypothesis that the original correlation matrix is an identity matrix, was significant for this particular set of data (χ2 (3) = 82.706, p = <.001). This means that the correlation matrix is not an identity matrix and that correlations between items were sufficiently large for PAF. Furthermore, the Kaiser-Meyer Olkin measure verified the sampling
31
adequacy for the analysis, KMO =.68, as this is above the acceptable limit of.5 (Field, 2013). This means that all assumptions for the PAF are met.
TABLE 4
Summary of exploratory factor analysis results for measurement scale I-deals (N = 105) Factor Loadings
Item Flexibility
I-deals
Developmental I-deals Flexibility in starting and ending the working day .92 .10
Individually customised work schedule .74 .04
Flexibility in work-related tasks .76 -.20
On-the-job activities .58 -.27
Training opportunities .12 -.74
Special opportunities for skill development -.08 -1.04
Career development .04 -.85
Eigenvalues 4.45 1.11
% of variance 63.61 15.92
Note. Factor loadings over.40 appear in bold
An initial analysis was run to obtain eigenvalues for each factor in the data. One factor had eigenvalues over Kaiser’s criterion of 1 and explained 67.56% of the variance. The scree plot showed an inflexion that would justify retaining one factor. Therefore, one factor was retained because of the use of fewer than 30 variables, the convergence of the scree plot and Kaiser’s criterion on this value. All three items load strongly on the one underlying factor (all above.616), representing “Daily Work Engagement”.
In order to compare the dependent variable to the other variables, one variable is computed that reflects the mean score of the three items that constitute daily work engagement (DengTOT).
DESCRIPTIVE STATISTICS
Table 5 shows the means, the standard deviations, and the correlations between the study variables. For all day-level variables, an aggregate score of the five days is used; therefore, for all day-level variables, participants were assigned a mean score of their five measurements.
32
Furthermore, only those who have taken the within survey more than three times that week were taken into account. The between-level and within-level surveys were send to 132 participants. In total, 117 respondents filled out the questionnaire on the between-level, reaching a response rate of 88.64%. On the within-level, in total, 111 Dutch workers responded to minimal three days of the dairy study, reaching a response rate of 84.09%. All in all, 127 measurement points were collected on the within-level survey. The reliabilities of multi-item scales were determined by computing Cronbach’s alpha values. All scales have acceptable alpha values and are greater than the suggested cut-off level of 0.7 (Hair, Black, Babin, Anderson & Tatham, 1998), ranging from.76 to.90.
Inconsistent with the expectations, no significant correlations were found between proactive personality, daily job crafting, and daily work engagement. A significant correlation was found between I-deals and daily work engagement, r =.21, p <.05. Besides, a significant correlation was found between I-deals and daily job crafting (r =.24, p <.05). Examining the control variables, age was significantly correlated with gender, org. tenure, daily job crafting (r = -.27, p >.05) and daily work engagement (r =.26, p <.01). Furthermore, org. tenure was significantly correlated with gender, age and daily job crafting (r = -.21, p <.05).
TABLE 5
Means, Standard Deviations, and Correlations among Variables (N = 105)
Variables Mean SD 1 2 3 4 5 6 7
1. Gender .61 .49 -
2. Organisational Tenure 7.10 9.26 -.21* -
3. Age 33.85 13.04 -.30** .74** -
4. Proactive Personality 3.72 .52 -.15 -.05 -.08 (.77) 5. Daily Job Crafting 2.59 .52 .00 -.21* -.27* .16 (.83)
6. I-deals 2.75 .98 -.09 -.01 .08 .05 .24* (.90)
7. Daily Work Engagement 3.38 .67 -.09 .13 .26** .19 .02 .20* (.76)
Note: *p <.05, **p <.01
Cronbach’s alphas are in parentheses.
Men are coded as 0, woman are coded as 1. Organisational Tenure is measured in years.
HYPOTHESIS TESTING
