Enhancing Blended Working Arrangements and Individual Work Performance
Wörtler, Burkhard
DOI:
10.33612/diss.170826946
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2021
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Wörtler, B. (2021). Enhancing Blended Working Arrangements and Individual Work Performance.
University of Groningen. https://doi.org/10.33612/diss.170826946
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Chapter 4
Do Blended Working
Arrangements Enhance
Organizational Attractiveness
and Organizational Citizenship
Behavior Intentions? An Individual
Difference Perspective
Note. This chapter is based on Wörtler, B., Van Yperen, N. W., & Barelds, D. P. H. (2020). Do blended working arrangements enhance organizational attractiveness and organizational citizenship behaviour intentions? An individual difference perspective. European Journal of Work and Organizational Psychology. Advance online publication. https://doi.org/10.1080/1359432X.2020.1844663
Abstract
In blended working arrangements (BWAs), employees have discretion over when and where they work. Although BWAs are proliferating worldwide, the lack of predefined temporal and locational structures is unlikely to appeal to every employee. To investigate with whom and when BWAs cause positive reactions, we conducted two experimental vignette studies among full-time employees. In Study 1, we used a 2 (BWAs: yes vs. no) × 2 (development support: yes vs. no) between-subjects design (N = 212) and, in Study 2, a within-subjects design with the same factors (N = 114). Additionally, in both studies, we measured individual differences in autonomy orientation and personal need for structure. Study 1 showed that, relative to traditional working arrangements, BWAs enhance organizational attractiveness and intention to demonstrate organizational citizenship behavior. In Study 2, in which employees could compare working arrangements, we only found effects of BWAs among employees high in autonomy orientation or low in personal need for structure. Development support for independent working was not found to moderate the effects of BWAs. By indicating which employees tend to prefer BWAs, our findings could help organizations determine employees’ suitability for such arrangements, which is likely to contribute to BWAs being effective.
The world of work is undergoing substantial changes due to societal developments and technological innovations leading – across European Union countries (Eurofound & The International Labour Office, 2017) and worldwide (e.g., Kossek & Michel, 2011) – to working arrangements that are increasingly flexible (see also Felstead & Henseke, 2017; Korunka & Kubicek, 2017; Spreitzer et al., 2017). This development is likely accelerated by the 2020 coronavirus (COVID-19) pandemic which has led to a substantial increase in employees working from home for prolonged periods (Kniffin et al., 2020). The current research examined the possible effects on organizationally relevant outcomes of contemporary working arrangements that provide employees with temporal and spatial flexibility, while also considering the role of individual differences in preferences and psychological needs in these effects from a person– environment fit perspective (e.g., Van Vianen, 2018).
The term blended working refers to a working arrangement that involves blending onsite and offsite working, a situation enabled through modern information and communication technology (ICT; Van Yperen et al., 2014; Van Yperen & Wörtler, 2017a). Blended working enhances employees’ control over where and when their work is accomplished (Van Yperen et al., 2014) and, thus, contrasts with traditional arrangements in which employees are required to be present at the employer’s site and work a relatively fixed schedule (Kalleberg et al., 2000). Various concepts that are similar to blended working have been used in the work and organizational psychological literature. A prominent example is telecommuting, which has also been referred to as telework, virtual work, and flexible work (see Allen et al., 2015). The terminology used differs across studies, and even a single label can have different definitions (Allen et al., 2015). Based on their review, Allen et al. (2015, p. 44) describe telecommuting as “a work practice that involves members of an organization substituting a portion of their typical work hours (ranging from a few hours per week to nearly full-time) to work away from a central workplace— typically principally from home—using technology to interact with others as needed to conduct work task”. Since blended working refers to a work arrangement where workers alternate between “traditional office working and working from home or another location at any time” (Van Yperen & Wörtler, 2017a, p. 157), it involves, but is not limited to, practices such as telecommuting. That is, blended working involves smooth time- and location-independent working where employees have discretion over when and where they work, as long as they get their work done (Van Yperen & Wörtler, 2017a).
Blended working arrangements (BWAs) have the potential to deliver various benefits. For example, they can extend older employees’ employability and smooth
their transition from employment to retirement (Dropkin et al., 2016; Van Yperen & Wörtler, 2017b) and are, therefore, seen as relevant to resolving societal issues linked to aging workforces (Phillips & Siu, 2012). Furthermore, a meta-analysis has linked blended working practices to various desirable outcomes including enhanced performance and job satisfaction (Gajendran & Harrison, 2007). However, BWAs also carry risks. Practices such as working remotely may lead to employees feeling isolated (Beauregard et al., 2019), possibly because maintaining workplace friendships likely requires face-to-face interaction (Sias et al., 2012). Further, ubiquitous ICT availability may lead to employees overusing technology and, consequently, experiencing high levels of stress and exhaustion (Murray & Rostis, 2007). Anecdotal reports indicate that these have been critical issues during the COVID-19 pandemic. This suggests that blended working practices, such as working from home, may not always be favorable and that some individuals might not react positively to BWAs.
In the present research, we pursued two aims. First, we sought to identify positive effects of BWAs, relative to traditional working arrangements, by considering outcomes that focus on benefiting an organization as a whole; namely, perceived organizational attractiveness and employees’ intention to show organizational citizenship behavior. Investigating these criteria is relevant because there is uncertainty as to whether BWAs provide organizational benefits (Rice, 2017). BWAs are likely to benefit organizations when they are used as a defined human resource (HR) practice in strategic HR management. A strategic perspective on HR management is adopted when “human resource practices, procedures, and systems are developed and implemented based on organizational needs” (Baird & Meshoulam, 1988). To enhance their effectiveness and success, and to gain a competitive advantage, today’s organizations need to be perceived as attractive if they are to recruit talented employees (Chapman et al., 2005; Werbel & DeMarie, 2005) who will accept increasing requirements for flexibility (Cañibano, 2019; Höge & Hornung, 2015) and show organizational citizenship behaviors (Podsakoff et al., 2014, 2009). To this end, it is valuable for organizations to know whether, and with whom (i.e., which types of individuals), BWAs increase organizational attractiveness and intentions to show organizational citizenship behavior.
Organizational attractiveness is “reflected in individuals’ affective and attitudinal thoughts about particular companies as potential places for employment” (Highhouse et al., 2003, p. 989). This is relevant for organizations because recruitment involves an attitude-formation process during which individuals form ideas of what being employed by an organization would be like (Breaugh, 2013). Thus,
an increased understanding of how organizational attractiveness can be enhanced should positively influence an organization’s recruitment success; for example, in terms of winning over talented applicants (Chapman et al., 2005). Indeed, applicants’ attraction to organizations was found to be positively associated with intentions to pursue employment with those organizations, which in turn predicted actual job pursuit behavior (Highhouse et al., 2003).
Organizational citizenship behavior (OCB) is formally defined as “Individual behavior that is discretionary, not directly or explicitly recognized by the formal reward system, and in the aggregate promotes the efficient and effective functioning of the organization” (Organ et al., 2006, p. 3). Here, we have considered intentions to show OCB (see also Williams & Shiaw, 1999) that is directed towards benefitting the wider organization (hereafter OCB intentions) such as taking action to enhance the organization’s reputation (Lee & Allen, 2002). Behavioural intention has been recognized as a proximal antecedent of actual behavior (Ajzen, 1991). Organizations value OCB because, as a facet of employees’ overall job performance (Rotundo & Sackett, 2002), such behavior has the potential to increase overall organizational effectiveness (Podsakoff et al., 2009; Yen & Niehoff, 2004).
Our main aim was then to investigate individual and contextual characteristics that might moderate the posited effects of BWAs. BWAs carry challenges including potential work intensification (e.g., Kelliher & Anderson, 2010), greater ambiguity (e.g., Raghuram et al., 2003), and blurred boundaries between work and non-work (e.g., Flecker et al., 2017). Such challenges can be linked to (a) constant connectivity to work through ICT (Leonardi et al., 2010) and (b) a lack of predefined temporal or locational work structures (e.g., Gerdenitsch, 2017). Substantial freedom to choose and a lack of predefined temporal and locational structures are unlikely to be unequivocally appealing for everyone (Gerdenitsch, 2017; Langfred & Moye, 2004). Thus, BWAs may not per se be perceived more favorably than traditional arrangements. When comparing BWAs and traditional working arrangements, employees with different characteristics are likely to differ in how appealing they deem these arrangements. In their review, Beauregard et al. (2019) suggested that conditions where BWAs would be effective include ones where employees prefer such arrangements. As such, empirically identifying with whom BWAs, relative to traditional working arrangements, tend to elicit greater positive reactions, would provide a better understanding of how to potentially boost the likelihood that BWAs actually deliver their intended benefits for employees and for organizations. Relative to the outcomes of blended working practices (see e.g., Allen et al., 2015),
knowledge of the individual differences on which the positive effects of blended working depend is rather limited (Anderson et al., 2015).
Psychological and organizational models and theories, and particularly person– environment fit theory (P–E fit theory; Kristof-Brown et al., 2005; Van Vianen, 2018), posit that individual outcomes are most accurately predicted when environmental and individual characteristics are considered jointly. Here, individual characteristics related to preferences and needs have been recognized as relevant when studying BWAs (Gerdenitsch, 2017; Van Yperen et al., 2014, 2016). Drawing on P–E fit theory, we argue that individual differences in autonomy orientation (Deci & Ryan, 1985) and in personal need for structure (PNS; Thompson et al., 2001) moderate the posited effects of BWAs. The P–E fit paradigm has played a central role in guiding HR practices for decades (Werbel & DeMarie, 2005). When adopting strategic HR management, HR practices are linked to the type of P–E fit that is most compatible with the organization’s strategy, and this is seen as providing the foundation on which to build a competitive advantage (Werbel & DeMarie, 2005). As such, if offering BWAs is part of an organization’s overall strategy, then identifying employees having a preference for BWAs and thus being more suitable for such arrangements (Beauregard et al., 2019), as well as training and developing them to work effectively in such arrangements, would be seen as systematic practices in the organization’s strategic HR management.
Autonomy orientation is an individual difference variable that reflects “a general tendency to experience social contexts as autonomy supportive and to be self-determined” (Gagné & Deci, 2005, p. 339). From a P–E fit perspective, an individual’s autonomy orientation is relevant when studying BWAs because individuals with a high autonomy orientation are particularly drawn to contexts that allow self-determination and choice (Deci & Ryan, 1985; Gagné, 2003). PNS refers to a desire for clarity and predictability as well as a preference for well-ordered contexts/ situations (Thompson et al., 2001). Thompson et al. (2001) conceptualized PNS as an individual difference variable, implying that the strength of an individual’s need for structure is meaningful when predicting various outcomes. From a P–E fit perspective, PNS is relevant when studying BWAs because, relative to more traditional working arrangements, they lack predefined structures regarding when and where the work is to be done (see also Gerdenitsch, 2017).
In addition to these individual difference variables, we investigated the moderating effect of leadership behavior aimed at developing employees’ skills and competence in working independently (hereafter labelled development support; Amundsen &
Martinsen, 2014), as a contextual factor, on the posited effects of BWAs. The reason for investigating development support was that research and wider discussions indicate that employees need to perceive support when working under BWAs, and that leadership behaviors have an important function in this (Gerdenitsch, 2017; Haines III et al., 2002).
Combining these aspects, we have examined whether the effects of BWAs on organizational attractiveness and OCB intentions vary as a function of autonomy orientation, PNS, and development support (see Figure 4.1). An enhanced understanding of for whom and when BWAs are appealing is urgent given that BWAs are becoming increasingly available and necessary due to rapid advances in ICT devices and infrastructure (e.g., Van Yperen et al., 2014) and unforeseen circumstances such as the COVID-19 pandemic.
Blended working arrangements
Outcome Variables ▪ Organizational attractiveness ▪ OCB intentions
Development support Individual Difference Variables
▪ Autonomy orientation ▪ Personal need for structure
Figure 4.1. Conceptual research model.
Organizational Attractiveness and OCB Intentions as Outcomes of BWAs
An integral feature of BWAs is that they allow greater job-related autonomy and control and, for this reason, they are regarded as a job resource (Thompson et al., 2015; Van Yperen et al., 2016). One would intuitively expect the prospect of gaining autonomy through BWAs to be generally appealing to employees because this is likely to be perceived as facilitating the achievement of work-related goals, a means to deal with job demands, and an opportunity for personal development, learning, and growth (Bakker & Demerouti, 2007). Individuals also strive for autonomy because experiencing a sense of autonomy is an inherent human need (Deci & Ryan, 2000) whose satisfaction is positively linked to experiencing positive affective-motivational states (e.g., Wörtler et al., 2020). BWAs could be perceived as an opportunity to fulfil this need (Deci et al., 2017; Gerdenitsch et al., 2015; Irak & Mantler, 2018). Furthermore, freedom to decide when and where to work allows
employees to more flexibly take care of non-work-related responsibilities (e.g., personal appointments) and meeting demands such as caring for children and elderly relatives (Thompson et al., 2015; Van Yperen & Wörtler, 2017a). Given these perceived advantages of BWAs over traditional working arrangements, it is likely that BWAs will enhance employees’ evaluations of organizational attractiveness. Indeed, a previous experimental study by Thompson et al. (2015) showed that opportunities for time- and location-independent working increased students’ perceptions of organizational attractiveness. Other research has revealed that blended working practices are positively related to organizational commitment (Eaton, 2003) and negatively related to turnover intention (Gajendran & Harrison, 2007). These findings indicate that BWAs are associated with employees’ bonding with an organization and, as such, provide indirect evidence that employees feel attracted to organizations that offer BWAs.
Thompson et al.’s (2015) study also showed that BWAs increased students’ perceptions that an organization valued their contributions and cared about their wellbeing. Such perceptions, in turn, are likely to evoke employee reactions that benefit the organization (Baran et al., 2012). Put another way, favorable organizational incentives that are perceived as supportive tend to be reciprocated by employees. For example, employees are likely to reciprocate by showing OCB (Baran et al., 2012). OCB that is directed towards the organization is “likely to be a direct function of what employees think about their work characteristics” (Lee & Allen, 2002, p. 133). Taking these together, we predicted that:
Hypothesis 1. There is a positive effect of BWAs, relative to an absence of BWAs,
on organizational attractiveness and on OCB intentions.
The Moderating Roles of Autonomy Orientation and Personal Need for Structure
Since the usage of BWAs (Shockley & Allen, 2010) and their perceived effectiveness (O’Neill et al., 2009; Van Yperen et al., 2014) are related to individual characteristics, BWAs may appeal only to certain individuals. Consistent with this idea, P–E fit theory emphasizes the importance of the “compatibility between an individual and a work environment that occurs when their characteristics are well matched” (Kristof-Brown et al., 2005, p. 281). Compatibility or fit between individuals and organizations has been labelled person–organization fit. Such a fit can be achieved when an organization accommodates employees’ needs, desires, and preferences (Kristof, 1996). Individuals form perceptions about their fit with an organization (Cable & Judge, 1996; Kristof, 1996), and the level of perceived/subjective fit is linked to various individual-level outcomes including organizational attraction
(Kristof-Brown et al., 2005; Yu, 2014) and OCB (Cable & DeRue, 2002; Kristof-(Kristof-Brown et al., 2005; Lauver & Kristof-Brown, 2001). Evidence from a meta-analysis suggests that the subjective perception of fit is among the strongest antecedents of an individual’s attraction to an organization (Chapman et al., 2005).
Findings from previous studies seem to be consistent with the underpinning idea of P–E fit theory. O’Neill et al. (2009) found that a need for autonomy was more strongly related to self-rated performance and to job satisfaction among employees who worked away from the conventional workplace through ICT than among employees who did not. In a sample of business students, Luse et al. (2013) found that personality predicted a preference for working in a team where members are geographically dispersed using technology. In particular, individuals who tended to be curious and willing to explore new ideas were found to prefer participating in such virtual teams. This may be because open individuals perceive virtual teams as an opportunity to explore new ideas within a non-traditional setting (Luse et al., 2013), which may be indicative of a subjective fit. Similarly, Anderson et al. (2015) demonstrated that the association between working remotely and job-related positive affect is stronger among employees high in openness to experience. Thompson et al. (2015) found that the positive effect of BWAs on organizational attractiveness was stronger for students who preferred to integrate their work and their non-work roles. Using a sample of workers from the US, Van Yperen et al. (2016) showed that an opportunity for blended working may maintain employees’ intrinsic work motivation in demanding work contexts, but only for workers having a strong desire to experience a work-related sense of choice and psychological freedom. In the current research, we posit, based on the tenets of P–E fit theory, that individual differences in autonomy orientation (Deci & Ryan, 1985) and PNS (Thompson et al., 2001) are particularly pertinent individual difference variables in the BWAs domain, and that these variables are likely to moderate the effects of such arrangements on organizational attractiveness and on OCB intentions.
Autonomy orientation as a moderator
The cost of autonomy perspective (Langfred & Moye, 2004) posits that even though autonomy is generally considered to be a favorable aspect of one’s job, it can be unfavorable if individuals do not endorse it or when it is perceived as a subtle form of control. For example, BWAs may be perceived as an implicit expectation to work overtime (Putnam et al., 2014). In this context, an important consideration is that individuals differ in their orientation to contextual elements, such as the opportunity for choice, when initiating/regulating their behavior (i.e., causality orientations; Deci & Ryan, 1985, 2000; Gagné, 2003). Individuals who have a strong
autonomy orientation are likely to perceive autonomy- supporting contexts, such as BWAs, as providing control rather than as controlling (Vansteenkiste et al., 2010). We posit that a strong autonomy orientation makes individuals prone to identify/recognize BWAs as an opportunity to experience a sense of choice and to show self-determined behavior (Gerdenitsch et al., 2015; Irak & Mantler, 2018). Autonomy-oriented individuals’ inclination to choose contexts that offer autonomy, such as jobs that allow initiative and making choices (Gagné et al., 2018), supports this assumption. Furthermore, Li et al. (2016) found that the association between perceived empowering leadership behavior (i.e., being given more work-related control) and employees’ OCB varied as a function of employees’ autonomy orientation. Overall, individuals with a high autonomy orientation are likely to perceive a fit with BWAs. On this basis, we expected:
Hypothesis 2. There is only a positive effect of BWAs, relative to an absence of
BWAs, on organizational attractiveness and on OCB intentions if autonomy orientation is high.
Personal need for structure (PNS) as a moderator
Individuals with a high PNS have a strong preference for well-ordered, structured situations (Thompson et al., 2001), and therefore an aversion to ambiguous situations. Since BWAs lack the clear, predefined structures found in more traditional working arrangements, they create greater ambiguity (e.g., Gerdenitsch, 2017; Van Yperen et al., 2014). On this basis, it has been suggested that PNS should be considered a key individual difference variable when studying BWAs (Gerdenitsch, 2017).
In the absence of structure, individuals with a high PNS tend to experience discomfort (Thompson et al., 2001), which should logically make them prone to avoid participating in BWAs. In a similar vein, research has shown that employees who express a strong need for structure tend to perceive blended working as ineffective (Van Yperen et al., 2014). Other research found that a low tolerance for ambiguity (i.e., a strong PNS) was associated with more feedback-seeking behavior (Ashford & Cummings, 1985). However, employees engaging in BWAs are usually expected not to rely on close supervision and to be capable of independently organizing and managing their work schedule and methods to fulfil their job requirements (Beauregard et al., 2019). Despite modern ICTs, working at non-traditional times and away from the main work site means it can take longer to receive feedback and help from co-workers and superiors, particularly if all or most staff, including
supervisors, are also engaged in blended working. Overall, individuals with a high PNS are unlikely to perceive a fit with BWAs. On this basis, we posited:
Hypothesis 3. There is only a positive effect of BWAs, relative to an absence of
BWAs, on organizational attractiveness and on OCB intentions if PNS is low.
The Moderating Role of Development Support
For employees to be effective in autonomy-supporting contexts, they need to be self-reliant and able to lead themselves effectively (Amundsen & Martinsen, 2014). BWAs provide substantial control over the work environment, requiring employees to be able to manage their time and organize their activities well (Beauregard et al., 2019; Lapierre & Allen, 2012). In providing development support, leaders aim to facilitate employees’ learning and development in order to enhance their skills in dealing with autonomous working (Amundsen & Martinsen, 2014). More specifically, development support involves demonstrating effective self-leadership behavior so as to continuously enhance employees’ competences to independently plan and organize their workdays and activities (Amundsen & Martinsen, 2014). As such, development support could be regarded as a competence-supporting technique aimed at increasing employees’ sense of effectiveness. Individuals thrive when they experience a sense of both autonomy and effectiveness (Deci & Ryan, 2000). Further, opportunities to make choices seem to be most beneficial to individuals who feel confident of making favorable decisions (Vansteenkiste et al., 2010). Accordingly, it has been suggested that being given the opportunity for self-determination and decision-making should be accompanied by development support to ensure that employees can adequately deal with the resulting work-related autonomy (Amundsen & Martinsen, 2014). One might therefore expect that BWAs will only be evaluated favorably when development support is provided because employees will then perceive their organization as being interested in facilitating their effective blended working experience. On this basis, we posited:
Hypothesis 4. There is only a positive effect of BWAs, relative to an absence of
BWAs, on organizational attractiveness and on OCB intentions if development support is provided.
The Present Research
We tested the above hypotheses in two survey-based experimental studies using the vignette methodology (Aguinis & Bradley, 2014). We relied on written vignettes to manipulate the contextual variables (BWAs and development support) because this research methodology has the advantage of offering both internal and external
validity (Aguinis & Bradley, 2014). That is, this methodology combines the merits of traditional lab experiments and traditional survey methods (Engelmann, 2017). In terms of achieving internal validity, if we had opted for real decision and judgment situations, we could not have isolated the effects of BWAs on organizational attractiveness and on OCB intentions from other factors that influence these desirable outcomes (see also Engelmann, 2017). In comparison, vignette studies allow an experimental design that manipulates the independent variable, which is crucial for making causal inferences (Rubin, 1986). Experimental vignette studies also ensure a certain degree of external validity because the research participants (i.e., the employees in the current research) are presented with concrete scenario descriptions. Here, specific and targeted judgments on organizational attractiveness and OCB intentions can be obtained, offering an advantage over responses to general survey questions (see also Engelmann, 2017).
Furthermore, findings from studies using the vignette methodology may provide strongest internal and external validity when both a between-subjects design and a within-subjects design are used. A between-subjects design is most appropriate when seeking to establish causal links because confounding effects can be excluded (Charness et al., 2012). As MacKinnon et al. (2012, p. 4) put it: “Random assignment of subjects to experimental conditions is the gold standard for making causal inference about the relationship between two variables”. Further, a between-subjects design also allows a conventional manipulation check on the manipulated factors because there are experimental conditions that can be compared based on the manipulation check scores. Conversely, a within-subjects design can optimize external validity in vignette research because, by administering multiple vignettes, individuals have a point of reference on which to contextually ground their judgments. Reference points enable individuals to appropriately judge the advantages and disadvantages of BWAs relative to an absence of BWAs. This enhances realism (i.e., external validity) because comparing organizational policies is what individuals would typically do during an actual job search.
Given the unique advantages of each design, we adopted both designs in the current research. Specifically, the first experiment (Study 1) used a between-subjects design including random assignment to the experimental conditions. Study 1 enabled us to (a) test a manipulation procedure for BWAs and development support through written vignettes and (b) provide initial and clear evidence that the hypothesized links between BWAs and organizational attractiveness and OCB intentions are indeed causal. The second experiment (Study 2) used a within-subjects design. Here,
by administering multiple vignettes, participants could compare organizational policies as they could during an actual job search.
Study 1
Method
Participants and design
Participants were recruited through Prolific, a commercial crowdsourcing platform that is focused on providing a subject pool for online research (Palan & Schitter, 2018). The participants received compensation equivalent to ₤2.20 for their time. We set system qualifications to only sample US citizens residing in the United States who were employed full-time and between 18 and 65 years of age. Of 355 submissions, 336 were complete. Complete submissions were removed from the dataset if at least one of the following criteria was met: Participants’ responses revealed that they did not meet all of the criteria requested through the system qualifications (n = 45), an attention check item was failed (n = 72), or participants disqualified their own submission, for example by indicating that they were sometimes dishonest (n = 49). Some of the removed submissions met multiple rejection criteria.
The final sample comprised 212 employees (57% male) who were randomly assigned to conditions in a 2 (blended working arrangements: yes vs. no) × 2 (development support: yes vs. no) between-subjects design. Participants’ ages ranged from 18 to 62 years (M = 35.90, SD = 9.63) and their job tenure ranged between one and 33 years (M = 4.94, SD = 4.58). In terms of the highest education level achieved, more than half of the participants (51%) had completed a four-year bachelor’s degree to conclude their education, 17% had completed a master’s degree, 16% had some college education but no degree, 9% an associate college degree, 3% a professional degree such as a JD or MD, 3% a high school degree or equivalent, 2% had not completed a high school degree, and 1% had a doctoral degree. A slight majority of the participants did not work in a leadership position (51%). The participants worked in a variety of industries of which the information technology sector (15%), the healthcare and social assistance sector (13%), and the financial and business consultancy sector (13%) were the most prevalent.
Procedure, manipulation, and materials
The participants accessed a survey through a hyperlink provided on Prolific. The survey software Qualtrics randomly assigned the participants to one of the four experimental conditions. At first, all participants completed measures assessing their autonomy orientation and their PNS before providing socio-demographic
information. After that, they were administered an experimental manipulation involving BWAs and development support (see below). After the manipulation, the participants completed a measure of organizational attractiveness and OCB intentions as well as an attentiveness check, manipulation checks, and finally evaluated their own response behavior. We describe the manipulation and measures below.
Experimental manipulation of BWAs and development support
Given our 2 (blended working arrangements: yes vs. no) × 2 (development support: yes vs. no) between-subjects design, we used four vignettes of hypothetical organizations that participants might well encounter when job searching (see Appendix). Some of the general instructions and vignettes were adopted from Thompson et al. (2015). When vignettes are used in between-subjects designs,
providing context has been recommended because each participant only reads one vignette, thereby disabling a referent point when making judgements (Aguinis & Bradley, 2014). To this end, we provided background information in the general instructions that was identical for all participants. Specifically, participants were asked to imagine they were seeking employment in a position that did not have leadership/manager/supervisor responsibilities and that they would be supervised by a line manager. To provide further context, participants were informed about the benefits package and salary (adopted from Thompson et al., 2015). This information also enhanced the realism of the vignettes since it informed the participants about relevant aspects in a job search (see also Harold & Ployhart, 2008).
The experimental manipulations addressed the working arrangements and leadership practices in the organization to manipulate the BWAs and development support factors. In the BWAs conditions, participants read that employees could decide when and where they worked, provided they got their work done. This description reflects the core features of blended working (i.e., time- and location-independent working; Van Yperen et al., 2014). Participants were informed that this working arrangement involves dealing with flexible working conditions (i.e., variable schedules for performing work-related activities, varied work settings, and blurred boundaries between work and non-work). In the no BWAs conditions, participants read that employees are required to work a fixed (8 a.m. – 5 p.m.) schedule at the main work site. Participants were informed that this working arrangement involves dealing with structured working conditions (i.e., a predictable and routine schedule for performing work-related activities, a fixed work setting, and clear boundaries between work and non-work).
In the conditions offering development support, participants read that their line manager’s primary function was to be an observable role model who regularly displays effective self-leadership skills, with the goal of continuously developing employees’ skills and competences to independently structure, schedule, and organize their workdays and work-related activities (see Amundsen & Martinsen, 2014). In the no development support conditions, participants read that their line manager’s primary function was to be an advisor who could be reached out to, with the aim that employees can reliably receive assistance when questions or issues arise concerning work-related activities.
Manipulation check
As a check on the BWAs manipulation, participants were asked to respond to four items (α = .98) on the nature of their working arrangement. The items being:
“Employees are free to work at any time they want to, provided they get their work done”, “Employees are required to work at the main work site” (reversed item), “Employees are required to adhere to a traditional (8 a.m. – 5 p.m.) work schedule” (reversed item), and “Employees deal with flexible working conditions”.
To check on the development support manipulation, participants were asked to respond to another four items (α = .74): “Your line manager’s primary function is
to be an observable role model by regularly displaying effective self-leadership skills”, “Your line manager’s goal is to reliably provide assistance to employees when questions or issues arise concerning work-related activities” (reversed item), “Your line manager’s goal is to continuously develop employees’ skills and competences to independently structure, schedule, and organize their workdays and work-related activities”, and “Your line manager’s primary function is to be an advisor who can be reached out to” (reversed item). The participants responded using a seven-point scale ranging from (1) strongly disagree to (7) strongly agree to respond to both manipulation check scales. The respective items were averaged after recoding reversed items to compute a manipulation check scale score for the ‘BWAs’ and ‘development support’ factors.
Organizational attractiveness
To measure perceptions of organizational attractiveness, we used five items developed by Highhouse et al. (2003). Using a seven-point scale ranging from (1)
strongly disagree to (7) strongly agree, the participants responded to items such as
“This organization is attractive to me as a place for employment”. The item scores were averaged to create a scale score (α = .95).
OCB intentions
To measure intentions to show organizational citizenship behavior directed towards the organization, we slightly adapted the eight-item scale developed by Lee and Allen
(2002). Using a seven-point scale ranging from (1) strongly disagree to (7) strongly
agree, the participants responded to items such as “I would attend functions that are
not required but that help the organizational image.” The item scores were averaged to create a scale score (α = .93).
Autonomy orientation
We used the 12-item General Causality Orientations Scale (GCOS-12; Deci & Ryan,
1985) that includes scenarios of social situations. For each of the 12 scenarios, the GCOS-12 includes a response to assess the respondent’s autonomy orientation. For example, the participants read the scenario “You had a job interview several weeks ago. In the mail you received a form letter which states that the position has been filled. It is likely that you might think:” The inserted corresponding response is designed to assess the respondent’s autonomy orientation. In the above example the response is “Somehow they didn’t see my qualifications as matching their needs.” The participants responded by using a seven-point scale ranging from (1)
very unlikely to (7) very likely. The item scores were averaged to create a scale score
(α = .71).
Personal need for structure (PNS)
Personal need for structure was measured using the 12-item PNS scale developed by Thompson et al. (2001). Using a response scale ranging from (1) strongly disagree to (7) strongly agree, participants rated their level of agreement with items such as “I find that a consistent routine enables me to enjoy life more.” The item scores were averaged to create a scale score (α = .90).
Statistical analysis
We used separate hierarchical multiple regression analyses for the dependent variables (organizational attractiveness and OCB intentions) to test the hypotheses. We included the measured individual difference variables (autonomy orientation and PNS) in separate regression models. Following guidelines provided by Cohen et al. (2003), we mean-centered autonomy orientation and PNS before computing the interaction variables. In the first step of the model building, we entered the BWAs and development support factors as well as the mean-centered autonomy orientation or PNS variable. Hypothesis 1 was tested in this step. In the second step, we added the two-way interaction variables to the model to test Hypotheses 2, 3, and 4.
Results
The means, standard deviations, and correlations among the continuous variables are shown in Table 4.1.
Table 4.1. Means, Standard Deviations, and Correlations of the Continuous Variables in Study 1
Variable M SD 2 3 4
1. Autonomy orientation 5.70 0.62 -.19 .05 .17
2. Personal need for structure 4.66 0.99 - .12 -.00
3. Organizational attractiveness 5.46 1.32 - .62
4. OCB intentions 5.18 1.05
-Note. N = 212. Correlations above .13 and .18 (in absolute values) are significant at the p = .05 and p = .01 level, respectively.
Manipulation checks
A 2 (blended working arrangements: yes vs. no) × 2 (development support: yes vs. no) analysis of variance (ANOVA) on the manipulation check scores of the BWAs revealed only a significant main effect of BWAs, F(1, 208) = 1443.24, p < .001,
η2
p = .87. Participants given the BWAs conditions scored higher (M = 6.35, SD = 1.01)
on the manipulation check than participants given the no BWAs conditions (M = 1.53, SD = 0.81). A similar two-way ANOVA on the manipulation check scores of development support revealed only a significant main effect of development support, F(1, 208) = 102.19, p < .001, η2
p = .33. Participants given the development
support conditions (M = 4.48, SD = 1.24) scored higher than participants given the no development support conditions (M = 2.78, SD = 1.22). These results clearly indicate that the experimental manipulations were effective.
Hypothesis testing
The results of the hierarchical regression analyses are shown in Table 4.2 (including autonomy orientation) and Table 4.3 (including PNS). In Hypothesis 1, we posited
that, relative to no BWAs, there would be a positive effect of BWAs on organizational attractiveness and on OCB intentions. Indeed, BWAs resulted in higher organizational attractiveness ratings (M = 6.03, SD = 0.98) than no BWAs (M = 4.81, SD = 1.36) and also in stronger OCB intentions (M = 5.43, SD = 1.02 vs. M = 4.88, SD = 1.02). As shown in Tables 4.2 and 4.3, the regression slopes corresponding to both mean differences differed significantly from zero. Thus, Hypothesis 1 was supported.
bl e 4 .2 . R es ul ts o f t he H ie ra rc hi ca l M ul tip le R eg re ss io n A na ly si s I nc lu di ng A ut on om y O ri en ta tio n i n S tu dy 1 O rgan iz at io na l at tr ac ti ve n es s O CB in te ntio n s ep Var iab le 1 2 1 2 B le nd ed w or ki ng a rr an ge m en ts ( B W A s) 1. 23 ** * (0 .1 6) 1 .2 9 ** * (0 .2 3) 0. 58 ** (0 .1 4) 0.6 5 ** * (0 .2 0) D ev elopm en t s upp or t ( D S) 0. 05 ( 0. 16 ) 0. 09 ( 0. 23 ) 0. 21 ( 0. 14 ) 0. 27 ( 0. 20 ) Au tonom y or ien ta tion 0.1 9 (0 .1 3) -0 .4 0 ( 0. 24 ) 0. 32 ** (0 .11 ) 0. 12 ( 0. 20 ) B W A s × D S -0. 05 (0. 32 ) -0 .1 0 ( 0. 28 ) B W A s × A ut onom y or ien ta tion 0. 30 ( 0. 26 ) 0. 03 ( 0. 22 ) D S × A ut onom y or ien ta tion 0.7 7 ** (0 .2 6) 0. 32 (0. 23 ) Fchang e 19 .5 2 ** * 3. 34 * 9. 08 ** * 0. 74 df Fch ang e 3 , 2 08 3 , 2 05 3 , 2 08 3 , 2 05 R² .22 ** * .26 ** * .1 2 ** * .1 3 ** * ∆R ² .22 ** * .0 4 * .1 2 ** * .01 ote . N = 2 12 . U ns ta nd ar di ze d co effi ci en ts (b ) a nd s ta nd ar d er ro rs (SE ; i n pa ren the se s) a re s how n. B len de d wo rk in g ar ra ng emen ts a nd d ev elo pmen pp or t w er e b ot h c od ed 0 = n o ( re fe re nc e c on di tio n) , 1 = y es . *p < . 05 , **p < . 01 , ** *p < . 001 .
Ta bl e 4 .3 . R es ul ts o f t he H ie ra rc hi ca l M ul tip le R eg re ss io n A na ly si s I nc lu di ng P er so na l N ee d f or S tr uc tu re i n S tu dy 1 O rgan iz at io na l at tr ac ti ve n es s O CB in te ntio n s St ep Var iab le 1 2 1 2 1 B le nd ed w or ki ng a rr an ge m en ts ( B W A s) 1.1 9 ** (0 .1 6) 1. 22 ** (0 .2 4) 0. 55 ** (0 .1 4) 0.6 8 * (0 .2 1) D ev elopm en t s upp or t ( D S) 0. 06 ( 0. 16 ) 0. 06 ( 0. 24 ) 0. 23 ( 0. 14 ) 0. 33 ( 0. 21 ) Pe rs on al n ee d f or s tr uc tu re ( PN S) 0. 10 ( 0. 08 ) 0. 30 ( 0. 16 ) -0 .0 3 ( 0. 07 ) 0. 01 ( 0. 14 ) 2 B W A s × D S -0 .0 7 ( 0. 33 ) -0 .2 3 ( 0. 29) B W A s × P N S -0 .3 0 ( 0. 17 ) -0 .1 5 (0 .1 5) D S × P N S -0 .0 8 ( 0. 17 ) 0. 07 ( 0. 15 ) Fchang e 19 .2 5 ** 1.1 0 6.0 9 * 0.6 3 df Fch ang e 3 , 2 08 3 , 2 05 3 , 2 08 3 , 2 05 R² .22 ** .23 ** .0 8 * .0 9 * ∆R ² .22 ** .01 .0 8 * .01 N ote . N = 2 12 . U ns ta nd ar di ze d co effi ci en ts (b ) a nd s ta nd ar d er ro rs (SE ; i n pa ren the se s) a re s how n. B len de d wo rk in g ar ra ng emen ts a nd d ev elo pmen su pp or t w er e b ot h c od ed 0 = n o ( re fe re nc e c on di tio n) , 1 = y es . *p < . 01 , **p < . 001 .
4
There was no evidence for an interaction effect between BWAs and autonomy orientation on either organizational attractiveness or OCB intentions. Similarly, the results did not provide evidence of an interaction effect between BWAs and PNS on either organizational attractiveness or OCB intentions. Therefore, Hypotheses 2 and 3 were rejected. Furthermore, we did not obtain evidence for an interaction effect between BWAs and development support influencing either organizational attractiveness or OCB intentions. Consequently, Hypothesis 4 was also rejected.
Study 2
Contrary to our expectations, the results of Study 1 did not provide evidence that the effects of BWAs on organizational attractiveness and on OCB intention vary as a function of autonomy orientation, PNS, or development support. A possible explanation is that the employees did not have the opportunity to compare all four organization descriptions as they could do during an actual job search. If there had been an opportunity to compare the organizations, the individual judgments might have been more realistic because the employees would have more points of reference when making judgments (Aguinis & Bradley, 2014; Charness et al., 2012). During comparisons, the features of each arrangement that are important to each employee may become more salient, and this may facilitate or lead to more accurate judgements and perceptions (such as of subjective fit). We therefore posit that comparing organization descriptions will be conducive to revealing the moderating influence of third variables. Therefore, we conducted a conceptual replication study in which we used a within-subjects design that allows organizations to be compared.
Method
Participants and design
Participants were recruited through Amazon’s crowdsourcing platform Mechanical Turk (hereafter MTurk), which has been evaluated as a reliable source for obtaining good quality data (e.g., Buhrmester et al., 2011). The participants received monetary compensation of $2.50 for their time. We sought to obtain a comparable sample to that used in Study 1. We again set system qualifications such that, to be able to participate, one had to be from the United States and employed full-time (35 or more hours per week). Following recommendations for obtaining good data quality using MTurk samples (Mason & Suri, 2012; Peer et al., 2014), we also set system qualifications that participants needed to have a record of at least 50 approved tasks on MTurk and a high ratio (above 97%) of approved versus submitted tasks.
We obtained 312 submissions of which 280 were complete. Completed submissions were removed from the dataset if participants met at least one of the following criteria: indicated that they were not employed for at least 35 hours per week (n = 27), were older than 65 years (n = 3), disqualified their submission at the end of the study, for example by indicating that they were sometimes dishonest (n = 18), provided an incorrect response to an instructed response item designed to detect careless responding (Meade & Craig, 2012; n = 17), or provided an incorrect response to an attention check item (n = 88). A submission was also excluded from the dataset if a participant did not correctly respond to five items that referred to the experimental manipulation (n = 117; see also the description of the items in the
Procedure and Measures section below). Some of the rejected submissions met more
than one of removal criteria and thus appear in multiple counts. Such high rates of inattentiveness are not unusual in online research and do not pose a problem if inattentive respondents are removed from the dataset (Fleischer et al., 2015). The final sample comprised 114 (52% male) employees. All these employees indicated that they were US citizens. Participants’ ages ranged from 22 to 65 years (M = 41.11, SD = 11.41) and their job tenure ranged from one to 42 years (M = 6.12,
SD = 6.22). In terms of the highest education level achieved, more than half (53%)
of the participants had completed a four-year bachelor’s degree, 16% a master’s degree, 11% had a college education but no degree, 10% an associate college degree, 6% a high school degree or equivalent, 3% a professional degree such as a JD or MD, and 2% had completed education without a high school degree. Slightly more than half (54%) of the participants did not work in a leadership position. The participants worked in a variety of industries of which the information technology sector (18%) and the retail trade and catering sector (also 18%) were the most prevalent. We used a full-factorial 2 (blended working arrangements: yes vs. no) × 2 (development
support: yes vs. no) within-subjects design.
Procedure and measures
Participants accessed the study via a hyperlink provided on MTurk, followed a similar procedure to that described for Study 1, and were asked to respond to the same measures of the study variables. More specifically, the participants first responded to measures assessing their autonomy orientation (α = .75) and PNS (α = .94). Each participant was then provided with all four vignettes (i.e., organizational descriptions) and were then expected to complete the measures of organizational attractiveness for each description (Cronbach’s alphas were between .91 and .95) and OCB intentions (Cronbach’s alphas ranged from .93 to .95). To counteract order and sequence effects, the four vignettes were presented
to the participants in a random order. Until the participants had submitted their ratings of organizational attractiveness and OCB intentions for all four hypothetical organizations, they could, at any time, revisit the vignettes, which were presented on a single screen. Once this task was completed, participants were expected to respond to five items that assessed whether the participants discriminated between manipulated characteristics of the organizations and characteristics that were held constant (see also Thompson et al., 2015). The items were: “Did the organizations differ on whether or not they allowed employees to decide when they work?” Did the organizations differ on whether or not they allowed employees to decide where they work?”, “Did the organizations differ in terms of the leadership practice?”, “Did the organizations differ on whether or not they offered a competitive salary?”, and “Did the organizations differ on whether or not they offered a benefits package?” The last two items were adopted from Thompson et al. (2015) and the first three were self-developed. Responses to these items could be either ‘yes’ or ‘no’. A conventional check on the manipulation was not feasible because, by the nature of the within-subjects design, the participants were not assigned to different experimental conditions.
Statistical analysis
As the dependent variables (organizational attractiveness and OCB intentions) were each assessed four times, we had a total of 456 observations/repeated measures nested within 114 individuals. This hierarchical structure of the data produces dependencies in the four ratings of organizational attractiveness and OCB intentions by each participant. To take account of this dependence, we carried out a multilevel regression analysis (Snijders & Bosker, 2012) using the nlme package (Pinheiro et al., 2020) developed for the statistical software R (R Core Team, 2019) to test our hypotheses. As we had predicted interaction effects involving the continuous
variables (autonomy orientation and PNS), we grand-mean-centered these variables before computing the interaction variables to facilitate interpretation of the interaction effects (Cohen et al., 2003).
As in Study 1, separate analyses were conducted for autonomy orientation and for PNS. Then, separately for the dependent organizational attractiveness and for OCB intentions variables, several nested random-intercept regression models were computed (Hoffman & Rovine, 2007). First, we fitted a null model that did not include any predictor variables. The null model served as a baseline/comparison for evaluating later model fits. In two subsequent steps, as in Study 1, we first entered the main effects and then the two-way interaction variables. Significant interaction effects identified were followed up through an analysis of simple slopes.
We examined the slopes at a high (i.e., one SD above the mean) and a low (i.e., one
SD below the mean) value of the continuous moderator variable (Cohen et al., 2003).
In all models we used the recommended restricted maximum likelihood (REML) estimator (Snijders & Bosker, 2012). We applied a deviance test to compare the fit of the models (Finch et al., 2014; Snijders & Bosker, 2012). This deviance test assesses whether adding predictor variables to a model results in a statistically significant improvement in model fit. To perform the deviance tests, we first had to re-specify all the models using maximum likelihood estimations (see Finch et al., 2014; Snijders & Bosker, 2012).
Results
The means, standard deviations, and correlations of the continuous variables are shown in Table 4.4.
Table 4.4. Means, Standard Deviations, and Correlations of the Continuous Variables in Study 2
M SD
Level 2
1 2
1. Autonomy orientation 5.77 0.67 –
2. Personal need for structure 4.67 1.25 -.33*
M SD
Level 1
1 2
1. Organizational attractiveness 5.48 1.30 –
2. OCB intentions 5.19 1.08 .59*
Note. Level 1: N = 456. Level 2: N = 114. *p < .001 (two-tailed).
Hypothesis testing
The results of the multilevel regression analysis related to the dependent organizational attractiveness and OCB intentions variables are summarized in Table 4.5 (including autonomy orientation) and Table 4.6 (including PNS). In Hypothesis 1, we posited that, relative to an absence of BWAs, there would be a positive
effect of BWAs on organizational attractiveness and on OCB intentions. Indeed, on average, participants rated an organization as more attractive if BWAs were provided (M = 5.91, SD = 1.18) than if they were not (M = 5.04, SD = 1.27). Likewise, OCB intentions were stronger if BWAs were provided (M = 5.41, SD = 1.06) than if they were not (M = 4.97, SD = 1.06). As shown in Tables 4.5 and 4.6, the multilevel regression results supported a positive effect of BWAs on both organizational attractiveness and OCB intentions (as evidenced by positive regression slopes that differed statistically significantly from zero). Thus, the results again supported
Ta bl e 4 .5 . R es ul ts o f t he M ul til ev el R eg re ss io n A na ly si s I nc lu di ng A ut on om y O ri en ta tio n i n S tu dy 2 O rgan iz at io na l at tr ac ti ve n es s O CB in te ntio n s St ep Var iab le 1 2 3 1 2 3 1 ‘N on e’ ( nu ll m od el ) 2 B le nd ed w or ki ng a rr an ge m en t ( B W A s) 0. 87 ** (0 .11 ) 0. 82 ** (0 .1 4) 0. 44 ** (0 .0 7) 0. 46 ** (0 .1 0) D ev elopm en t s upp or t ( D S) -0 .1 5 (0 .1 1) -0 .2 0 ( 0. 14 ) -0 .0 7 ( 0. 07 ) -0 .0 6 ( 0. 10 ) Au tonom y or ien ta tion 0 .1 4 ( 0. 10 ) -0 .2 7 ( 0. 15 ) 0 .52 ** (0 .11 ) 0. 33 * (0 .1 3) 3 B W A s × D S 0 .1 0 ( 0. 20 ) -0 .0 2 ( 0. 13 ) B W A s × A ut onom y or ien ta tion 0 .9 0 ** (0 .1 5) 0 .4 1 ** (0 .1 0) D S × A ut onom y or ien ta tion -0 .0 8 ( 0. 15 ) -0 .0 4 ( 0. 10 ) Le ve l 1 v ar ia nc e ( σ 2) 1. 55 1 .3 0 1 .1 9 0 .6 1 0 .5 4 0 .52 Le ve l 2 v ar ia nc e ( τ00 ) 0 .1 3 0 .1 9 0 .2 2 0 .5 7 0 .4 7 0. 48 D ev ia nce 15 26 .7 8 14 61. 86 14 28 .3 0 12 42 .6 8 118 1. 31 11 64 .5 9 df 3 6 9 3 6 9 ∆D ev ia nce 6 4.9 3 ** 3 3. 56 ** 6 1. 36 ** 16 .7 3 ** N ote . N = 4 56 ob se rv at io ns ar e ne st ed in N = 1 14 em pl oy ee s. U ns ta nd ar di ze d co effi ci en ts (γ ) a nd st an da rd er ro rs (SE ; i n pa re nt he se s) ar e sh ow n. B le nd ed w or ki ng a rr an ge m en ts a nd d ev el op m en t s up po rt w er e b ot h c od ed 0 = n o ( re fe re nc e c on di tio n) , 1 = y es . *p < . 05 . **p < . 001 .
4
bl e 4 .6 . R es ul ts o f t he M ul til ev el R eg re ss io n A na ly si s I nc lu di ng P er so na l N ee d f or S tr uc tu re i n S tu dy 2 O rgan iz at io na l at tr ac ti ve n es s O CB in te ntio n s ep Var iab le 1 2 3 1 2 3 ‘N on e’ ( nu ll m od el ) B le nd ed w or ki ng a rr an ge m en t ( B W A s) 0. 87 * (0 .11 ) 0. 82 * (0 .1 4) 0. 44 * (0 .0 7) 0. 46 * (0 .0 9) D ev elopm en t s upp or t ( D S) -0 .1 5 (0 .1 1) -0 .2 0 ( 0. 14 ) -0 .0 7 ( 0. 07 ) -0 .0 6 ( 0. 09) Pe rs on al n ee d f or s tr uc tu re ( PN S) 0 .0 4 ( 0. 05 ) 0. 29 * (0 .0 8) -0 .0 3 ( 0. 06 ) 0 .1 2 ( 0. 07 ) B W A s × D S 0 .1 0 ( 0. 20 ) -0 .0 2 ( 0. 13 ) B W A s × P N S -0 .55 * (0 .0 8) -0. 30 * (0 .0 5) D S × P N S 0 .0 5 ( 0. 08 ) -0 .01 (0 .0 5) Le ve l 1 v ar ia nc e ( σ 2) 1 .5 5 1. 30 1 .1 5 0 .6 1 0 .5 4 0 .5 0 Le ve l 2 v ar ia nc e ( τ00 ) 0 .1 3 0 .2 0 0 .24 0 .5 7 0 .5 9 0 .6 1 D ev ia nce 15 26 .7 8 14 63 .1 5 14 17. 78 12 42 .6 8 12 01 .8 5 11 70 .9 5 df 3 6 9 3 6 9 ∆D ev ia nce 6 3. 63 * 4 5. 37 * 4 0. 84 * 3 0.9 0 * ote . N = 4 56 ob se rv at io ns ar e ne st ed in N = 1 14 em pl oy ee s. U ns ta nd ar di ze d co effi ci en ts (γ ) a nd st an da rd er ro rs (SE ; i n pa re nt he se s) ar e sh ow n. le nd ed w or ki ng a rr an ge m en ts a nd d ev el op m en t s up po rt w er e b ot h c od ed 0 = n o ( re fe re nc e c on di tio n) , 1 = y es . *p < . 001 .
In Study 2, in contrast to Study 1, we found evidence for the predicted two-way interaction effects which we posited in Hypotheses 2 and 3. In Hypothesis 2, we proposed that there would only be a positive effect of BWAs, relative to an absence of BWAs, on organizational attractiveness and on OCB intentions if autonomy orientation was high. Here, the results did support an interaction effect between BWAs and autonomy orientation on organizational attractiveness and on OCB intentions. Figure 4.2 shows that BWAs had a positive effect on organizational attractiveness when autonomy orientation was high, γs= 1.42, SE = 0.18, p < .001 but that there was no evidence for such an effect when autonomy orientation was low, γs= 0.22, SE = 0.18, p = .21. Similarly, Figure 4.3 shows that BWAs had a positive effect on OCB intentions when autonomy orientation was high, γs= 0.73, SE = 0.12, p < .001 but no such effect when autonomy orientation was low, γs= 0.18, SE = 0.12, p = .13.
Figure 4.2. Moderating effect of autonomy orientation on the effect of blended working arrangements on organizational attractiveness in Study 2. ‘Low’ and ‘high’ represent values of 1 SD below and above the mean, respectively.
Figure 4.3. Moderating effect of autonomy orientation on the effect of blended working arrangements on OCB intentions in Study 2. ‘Low’ and ‘high’ represent values of 1 SD below and above the mean, respectively.
In Hypothesis 3, we posited that there would only be a positive effect of BWAs, relative to an absence of BWAs, on organizational attractiveness and on OCB intentions if PNS was low. The results indeed supported an interaction effect between BWAs and PNS on organizational attractiveness and on OCB intentions. Figure 4.4 shows a positive effect of BWAs on organizational attractiveness when PNS was low, γs= 1.51, SE = 0.17, p < .001. However, when PNS was high, there was no evidence for a positive effect of BWAs on organizational attractiveness, γs= 0.13, SE = 0.17, p = .45. Similarly, Figure 4.5 shows that the results supported a positive effect of BWAs on OCB intentions when PNS was low, γs= 0.83, SE = 0.11, p < .001 but not when PNS was high, γs= 0.08, SE = 0.11, p = .47. Further, as shown in Tables 4.5 and 4.6, no evidence was found for an interaction effect between BWAs and development support that influenced either organizational attractiveness or OCB intentions. Consequently,
Figure 4.4. Moderating effect of personal need for structure (PNS) on the effect of blended working arrangements on organizational attractiveness in Study 2. ‘Low’ and ‘high’ represent values of 1 SD below and above the mean, respectively.
Figure 4.5. Moderating effect of personal need for structure (PNS) on the effect of blended working arrangements on OCB intentions in Study 2. ‘Low’ and ‘high’ represent values of 1 SD below and above the mean, respectively.
Discussion
Blended working arrangements (BWAs) provide employees with substantial control over where and when they do their work and are proliferating following societal and technological developments. We have investigated whether BWAs,
relative to traditional working arrangements, increase employees’ perceptions of organizational attractiveness and their intention to show OCB directed towards the organization. Moreover, we have investigated whether these effects vary as a function of individual differences in autonomy orientation and PNS as well as contextual development support aimed at facilitating autonomous working. By considering these potential contingency factors, we have nuanced understanding of the organizational benefits (i.e., perceived organizational attractiveness and employees’ OCB intentions) associated with BWAs.
Across two experimental vignette studies, both utilizing samples of full-time employees, using a between-subjects design (Study 1) and a within-subjects design (Study 2), we provide consistent evidence that organizations that offer BWAs, rather than traditional working arrangements, are perceived as more attractive potential employers. This finding conceptually replicates the main effect found by Thompson et al. (2015). Replications are “critical for establishing the generalizability of an initial observation and the theory it purports to support” (Crandall & Sherman, 2016, p. 94). Whereas Thompson et al. (2015) used a student sample, our findings generalize the existence of the effect to the population of full-time employed individuals. Both of our studies also show that BWAs elicit stronger intentions to show organization-directed OCB, that is, discretionary behaviors that benefit an organization (e.g., Podsakoff et al., 2009). This finding dovetails with previous findings on potential organizational benefits of new ways of working, such as greater organizational commitment and lowered turnover intentions (Eaton 2003; Gajendran & Harrison, 2007).
Future research could seek to identify the mechanisms underlying the effects of BWAs on perceived organizational attractiveness and on OCB intentions. On the affective level, a theoretically conceivable explanation is that positive affective-motivational states mediate these effects. Specifically, autonomy-supporting work contexts, such as BWAs, are likely to elicit positive affect linked to the anticipated experience of work-related self-determination and choices. This is likely to trigger both a positive attitude towards the organization and a motivation to show positive work behaviors (see Deci et al., 2017). On a rational level, being offered BWAs rather than traditional arrangements will likely increase perceptions of fairness and/or
support (Lee & Kim, 2018; Thompson et al., 2015). Such perceptions may not only increase the perceived organizational attractiveness but, based on the social-exchange principle, also OCB intentions (e.g., Baran, 2012).
However, our main finding is that BWAs, relative to traditional working arrangements, do not always evoke positive reactions. Specifically, in Study 2, the positive effects of BWAs on both organizational attractiveness and OCB intentions were limited to (a) employees with a strong autonomy orientation and (b) employees who had a weak PNS. Given the very similar samples used in our two studies, the fact that these moderation effects were only found in Study 2 requires an explanation. Our suggestion is that the between-subjects design in Study 1 did not provide the employees with points of reference when making judgements. The results of Study 1 suggest that BWAs have a general, positive appeal. If the alternatives are unknown, an organization that offers BWAs tends to be perceived as a more attractive potential employer than an organization that offers traditional working arrangements. However, if the alternatives are known, enabling comparisons between working arrangements, individual differences become important (Study 2). In such an explicit comparison, the advantages of traditional arrangements, including clarity and structure, can be more clearly identified. That is, when the features of both working arrangements are salient during their comparison, employees can better judge which arrangements fit their preferences and needs. Consistent with this reasoning is the argument that a within-subjects design “helps to show the effects of a manipulation within one individual and is useful in terms of uncovering judgment processes of a single individual” (Aguinis & Bradley, 2014, p. 361). In other words, within-subjects designs may be better at detecting the moderating effects of individual differences.
Our moderation findings are consistent with the meta-analysis finding that an individual’s perceived fit with an organization is a predictor of attraction-related outcomes (Chapman et al., 2005) and they add to the literature on individual differences as moderators of the benefits of BWAs. In terms of autonomy orientation, our findings are consistent with those indicating that employees who express the desire for autonomy in their job tasks are (a) more likely to perceive blended working as effective (Van Yperen et al., 2014; see also O’Neill et al., 2009) and (b) less likely to lose their intrinsic work motivation in a demanding work context if there is an opportunity for blended working (Van Yperen et al., 2016). Our findings are also in line with those obtained by Li et al. (2016). These authors found that employees’ perceptions of being empowered by their supervisors (i.e., receiving greater work autonomy/control) were positively associated with increased
