When do job crafting interventions work? The moderating roles of workload, intervention intensity, and participation

Highlights

• High workload prior to a job crafting intervention promotes crafting to reduce hindrance demands

• Low workload prior to a job crafting intervention promotes crafting to increase job resources

• Both job crafting interventions were effective regardless of intensity

• Researchers and practitioners can promote intervention effectiveness by motivating individuals to actively participate

Abstract

Job crafting refers to self-initiated, proactive strategies to change work characteristics to better align one's job with personal needs, goals, and skills. This study evaluated the conditions under which job crafting interventions are effective for increasing job crafting behaviours. We assessed the impact of initial workload on the effectiveness of two interventions – a less intense, knowledge-reflection intervention (N = 39), and a more intense, knowledge-reflection-action intervention that involved completing Job Crafting Boosts over 4 weeks (N = 50). Irrespective of intervention, longitudinal growth modelling analyses revealed that those with high initial workload engaged in more crafting behaviours to decrease hindering demands, whereas those with low initial workload engaged in more crafting behaviours to increase structural resources. No significant differences were observed between intervention groups in any job crafting behaviours. A further goal of the study was to examine, for those in the knowledge-reflection-action intervention, how much the participants actively participated in Job Crafting Boosts. Engagement varied considerably, with the average participation being below the target of three Job Crafting Boosts per week. Content analysis of open-ended responses to questions revealed that, for those who did engage in the Job Crafting Boosts, there were positive experiences, including insights about making positive changes at work, increased vigour and motivation, increased productivity, and resolved problems. This research shows that: interventions to reduce hindering demands should be targeted at those high in initial workload; interventions to increase job resources should be targeted at those low in initial workload; and intervention intensity does not impact intervention effectiveness.

Introduction

Job crafting is a powerful way by which individuals can change their work design. Job crafting refers to the self-initiated strategies that individuals use to change aspects of their work to align them more with their personal needs, goals, and skills (Tims, Bakker, & Derks, 2013; Wrzesniewski & Dutton, 2001). Examples of job crafting include seeking support from colleagues, asking for feedback from a supervisor, and/or enquiring about training and development opportunities. Many studies have shown that, by proactively crafting aspects of their work in these ways, individuals experience improved well-being, work engagement, and performance (see meta-analyses of Lichtenthaler & Fischbach, 2018 and Rudolph, Katz, Lavigne, & Zacher, 2017 and reviews of Lazazzara, Tims, & Gennaro, 2019, and Zhang & Parker, 2019). This type of changing work design is therefore a bottom-up approach by which individuals tailor their jobs to meet their own needs, goals, and skills. Job crafting contrasts with top-down work redesign approaches that involve implementing changes that affect everyone in a whole department or organization, whether or not the employee embraces the change (Hornung, Rousseau, Glaser, Angerer, & Weigl, 2010).

Given the growing amount of research that demonstrates the associations between job crafting behaviours, antecedents, and outcomes (Zhang & Parker, 2019), researchers have developed interventions in which employees learn to craft their jobs in order to improve their work motivation and performance (e.g., Gordon, Demerouti, Le Blanc, & Bipp, 2015; Van Wingerden, Bakker, & Derks, 2016). A recent meta-analysis identified 14 job crafting intervention studies and found positive and significant effects of job crafting interventions on overall job crafting, as well as on specific job crafting behaviours to increase job challenges and to decrease hindering job demands (Oprea, Barzin, Vîrgă, Iliescu, & Rusu, 2019). Additionally, positive, significant effects were found on work engagement and contextual performance, supporting the value of job crafting interventions for positively changing work behaviours and impacting employee outcomes, such as well-being and job performance.

However, results from individual intervention studies reveal inconsistent findings across different types of crafting. Oprea et al.’ (2019) meta-analysis reported confidence intervals for the effect of interventions on job crafting and sub-dimensions which were sufficiently wide to suggest moderators (overall job crafting, g = 0.26, 95%-CI = 0.09–0.44; seeking job resources, g = 0.19, 95%-CI = 0.01–0.38; reducing hindrance demands, g = 0.57, 95%-CI = 0.15–0.99). As an example of the inconsistent effects in individual studies, Dubbelt, Demerouti, and Rispens (2019) and Van Wingerden et al. (2017a) reported that their job crafting intervention increased seeking resources and decreasing demands behaviours, but did not increase seeking challenges. In another study, Van Wingerden et al. (2017b) reported that the job crafting intervention increased seeking challenges but did not increase seeking resources, nor did it affect decreasing hindering demands.

Given these inconsistencies across job crafting intervention studies, it is important to consider why interventions are, or are not, successful, and to discuss how well interventions are designed or implemented (Briner & Reynolds, 1999; Nielsen & Miraglia, 2017). That is, research has yet to answer, under what conditions job crafting interventions are most successful in inducing more crafting, that is, what factors – such as workload and intervention intensity– shape intervention effectiveness? Without this information, the field is at risk of settling on erroneous conclusions, which hinders the effective design of future interventions. It is therefore important to shed light on the implementation of the intervention, alongside traditional analyses of intervention effectiveness. At the same time, it is important to understand which aspects of job crafting are affected by crafting interventions.

In this study, for a sample of managers and professionals, we report on the implementation of two job crafting interventions, focusing on the question as to under which conditions job crafting interventions succeed in stimulating particular job crafting behaviours. Specifically, we pursue four goals. First, we examine an important boundary condition that may shape the type of crafting behaviours that individuals engage in, namely workload. Individuals need initial resources to engage in proactive behaviours because these behaviours can consume considerable resources, such as time and mental energy (Parker, Johnson, Collins, & Nguyen, 2013). We draw on conservation of resources (COR; Hobfoll, 1989) theory to predict that workload will shape the type of job crafting that is stimulated by job crafting interventions. If workload is high, individuals will be motivated to invest in types of job crafting that protect against future resource loss, such as decreasing their hindering job demands, whereas if workload is low, we expect that individuals will be motivated to engage in types of job crafting that will build their resources, such as increasing structural job resources. To our knowledge, workload is thus far unexplored as a situational factor that may facilitate or hinder engagement in particular job crafting behaviours, even though there is strong theoretical and empirical evidence that workload has important ramifications for how one feels and behaves at work (Jex, 1998; Prem, Paškvan, Kubicek, & Korunka, 2018).

The second goal of this study is to evaluate whether intervention intensity influences the extent to which individuals engage in job crafting behaviours post-intervention, to evaluate whether intervention intensity might be able to explain differences in intervention effectiveness. We compare two job crafting interventions that vary in intensity. The first, less intense, intervention, which we refer to as a ‘knowledge-reflection’ intervention, provided education around job crafting, and encouraged participants to analyse and reflect on their own work design. The second, more intense intervention, which we refer to as a ‘knowledge-reflection-action’ intervention, incorporated all elements of the ‘knowledge-reflection’ intervention, and additionally included weekly encouragement over four weeks to engage in specific job crafting activities, which we call Job Crafting Boosts. Specifically, participants were encouraged to engage in at least three Job Crafting Boosts per week by selecting from a curated list, and were reminded weekly to engage in these Boosts by the researchers. Each Boost focused on a specific type of crafting, for example, by building job or social resources such as feedback, training opportunities, and colleague support, or reducing ‘hindering’ demands, such as eradicating excessive emails or mundane tasks. Examples include learning a new skill, brainstorming with colleagues, identifying interesting projects and negotiating involvement, and reducing distractions. Drawing on intervention research in different areas, intervention intensity has been associated with better outcomes following the intervention (e.g., Duhon, Mesmer, Atkins, Greguson, & Olinger, 2009; Schwichtenberg & Poehlmann, 2007). However, this idea has not received empirical scrutiny in designing job crafting interventions.

The third goal of this study is to delve more deeply into how much participants actually engage in an intervention and for how long. To our knowledge, studies have not yet measured how many job crafting activities individuals have partaken in, or what types. This means that we do not know to what extent individuals carry out crafting behaviours at work after engaging in a job crafting intervention. Addressing this goal was possible in our knowledge-reflection-action intervention because we assessed the number and type of Job Crafting Boosts that each participant completed at work. Building on research about the transfer of training, being able to perform the learned behaviour in the work context is crucial for an intervention to be considered successful (Van der Locht, Van Dam, & Chiaburu, 2013; Volet, 2013).

Our fourth goal is to evaluate participants' subjective experiences of the knowledge-reflection-action intervention. Through the short online evaluation survey that participants completed following each Job Crafting Boost, and which acted as an opportunity for further reflection following each Boost, we gained quantitative and qualitative information enabling us to contextualise the results from the surveys and gain further insight into how workload impacts individuals' perceptions of effectiveness of this more intensive intervention. Nielsen and Randall (2013) highlighted the importance of evaluating participants' subjective experiences in addition to more traditional analyses of evaluation effectiveness (e.g., between-subjects, repeated measures analyses of variance) to prevent erroneous conclusions. For example, it may be possible that statistically significant results for intervention effectiveness are not found, but that participants themselves perceive that the intervention has been effective. Results from this type of holistic analysis, in which both statistical effectiveness of interventions and also participation is taken into account, may also indicate how practical such an intervention is for participants, with low participation rates suggesting a less intense intervention is more feasible, informing the design of future interventions.

Taken together, by investigating the role of workload, the intensity of the intervention, and participation in / subjective experiences of the knowledge-reflection-action intervention, our study helps to tease apart theoretical and practical reasons for job crafting intervention success, as well as helping to explain why some interventions affect particular types of crafting but not others. We support our findings with both quantitative and qualitative data, increasing the robustness of our conclusions. In what follows, we elaborate the theory underpinning these goals and develop specific hypotheses.

Work design refers to the nature and organization of tasks, roles, relationships and responsibilities that individuals hold at work (Parker, 2014). Much research has shown that work which is high in positive job characteristics, or job resources, such as autonomy, feedback and social support, with moderate job demands, such as workload and time pressure, is beneficial for work motivation, well-being and performance (Bakker & Demerouti, 2007; Humphrey, Nahrgang, & Morgeson, 2007; Parker, Morgeson, & Johns, 2017). Job crafting is one way in which work can be redesigned to increase work quality. Wrzesniewski and Dutton (2001) first coined the term job crafting, defining it as ‘the physical and cognitive changes individuals make in the task or relational boundaries of their work’ (p.179). From this theoretical perspective, individuals self-initiate changes to the type and scope of tasks they conduct at work, the quality or frequency of interaction with others at work, and/or the way they cognitively frame or view their job.

A follow-up school of thought approaches job crafting from job demands-resources theory (JD-R; Bakker & Demerouti, 2007). This perspective proposes that individuals craft their jobs to balance their job demands and resources (see Tims & Bakker, 2010). Job demands refer to aspects of the job which require sustained cognitive, emotional or physical effort, such as workload, dealing with time pressure, or challenging customer interactions (Bakker & Demerouti, 2007). Job resources refer to aspects of the job, or work characteristics, which help individuals achieve work goals, reduce job demands, or stimulate growth and development (Bakker & Demerouti, 2007), such as job autonomy, feedback, and social support. Tims, Bakker, and Derks (2012) empirically validated four types of crafting: i) increasing structural job resources, when individuals increase job resources such as autonomy, variety, and development opportunities; ii) increasing social job resources, which refers to increasing relational characteristics such as social support and feedback from others; iii) increasing challenging job demands, which is when individuals take on extra tasks, roles or responsibilities which are stimulating as opposed to a hindering; iv) decreasing hindering job demands, when individuals reduce the number of tasks which require sustained effort and are not motivational. Recently, these two job crafting perspectives have been integrated, with researchers showing how they conceptually and empirically relate to each other (Bruning & Campion, 2018; Zhang & Parker, 2019). However, since most empirical intervention research tends to adopt Tims et al.’ (2012) perspective (see Oprea et al., 2019), for reasons of comparability, we adopt the JD-R perspective on job crafting in this research.

Intervention studies are stronger tests of theory than other research designs due to their longitudinal nature and manipulation of key variables of interest (Higgins & Green, 2011). Intervention studies are also important from a practical perspective, demonstrating to organisations and individuals that it is possible to achieve change. In their meta-analysis, Oprea et al.’ (2019) found that creating Job Crafting Plans involving both organizational and personal goals boosted the effectiveness of interventions. Occupation did not have a moderator effect on overall job crafting. These results should be interpreted with caution, however, given the low study numbers (K = 14) which reduces robustness, and the limited number of occupations and settings upon which moderator analyses were based. Other moderators were not tested, suggesting that there is still much to be understood about how and why job crafting interventions work. This includes understanding the impact of workload, and whether the intensity of interventions (i.e. how demanding an intervention is of an individual's time and effort), or the extent that individuals participate (i.e. whether they actually carry out job crafting behaviours) have an impact on intervention effectiveness for increasing job crafting behaviours.

Job crafting intervention studies have demonstrated mixed results. For example, some have reported increases in all types of job crafting behaviours following the intervention (e.g. Van Wingerden et al., 2016, Van Wingerden et al., 2017c; Demerouti, Xanthopoulou, Petrou, & Karagkounis, 2017; Sakuraya et al., 2016), whereas others have shown some positive effects on some job crafting behaviours some of the time (e.g. Van Wingerden et al., 2017a; Van Wingerden et al. 2017b; Demerouti et al., 2017; Van den Heuvel, Demerouti, & Peeters, 2015; Gordon et al., 2018). This suggests that there are important differences in the types of crafting fostered across studies. Why these differences exist, however, is not well understood. We focus on the key role of workload in helping to understand the diverse effects of job crafting interventions, as well as the role of intervention intensity.

JD-R theory suggests that job crafting is only possible when job demands are manageable, implying that people have the capacity and the resources to engage in job crafting behaviours (Bakker, 2011). Workload, as a key job demand, is therefore an important moderator to consider in job crafting interventions. Previous empirical interventions have not considered the moderating impact of job demands on job crafting behaviours.

We draw on COR theory to help explain why workload is an important moderator of job crafting behaviours. COR theory proposes that individuals try to protect their current resources as well as gain resources (Hobfoll, 1989). From this perspective, resources refer to: objects, such as homes, cars, and other personal belongings; personal characteristics, such as self-esteem, your attitude towards events which happen to you, and a sense of mastery over challenges faced; conditions or roles, such as marriage, and work and family roles, and; energies, such as time, money and knowledge (Hobfoll, 1989). Hobfoll suggested that individuals value the resources they have and use them to obtain more of these same resources. Threats from the environment, however, can make people afraid that they will lose these valued resources. In the work domain, threats to one's role, work relationships, and sense of mastery over one's job could occur in the form of job insecurity, work overload, or negative feedback about your work performance from a supervisor, for example. Losing resources can lead to psychological stress (Hobfoll, 1989). In these situations, individuals are motivated to replenish their resources.

When job demands are low or moderate, COR theory predicts that individuals are able to craft their jobs using the resources they have, and therefore are more likely to focus on accumulating more resources for the future so as to proactively protect against possible future resource loss, or to enhance current resources in accordance with personal goals and desires (Hobfoll, 1989). We therefore expect to see that individuals with low to moderate demands will be more likely to use their resources to gain more resources, by increasing behaviours directed towards increasing structural job resources, such as more autonomy from colleagues, social resources, such as developing working relationships with colleagues, and challenging job demands, such as starting new projects.

When job demands are high, however, COR theory suggests that individuals are likely to feel threatened that they will lose their current resources, and are too resource deficient to invest energy in self-initiating behaviours to build more resources. In this situation, individuals are likely to engage in behaviours directed at protecting the resources they do have, such as by managing or reducing their demands. For example, individuals might reduce a high workload by delegating tasks to others or negotiating new deadlines. Proactivity theory supports this view, by proposing that proactive behaviour, such as job crafting, requires initial energy and resources, which are then invested in creating more resources (Parker, Bindl, & Strauss, 2010). According to this view, when workload is high, job crafting aimed at increasing job resources and job challenges will be limited. Instead, individuals are likely to invest any resources they do have in managing those demands, and thus are likely to engage in behaviour to reduce hindering job demands.

Furthermore, person-job fit may be achieved when individuals proactively craft their jobs so that work characteristics are aligned with their needs, goals, and skills (Kristof-Brown, Zimmerman, & Johnson, 2005; Tims and Knight, 2019). Individuals can fulfil these needs and desires by moulding their jobs through job crafting (Tims, Derks, & Bakker, 2016). For example, an individual who particularly likes working with people might actively seek work roles or tasks which involve face-to-face contact with others, or an employee who is keen to learn and develop might actively discuss training opportunities with a supervisor. When workload is manageable, individuals have the capacity to reflect on their work tasks, relationships, and goals, and self-initiate changes to strive towards these, building job resources. When workload is high, however, individuals are likely to perceive a misalignment between person-job fit, motivating them to try and reduce their demands and improve fit.

Despite these strong theoretical underpinnings for workload as a moderator of job crafting interventions, few studies have investigated the role of workload in shaping the success of such interventions, and the results are mixed. A qualitative study found results consistent with theory, with teachers experiencing a high workload and a high pressure environment most commonly reporting job crafting activities directed at reducing job demands. Nevertheless, 55% of these activities were not successfully completed, with lack of time and time pressure cited as explanatory reasons (Van Wingerden, Derks, Bakker, & Dorenbosch, 2013). As such, demands did not decrease following the intervention. This study did not compare the results to a low workload sample, limiting conclusions. In an empirical intervention study, Van Wingerden et al. (2017a) also found that levels of workload did not change following a job crafting intervention, although behaviours to manage hindering job demands increased. The authors speculated that this could have been due to participants becoming more aware of demands following job analysis but it is unclear why workload did not subsequently reduce.

Other, non-intervention studies, have found different results when investigating workload as a moderator of job crafting behaviours. A two-wave, cross-lagged panel study found that role overload was negatively related to job crafting, but this relationship was moderated by perceived adaptivity such that when role overload was high, those who perceived their adaptivity to be high were also able to job craft (Solberg & Wong, 2016). Similarly, Petrou, Demerouti, Peeters, Schaufeli, and Hetland (2012) found that individuals with active jobs characterized by high levels of job resources (i.e., job autonomy) and high levels of job demands (i.e., work pressure) are more likely to invest in crafting job resources and less likely to invest in activities to reduce their job demands (no effect was found for increasing challenging job demands). These cross-lagged and correlational studies, respectively, are unable to determine causality, limiting conclusions and theory development.

In sum, there are limited and mixed results relating to the moderating effect of workload, despite strong theoretical reason to expect such a relationship. Further, there are methodological challenges with both existing intervention and non-intervention studies that limit the conclusions that can be drawn. In accordance with COR theory and proactivity theory, we hypothesize the following moderating effect of work load:

Hypothesis 1

Compared to those with a lower workload, those with a higher initial workload (Time 0) will report, post-intervention, an increase in their levels of job crafting behaviours to decrease hindering demands relative to their pre-intervention levels.

Hypothesis 2

Compared to those with a higher workload, those with a lower initial workload (Time 0) will report, post-intervention, an increase in their job crafting behaviours to increase structural or social job resources, or challenging job demands, relative to their pre-intervention levels.

We expect that the effect of the intervention will be stronger for those who complete an intense intervention (knowledge-reflection-action intervention) involving Job Crafting Boosts (activities) and post-Boost evaluation questions. This is because Boosts are intended to actively prompt and remind individuals to engage in crafting behaviours, and the evaluation questions are intended to encourage reflection, learning, and motivation to carry out more Boosts. COR theory and proactivity theory can help explain why an intensive intervention is likely to have a stronger effect on crafting behaviours than a less intensive intervention. The central tenet is that individuals need initial resources in order to proactively craft (Parker et al., 2010; Tims et al., 2013). Proactivity theory states that proactivity is about ‘making things happen’ (Parker et al., 2010, p. 827), rather than passively ‘letting things happen’. It involves setting a future-oriented goal as well as striving to achieve that goal, and is driven by ‘can do’, ‘reason to’, and ‘energised to’ motivational states (Parker et al., 2010). Parker et al. (2010) argued that the uncertainty associated with self-initiating new behaviours requires high levels of motivation, self-efficacy and positive affect to overcome the ‘risks’ involved in this sort of behaviour (Parker et al., 2010). Providing Boosts makes it easier for individuals to engage in job crafting in the knowledge-reflection-action intervention as the amount of initial effort required to initiate crafting behaviours is reduced. This is because, by providing a list of Boosts which individuals can choose from, and by encouraging individuals to carry them out during their work week, individuals do not need to invest cognitive energy to create their own Boosts, nor do they need to rely solely on their own motivation to carry them out. Altogether, the amount of self-initiative required to engage in these activities is reduced. In contrast, in the knowledge-reflection intervention, individuals would need to plan, as well as carry out, their own ‘Boosts’, or job crafting activities, which would require more initial energy investment, and a higher level of self-initiative throughout the intervention. We would therefore expect the knowledge-reflection-action intervention to result in higher perceptions of job crafting behaviours than the knowledge-reflection intervention.

Furthermore, individuals who engage in Boosts will experience ‘can do’, ‘reason to’ and ‘energised to’ motivational states (Parker et al., 2010) which will drive their crafting behaviour and encourage them to engage in further crafting. More specifically, successfully completing Boosts is encouraging, enhancing self-efficacy and individuals' beliefs that they can change their work design (‘can do’) further, through completing more Boosts. This means individuals are more likely to try out new ways of working. Completing the evaluation questions following each Boost reminds individuals why they are important, thus making salient reasons why they would want to change aspects of their work (‘reason to’). Finally, individuals are likely to experience positive emotions in response to the idea of changing their work for the better and thus invest energy in these changes (‘energised to’). By providing Boosts, we facilitate the experience of these proactive motivational states in individuals in the knowledge-reflection-action intervention and therefore promote engagement in job crafting behaviours.

In line with these arguments, we predict that under conditions of high workload, individuals in the knowledge-reflection-action intervention will be more likely to engage in Boosts to reduce hindering demands than those in the knowledge-reflection intervention. Although COR theory predicts that both groups will be motivated to protect and manage resources when workload is high, we expect to see more engagement with these behaviours in the knowledge-reflection-action intervention as providing Boosts is theorised to decrease the amount of energy individuals need to invest in these proactive behaviours. Similarly, under conditions of low workload, the amount of energy initially required to engage in Boosts in the knowledge-reflection-action intervention will also be reduced, facilitating greater engagement in Boosts directed at gaining resources, than in the knowledge-reflection intervention. Our hypotheses are as follows:

Hypothesis 3

Under conditions of high workload pre-intervention (Time 0), a knowledge-reflection-action (high intensity) job crafting intervention will be more effective than a knowledge-reflection (low intensity) intervention for decreasing hindering demands post-intervention.

Hypothesis 4

Under conditions of low workload pre-intervention (Time 0), a knowledge-reflection-action (high intensity) job crafting intervention will be more effective than a knowledge-reflection (low intensity) intervention for increasing structural job resources, challenging job demands, and social job resources post-intervention.

Previous job crafting intervention studies have rarely, if at all, investigated how much individuals actually engage in crafting after the initial workshops or training. If individuals do not actively participate in interventions, or planned activities are not carried out, intervention effectiveness may be impeded (Carroll et al., 2007; Nielsen & Miraglia, 2017). Carroll et al. (2007) argued that the content of an intervention may be considered its ‘active ingredients’, with participation being one of those key ingredients. According to this view, acceptance of the intervention by participants is crucial to their engagement. Similarly, Argyris (1995) argued that knowledge and practices which are verbalised need to be internalized, adopted, and practised by individuals in order to effect change in organisations. In support, Nielsen and Randall (2013) found that in a teamworking intervention, participation, and changes in work procedures predicted post-intervention autonomy, social support and well-being. In addition, poor implementation, including lack of employee participation, may lead to insignificant findings which obscure the positive impact perceived by people who did invest in the intervention. This is important as it suggests that the intervention could have had wider effects under more conducive circumstances.

Given the above, as part of our analyses, we assess the degree of participation in the ‘knowledge-reflection-action’ intervention, in order to understand whether and how actually carrying out Job Crafting Boosts affects results.¹ In particular, we explore the impact of initial workload level on the extent of participation, in terms of the number and type of Job Crafting Boosts completed, and individual experiences of taking part. COR theory suggests that those with a higher workload are less likely to engage in crafting job resources and taking on new activities. In our study, there were more Boosts overall aimed at increasing resources and challenging job demands than at decreasing hindering demands. We therefore expect those with a higher workload to complete fewer boosts overall. This is likely to impact the success of the knowledge-reflection-action intervention. Further, and in line with hypotheses 2 and 3, we predict that those with a higher initial workload are likely to engage in more Job Crafting Boosts to reduce hindering demands than those with a lower initial workload. Conversely, those with a lower initial workload will carry out a greater number of Job Crafting Boosts to increase structural or social job resources, or challenging job demands, than those with a higher initial workload. Our final hypotheses are thus as follows:

Hypothesis 5

Participants in the knowledge-reflection-action intervention with a higher initial workload (Time 0) will carry out fewer Job Crafting Boosts overall than those with a lower initial workload.

Hypothesis 6

Participants in the knowledge-reflection-action intervention with a higher initial workload (Time 0) will carry out more Job Crafting Boosts to decrease hindering demands than those with a lower initial workload.

Hypothesis 7

Participants in the knowledge-reflection-action intervention with a lower initial workload (Time 0) will carry out more Job Crafting Boosts to increase structural or social job resources, or challenging job demands, than those with a lower initial workload.

In addition, research suggests that the effectiveness of interventions will be shaped by individuals accepting the intervention, applying their newly acquired knowledge to change their work practices (Argyris, 1995; Carroll et al., 2007). Therefore, beyond testing the above hypotheses, we examined themes from an open text evaluation question completed after each Boost to explore people's experiences of engaging in the intervention. Responses provided insight into how individuals felt about the intervention, and their learning and engagement.

We base our study on Australian participants, who we believe face similar work challenges, such as high workloads, to those reported in European countries where much of the job crafting research has been conducted (Oprea et al., 2019). This is because Australia, like European countries, is a developed country with the majority of workers in secondary and tertiary industries. We therefore believe there is an opportunity for job crafting to be used by Australian workers as a means to protect and gain resources, and manage demands.