The dark and potentially bright sides of work-avoidance goal orientation

Ebert, Thea; Först, Rieke; Bipp, Tanja

doi:10.3389/forgp.2024.1445014

ORIGINAL RESEARCH article

Front. Organ. Psychol., 13 December 2024

Sec. Work Motivation and Participation

Volume 2 - 2024 | https://doi.org/10.3389/forgp.2024.1445014

The dark and potentially bright sides of work-avoidance goal orientation

$\r\nThea Ebert$ Thea Ebert

Rieke Först^*

Tanja Bipp

Department of Psychology, Ruprecht-Karls-Universität Heidelberg, Heidelberg, Germany

Goal orientations represent a key approach for explaining employee motivation and performance. While a trichotomous framework (learning, performance-approach, and performance-avoidance goal orientations) dominates the work literature, goals regarding the work process, for example, minimal work expenditure, fall outside the scope. This is reflected in a lack of theoretical differentiation and validated instruments for assessing work avoidance goals in the work context. However, a minimization of effort among employees can pose a risk to the goals of many organizations. Therefore, we extend goal orientation theory at work by building on educational research and introducing the dimension of work avoidance goal orientation to the work context. In three studies of German employees (N₁ = 115, N₂ = 224, N₃ = 121), including cross-sectional and cross-lagged data, we developed a reliable and construct-valid scale to assess work-avoidance goal orientation at the workplace. Furthermore, we reveal contradicting implications from an organizational (prediction of absenteeism/withdrawal behaviors) and an individual perspective (protecting influences in the form of reduced demands), thereby offering starting points for future research and organizational practices that seek to better differentiate in the lower performance and motivation ranges of employees.

1 Introduction

In the past decades, goal orientation (GO) theory has become a central approach to explaining behavior in achievement situations (Vandewalle et al., 2019). Representing a motivational construct that refers to the goals individuals pursue in achievement settings, GO represents an essential predictor for work-related outcomes such as performance (Nerstad et al., 2018), engagement (Daumiller and Dresel, 2020), or wellbeing (Rinas et al., 2022). Beyond direct relationships, GO is increasingly included in mediational and moderational analyses to inform more comprehensive research on work-related processes. For instance, it has been found to predict work engagement directly and indirectly through job crafting (Matsuo, 2019) and to be a moderator on cognitive appraisal in stressor–performance relationships (Ma et al., 2021).

Although different theoretical conceptualizations of GO have been suggested, a three-factor model is frequently applied in organizational contexts (Vandewalle et al., 2019). It distinguishes learning (the goal to develop one's competence by increasing skills), performance-approach (the goal to demonstrate competence by outperforming others and gaining favorable judgments), and performance-avoidance goals (the goal to avoid showing a lack of competence and avoiding negative judgments). However, this trichotomous framework neglects one possible component of achievement motivation: the goal of avoiding work. Work-avoidance (WA GO) refers to the desire to reduce effort, do as little as possible, and not work hard (Duda and Nicholls, 1992) and has been part of several conceptualizations of achievement goals in educational contexts (e.g., Meece et al., 1988). While learning and performance-approach goals have proven helpful in understanding high achievement (Linnenbrink-Garcia et al., 2008), performance-avoidance and WA goals have allowed differentiation at the lower end of the achievement spectrum in students (Tuominen-Soini et al., 2011; Wormington and Linnenbrink-Garcia, 2017). However, this side of achievement goals has not been widely applied to work settings, except with a specific educational lens (e.g., WA GO of teachers or university instructors; Daumiller et al., 2019). Extending GO theory at work to incorporate WA GO is essential for achieving a comprehensive understanding of motivational goals at work, including their roles in both high and low achievement.

Furthermore, recent GO research has moved beyond simplistic categorizations of GO as purely beneficial or detrimental, instead advocating for a more nuanced investigation of how different orientations relate to various outcomes (Vandewalle et al., 2019). Consequently, we contribute to exploring whether WA GO of employees has potentially contradictory implications for individuals and organizations: While, for example, reducing effort in the work process may be counterproductive and negatively impact organizational outcomes, it could also help individuals avoid excessive workloads and reduce burnout risks.

To address these questions, our research extends GO theory in the workplace by adding WA GO as a fourth dimension and examining its outcomes from both organizational and individual perspectives. Specifically, our studies have three major aims: (1) to develop and validate a scale for assessing WA GO among employees, demonstrating its distinction from existing GO dimensions; (2) to investigate the predictive role of WA GO for negative work behaviors, focusing on counterproductive work behavior (CWB); and (3) to further examine its predictive validity through potentially protective influences of WA GO for individuals in terms of reduced demands and thus decreased exhaustion.

2 Theory and hypotheses

2.1 Extending GO theory at work: WA goals

Although GO research has its roots in studying motivational differences among students (e.g., Nicholls et al., 1985), the concept refers to explaining behavior in achievement situations. This also applies to the workplace in general, where employees often have to meet high and rapidly changing demands and increasingly intense work (e.g., Mauno et al., 2023) while their supervisors or peers evaluate their performance and development. Supporting the theoretical core of GO theory that goals act as proximal antecedents in terms of energizing, directing, and guiding behavior in achievement situations, they are related to a wide range of work-related outcomes, such as job performance, deviance, or employees' wellbeing (Payne et al., 2007; Cellar et al., 2011). Comparable to evidence from educational contexts, learning GO was mainly found to have favorable effects in the work context, supporting the notion of GO theory as a comprehensive approach to explaining achievement motivation across domains (Van Yperen et al., 2014).

Although the trichotomous framework of GO is vastly utilized and supported by empirical evidence, already more than 20 years ago, several authors noted an essential gap in achievement goal theory: These goals do not seem to cover the whole spectrum of possible motivational orientations (Brophy, 1983; Nicholls, 1989). Therefore, on a theoretical level, adding WA GO to the GO framework has been suggested (e.g., Duda and Nicholls, 1992). Originally, WA GO was postulated as the absence of an achievement goal (e.g., Elliot, 1999) or the desire to reduce effort, do as little as possible, and not work hard (Duda and Nicholls, 1992). To allow deeper comprehension of this GO, comparing WA GO to related constructs is essential. Given that achievement motivation research involves a wide range of concepts, we employ motivated action theory (DeShon and Gillespie, 2005) to draw parallels between WA GO and similar constructs at different levels of abstraction. This hierarchically structured model categorizes GO at an intermediate level of achievement motivation concepts, which are informed by more general self or principal goals and, in turn, manifest themselves in more specific action goals and behaviors. At the most basic level of self or principal goals, we recognize connections with the general achievement motive, especially considering that its “fear of failure” facet is also strongly avoidance-oriented. Based on the notion that specific GOs manifest themselves based on such general motives (Bipp and van Dam, 2014), the achievement motive could represent a related prerequisite for WA GO that has not yet been clarified. At the GO level, WA GO overlaps with the performance-avoidance dimension. However, a significant difference arises from the reasons underlying the avoidance tendencies. In a performance-avoiding GO, avoidance arises from a fear of failure and appearing incompetent (Noskeau et al., 2021). In contrast, a WA GO is characterized by a lack of interest in demonstrating competence (Middleton and Midgley, 1997). Success in the sense of a WA GO is defined as the minimal effort possible instead of any measure of achievement (Elliot, 1999). Regarding the most specific level of action plan goals and behaviors, WA GO mainly shows similarities to procrastination. Procrastination as a work-avoiding behavioral tendency (Yan and Zhang, 2022) could result from a dispositional WA GO. However, the core motivation of a WA GO is to avoid or reduce work altogether in contrast to postponing it and waiting until just before a deadline to deal with it.

Previous research in educational contexts provides valuable insights into the potential consequences of a WA GO. Students who score high on WA GO tend to cognitively disengage from classes (Nicholls et al., 1985), show reduced intrinsic motivation (Hidi and Harackiewicz, 2000), and ultimately receive worse grades (Spinath et al., 2002). In a comprehensive GO profile study, students pursuing predominantly a WA GO before other GO dimensions showed the lowest levels of social and emotional wellbeing, with students endorsing particularly low-performance-approach goals (Wormington and Linnenbrink-Garcia, 2017). In contrast, students with a distinctively low WA GO reported the most adaptive achievement outcomes and learning strategies.

While outcomes like disengagement or lower performance in educational contexts mainly affect individual pupils and their success in school, the effects of employees' WA GO could reach significantly further. For example, students showing a high WA GO and associated behavior do not cause additional expenditure for other students. However, a WA GO for employees, fueling avoidance and withdrawal behaviors, such as disengagement, emotional withdrawal, quiet quitting, or avoidance crafting, may lead to colleagues having to deal with excessive workloads as a result of compensation for avoided tasks (Tims et al., 2015), as well as lower organizational outcomes due to productivity losses (Ogunfowora et al., 2022). At the same time, an individual employee showing a high WA GO and associated behavior might, in fact, experience workload relief when compensation through colleagues is possible. In contrast, students must generally bear the consequences of their disengagement themselves, leaving little room for protective influences of a WA GO in educational contexts. However, compared to the thoroughly researched learning and performance orientations, the literature regarding WA GOs at work is sparse and inconsistent (King and McInerney, 2014). To our knowledge, only a handful of studies offer preliminary ideas about WA GOs at work, and they mainly focus on teachers or university lecturers and, thus, a unique form of employees in a particular organizational environment. For example, in a longitudinal study, Wang et al. (2017) found a negative association between teachers' WA GO and positive affect while teaching. Daumiller et al. (2019) replicated this finding for university instructors and reported a relation between WA GOs and perceiving the act of seeking help as a threat. Recently, WA GO has been shown to be associated with higher burnout in researchers (Daumiller and Dresel, 2020). Regarding more distal outcomes, the WA GO of instructors showed a negative relation with students' ratings of teaching (Daumiller et al., 2016), as well as a higher desire to quit (Jagacinski et al., 2020).

To assess the construct reliably and evaluate its utility for the general workplace, we first develop a measurement instrument for WA GO—as a distinctive GO—specifically for employees. Although individual existing GO measures also include some items on WA GO (e.g., Daumiller et al., 2016, 2019), these have been developed explicitly for use in the educational context and thereby do not include perspectives from a broader work-related background. Furthermore, evidence for the construct validity for these scales is sparse but is needed to be widely used for research and in practice. In line with findings from the educational context that a WA GO can be separated from other GOs (King and McInerney, 2014), we expect that employees' WA GOs also show construct validity in this sense. Therefore, we expect our newly developed WA GO scale for the general work context to measure WA GO as a distinctive dimension of GO in employees.

Hypothesis 1: WA GO can be separated from learning, performance-approach, and performance-avoidance GOs at work.

2.2 Organizational perspective: WA GO and CWB

The literature on organizationally relevant outcomes of GO primarily focuses on performance. Research has consistently demonstrated positive associations, especially between learning and performance-approach GOs, and various performance measures (Payne et al., 2007; Van Yperen et al., 2014). Transferred again from educational contexts, findings suggest to the contrary that a WA GO predicts disaffection and negatively influences engagement and achievement (King, 2014). As mentioned earlier, Daumiller et al. (2016) showed that the WA GO of instructors had a negative relation to students' ratings of teaching. Compared to the traditional trichotomous GO structure, WA GO has been found to account for additional variance in negatively predicting organizational citizenship behavior (OCB) of employed college students (Jagacinski et al., 2020).

Therefore, empirical evidence from educational research and initial findings in the specific organizational context of teachers point primarily to the negative consequences of WA GO on performance outcomes. However, most of these studies conceptualize performance only through task performance or OCB, limiting the understanding of performance to productive behavior with varying content and magnitude. In contrast, Rotundo and Sackett (2002) suggest a three-component approach to performance, adding CWB to the aforementioned two. Thus, CWB is perceived as “voluntary behavior that violates significant organizational norms and in doing so threatens the wellbeing of an organization, its members, or both” (Robinson and Bennett, 1995, p. 556). To address whether employees' WA GO not only does not support the organization's productivity through lacking performance but can also harm it directly, we examine the predictive power of WA GO for two specific counterproductive work behaviors.

According to Marcus et al. (2002), the most noticeable feature of CWB research so far is its diverse nature. Conceptualizations differ vastly in their degree of abstraction, ranging from simple aggregations of several behavioral dimensions to a general perspective of counter productivity. To investigate the utility of WA GO, we focused on organizationally rather than interpersonally directed CWB (Robinson and Bennett, 1995), specifically the two facets absenteeism/withdrawal and its contemporary, technology-driven continuation, cyberloafing.

First, absenteeism/withdrawal reflects in its core behaviors of avoidance, refusal, or withdrawal at the workplace, such as staying away from work without an excuse or working less in absence of the supervisor (Marcus et al., 2002). Therefore, WA GO and absenteeism/withdrawal share substantive similarities in terms of construct definition and operationalization. WA GO implies motivational tendencies like doing as little as possible at work (Duda and Nicholls, 1992) or avoiding any or challenging tasks (Dowson and McInerney, 2001), whereas absenteeism/withdrawal includes certain behaviors, as shirking unpleasant tasks or pretending to be busy to avoid new tasks. Adapting the perspective of motivated action theory (DeShon and Gillespie, 2005) once again, absenteeism/withdrawal behavior constitutes a plausible action plan or behavior following a WA GO as a GO focusing on reducing effort (Duda and Nicholls, 1992; Jagacinski et al., 2020) in the work process. In this sense, we suggest that absenteeism manifests WA GO on a behavioral level at the workplace.

Hypothesis 2: WA GO positively predicts subsequent absenteeism/withdrawal behavior.

Second, absenteeism/withdrawal behaviors might be particularly easy to realize in modern workplaces characterized by flexible, technology-driven work environments. Griffiths (2003) highlighted that 59% of internet use at work was not related to work. Therefore, we complement the well-established absenteeism/withdrawal facet of CWB with a contemporary expansion. Lim (2002) defined cyberloafing as employees' voluntary use of their organization's internet access for non-work purposes during work time or, more simply, “the IT way of idling on the job” (p. 678). Several authors consider cyberloafing to be a newer version of traditional CWB behaviors (Block, 2001) or, even more specifically, another form of withdrawal facilitated by information technology (Mercado et al., 2017). Again, drawing on motivated action theory (DeShon and Gillespie, 2005), cyberloafing could be a purposeful action following WA GO and the motivation of doing as little as possible (Duda and Nicholls, 1992) at work, similar to absenteeism/withdrawal. Therefore, we expect these motivational tendencies of a WA GO to manifest in the work-avoidant behavior of cyberloafing.

Hypothesis 3: WA GO positively predicts subsequent cyberloafing behavior.

2.3 Individual perspective: WA, job demands, and exhaustion

As WA GO is defined as the desire to reduce effort, do as little as possible, and not work hard (Duda and Nicholls, 1992), it contrasts with the learning and performance GOs regarding competence development and achievement. Several studies in educational contexts point to associations of a WA GO with multiple adverse outcomes, including lowered engagement (e.g., King and McInerney, 2014), less positive affect (e.g., Daumiller et al., 2019), increased shame and boredom (Rinas et al., 2022), reduced intrinsic motivation (Hidi and Harackiewicz, 2000), more superficial information processing (Nolen, 1988), and worse grades (Spinath et al., 2002). In principle, these findings can also be interpreted as mainly detrimental effects of WA GO for employees.

However, many employees in modern work environments struggle to cope with constantly changing, increasing, and more intensive demands (Korunka et al., 2015) and the resulting threats to performance, wellbeing, and health (Alarcon, 2011). Because avoidance-oriented working behaviors can also counter such overload (Bakker and de Vries, 2021), we extend the previously discussed findings and propose potentially beneficial effects of WA GOs for individual workers in terms of reduced demands. Individuals with a high WA GO avoid challenging achievement situations (Dowson and McInerney, 2004) and pursue the central goal of reducing effort (Jagacinski et al., 2020), thereby lowering quantitative demands, such as their workload. Furthermore, they seek possibilities for not working hard (Duda and Nicholls, 1992) and prefer easy tasks (Meece et al., 1988), mitigating qualitative demands (e.g., task complexity) as well. Consequently, through the different inherent facets of avoidance, a WA GO can be assumed to motivate a reduction of both quantitative and qualitative demands for individual employees in the work context.

Hypothesis 4: WA GO is associated with reduced (a) quantitative and (b) qualitative demands at work.

Following a central preposition of the job–demands–resources model (Bakker et al., 2023), previous research supports a strong association between demands and employee exhaustion. For instance, meta-analytic findings by Alarcon (2011) show a correlation of ρ = 0.49 between workload and emotional exhaustion as a symptom of burnout. As argued earlier, we adapt the novel view of WA GO as a potential mitigating mechanism in the form of reducing quantitative and qualitative work demands, thereby also indirectly lowering exhaustion.

In addition to these anticipated indirect effects, stress research and theory also suggest other associations of a WA GO with exhaustion. Several studies have found that personality traits are associated with burnout (e.g., Alarcon et al., 2009; Swider and Zimmerman, 2010; Roloff et al., 2022), which can be explained by their effect on the subjective perception and appraisal of work-related and potentially stressful situations (Ma et al., 2021). Given the trait-like nature of GOs in general (e.g., Towler and Dipboye, 2001; Bell and Kozlowski, 2002) and the empirical evidence for the stability of the WA GO (King, 2014), we also assume such influences on the perception of stressful situations and thus exhaustion. In this sense, we expect employees with a high WA GO to perceive stressful job situations as less exhausting as they do not place any achievement expectations on themselves.

Hypothesis 5: WA GO has negative (a) direct, as well as indirect, associations—via (b) quantitative and (c) qualitative demands—with exhaustion.

2.4 Overview of current studies and research designs

We designed three studies to introduce the concept of WA GO to the workplace and test our hypotheses. In Study 1, we built on previous work in educational contexts and developed a scale to assess WA GOs in employees, investigating its construct validity in a cross-sectional survey design by separating it from the three-dimensional GO model at work (H1). In Study 2, we provided further evidence on scale validation and took an organizational perspective: Within a cross-lagged study design, we anticipated predictive effects of WA GOs for two types of following CWB (absenteeism/withdrawal and cyberloafing; H2 and H3). Additionally, taking a possible bright side of WA GO into account, we conducted a further cross-sectional survey and adopted an individual perspective in Study 3, considering a WA GO as a possible coping mechanism for employees in terms of reduced job demands and, consequently, lowering exhaustion (H4 and H5).

3 Study 1: scale development and construct validity

3.1 Method

3.1.1 Participants and procedure

Participants were 115 German employees (73.9% female) from various organizations and industries. They were between 20 and 63 years old (M = 32.29, SD = 11.74) and worked on average 34.09 hours/week (SD = 13.74). Research assistants recruited participants from private and professional networks for the cross-sectional online survey. Participants could participate in a raffle (€20 gift vouchers, approximately US $ 21) as an incentive. Informed consent was obtained (according to the guidelines of the German Psychological Society) after participants received information about how their participation was voluntary, how data privacy would be protected, and how anonymized data would be used.

3.1.2 Measures

3.1.2.1 WA GO

We took a deductive approach to construct the new WA GO scale for the work context. A search of existing construct definitions in the literature (almost exclusively from the educational context; e.g., Meece et al., 1988; Nicholls et al., 1990; Skaalvik, 1997; Spinath et al., 2002; Dowson and McInerney, 2004; Daumiller et al., 2016) resulted in an operational definition: Individuals with a high degree of WA GO strive to keep the workload to a minimum, which is why they prefer simple tasks and avoid challenging situations. The aim is not to develop or demonstrate one's competence but to minimize the effort as much as possible.

With an initial pool of 49 items from the existing literature, we translated all items to German and adapted the wording to fit a general work context. Afterward, all items were rated for conformity with the construct definition, and 18 items were selected to be included in the data collection. We calculated various item characteristics following classical test theory (CTT, e.g., selectivity and skewness), evaluated item contents again, and checked factor loadings in a series of confirmatory factor analyses (CFA). Following this multistep, multi-criterion interactive process, we obtained a final scale encompassing five items (α = 0.86; Ω = 0.81), for example, “At work, my goal is to do as little as possible” and “At work, I am all about not working so hard” (see Supplemental material, Table 1 for all original items and German translations). Items were rated on a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree).

3.1.2.2 Learning, performance-approach, and performance-avoidance GOs

A prior validated German version (Theis and Bipp, 2020) of Vandewalle's (2001) 12-item scale was used to assess the established trichotomous GO framework, with four items assessing each of the three dimensions. Items were presented randomly, and participants rated their approval on a 7-point Likert scale, ranging from 1 (strongly disagree) to 7 (strongly agree).

3.2 Results and discussion

The final five-item WA GO scale showed very satisfactory psychometric qualities and good reliability (α = 0.86, Ω = 0.81). In line with prior findings (Payne et al., 2007), the other GOs had substantial intercorrelations (cf. descriptive statistics and intercorrelations in Table 1). As expected, WA GO revealed the highest correlation with performance-avoidance goals (r = 0.44), followed by learning goals (r = −0.39).

Table 1

Table 1. Descriptive statistics and intercorrelations of study variables (Study 1).

To assess the WA GO scale's distinctiveness and verify its construct validity, we conducted and compared three CFAs. The tested models and their corresponding fit indices are displayed in Table 2. Supporting H1, a four-factor model with distinct but correlated learning, performance-approach, performance-avoidance, and WA GO factors provided the best fit to our data with the following comparisons: four- versus three-factor model: $Δ χ_{(3)}^{2}$ = 71.74, p < 0.001, and four- vs. one-factor model: $Δ χ_{(6)}^{2}$ = 437.95, p < 0.001. Average variance extracted (AVE) coefficients for all four factors exceeded 0.50 as recommended by Fornell and Larcker (1981): AVE_{learning GO} = 0.58, AVE_{performance − approach GO} = 0.61, AVE_{performance − avoidance GO} = 0.52, and AVE_WAGO = 0.56.

Table 2

Table 2. Goodness-of-fit indices for the four-, three- and one-factor models of GO (Study 1).

In conclusion, the first test of the scale to assess WA GOs at work yielded highly satisfactory results. The scale is internally consistent and can be clearly separated from established GO dimensions. Thus, the WA GO scale forms a reliable and valid instrument to assess WA GO in the workplace, allowing us to expand the scope of the existing GO literature and applications at work.

4 Study 2: organizational perspective

4.1 Method

4.1.1 Participants and procedure

We created an online survey with two measurement points (4 weeks apart). Such cross-lagged research designs are the most widely used approach to examine the prospective effects of one construct on another while controlling for autoregressive effects (Orth et al., 2021). Thus, this approach allowed us to investigate the possible effects of individual differences in WA GOs on subsequent CWBs. Additionally, the autoregressive coefficients provided insights into whether the WA GO scale exhibited the intended trait-like stability, comparable to the GO of the trichotomous framework (Payne et al., 2007). We collected data from N = 224 German employees matched T1 and T2 (65.9% female), with ages ranging from 19 to 64 (M = 37.73, SD = 13.05) years. Participants worked on average 35.31 hours/week (SD = 8.28) and stemmed from different companies that represent a breadth of industries, including public administration and government (13.9%) or health services (12.7%). Corresponding to Study 1, we obtained informed consent after participants received information about how their participation was voluntary, how data privacy would be protected, and how anonymized data would be used.

4.1.2 Measures

4.1.2.1 WA GO

We applied the five-item version of the WA GO scale from Study 1 (e.g., “At work, my goal is to do as little as possible” and “At work, I am all about not working so hard”) at both measurement times. Again, our WA GO scale provided high internal consistency in the current study (α = 0.82, Ω = 0.83). To cross-validate the scale, we used data from T1 and applied item response theory (IRT) criteria. To test for homogeneity of items, the Martin-Löf statistic was computed, which yielded no significant results, indicating that the scale can be assumed as unidimensional. Furthermore, we analyzed item thresholds to examine the items' difficulties, graphical displays of the category characteristic curves are included in the Supplementary Figure 1. Further item fit indices (see Supplementary Table 2) also supported the decision of the final scale items. All items demonstrated strong discrimination values, ranging from 1.90 to 3.09, indicating that they effectively differentiate between individuals with varying levels of WA GO (Baker, 2001). Regarding item fit, all chi-square statistics (S-X2 index; Orlando and Thissen, 2000) were non-significant, and the root-mean-square-error of approximation (RMSEA) values were all below.05, indicating excellent fit. Additionally, infit/outfit values fell within acceptable ranges, further confirming a good fit for all items, with no evidence of misfit.

To validate and distinguish our scale for the work context from a specific scale to assess WA GOs in university contexts by Daumiller et al. (2016), we also administered their four items. In a CFA, the model with two latent factors (latent correlation r = 0.74) for the respective scales showed significant improvement compared to a one-factor model that combines all items (p < 0.001).

4.1.2.2 Absenteeism/withdrawal

We measured absenteeism/withdrawal with the 13-item German scale of counterproductive work behavior by Marcus et al. (2002) at both times of measurement. We adapted the original instruction of the scale in two ways. First, we changed the time period of the instruction to adhere to our study outline (behavior in the last 4 weeks; interval between T1 and T2). Second, all items (e.g., “I exceeded a break for more than 5 min”) were rated on a 7-point (instead of a 6-point) Likert scale from 1 (never) to 7 (always) to obtain consistency in the answering format with the other scales in the study.

4.1.2.3 Cyberloafing

We used the short version of the cyberloafing behavior scale by Mercado (2017), which was adapted to German using forward-backward translations (e.g., “Browse nonwork-related websites while you should be working”). Participants estimated their personal frequency of seven different behaviors on a 7-point Likert scale ranging from 1 (never) to 7 (several times a day) at both times of measurement.

4.2 Results and discussion

Tables 3, 4 present the descriptive statistics and intercorrelations among study variables.

Table 3

Table 3. Descriptive statistics of variables for both times of measurement (Study 2).

Table 4

Table 4. Bivariate correlations of central variables for both times of measurement (Study 2).

We tested our hypotheses regarding the predictive value of WA GO for absenteeism/withdrawal and cyberloafing, applying path analysis in AMOS 26. As the hypothesized model, including all auto-regressive and crossed-lagged paths, displayed an unsatisfactory fit; $χ_{(4)}^{2}$ = 21.70, p =0.01; χ²/df = 5.43; CFI = 0.97; RMSEA = 0.141; Standardized Root Mean Square Residual (SRMR) = 0.054; we adjusted the model and removed all non-significant paths. The adjusted model (Figure 1) exhibits a significantly improved and acceptable model fit: $χ_{(9)}^{2}$ = 15.42, p = 0.01; χ²/df = 1.71; CFI = 0.98; RMSEA = 0.096; SRMR = 0.039. In particular, we found support for H2: T1 WA GO was positively related to T2 absenteeism/withdrawal (β = 0.13). In contrast, the cross-lagged effect from WA GO to cyberloafing did not reach significance (H3 rejected). Furthermore, WA GO shows a high autoregressive effect, supporting the theoretically assumed stability of this (trait-like) GO dimension. We also found a significant cross-lagged effect of cyberloafing on absenteeism/withdrawal, demonstrating the interdependence of the two separate CWB constructs.

Figure 1

Figure 1. N = 224. ^*p < 0.05, ^**p < 0.01.

In summary, the WA GO scale has proven to be reliable and valid in a second sample, also holding against further evaluations based on IRT criteria. In terms of predictive validity for organizationally relevant outcomes, WA GO of employees revealed to foster counterproductive behaviors in the workplace in terms of higher absenteeism/withdrawal and thus can have tangible, detrimental effects for the organization.

5 Study 3: individual perspective

5.1 Method

5.1.1 Participants and procedure

Data were collected through an online survey. The final sample consisted of 121 German employees (64.5% female) between 20 and 63 years old (M = 37.35, SD = 11.55). The average of working hours/week was 35.52 (SD = 14.18). Study participation was promoted through professional and private networks. Participants could leave their email addresses to receive an update on the study results and/or participate in a lottery for a €10 gift voucher at the end of the survey. We obtained informed consent after participants were informed about how their participation was voluntary, how data privacy would be protected, and how anonymized data would be used, as in the two previous studies.

5.1.2 Measures

5.1.2.1 WA GO

We measured WA GO using the 5-item scale validated in Studies 1 and 2. Again, the scale provided high internal consistency (α = 0.87, Ω = 0.85).

5.1.2.2 Job demands

We used two subscales from the short scale for workplace analysis (Kurzfragebogen zur Arbeitsanalyse; KFZA) (Prümper and Schneeberg, 2020) consisting of two items each to measure quantitative (e.g., “I am often under time pressure”) and qualitative demands (e.g., “This work involves tasks that are too complicated; e.g. due to no or unclear job descriptions or a lack of qualifications”) at work. Participants indicated their answers on a 5-point Likert scale, ranging from 1 (not true at all) to 5 (completely true). Quantitative demands showed good internal consistency (α = 0.83, Ω = 0.87). However, for qualitative demands, Cronbach's alpha was only borderline acceptable (α = 0.68, Ω = 0.78).

5.1.2.3 Exhaustion

To measure exhaustion, we used the respective subscale from the Oldenburg Burnout Inventory by Demerouti and Bakker (2008). Item order was chosen at random, and participants were asked to indicate their approval on all eight items (e.g., “After work, I usually feel tired and worn out”) on a 4-point scale from 1 (strongly agree) to 4 (strongly disagree). The scale showed good internal consistency (α = 0.84, Ω = 0.89).

5.2 Results and discussion

Descriptive statistics and intercorrelations of the study variables are displayed in Table 5. To test our hypotheses, we conducted path analyses in AMOS 26. The hypothesized model included all possible paths, meaning indirect paths of WA GO through quantitative and qualitative demands on exhaustion as well as a direct one of WA GO on exhaustion. Quantitative and qualitative demands were allowed to correlate. This original model proved to be statistically just identified, including significant standardized paths from work avoidance to quantitative demands (β = −0.25) and directly with exhaustion (β = 0.26), as well as from quantitative demands to exhaustion (β = 0.41). However, the remaining two paths from WA GO to qualitative demands and from qualitative demands to exhaustion, were non-significant.

Table 5

Table 5. Descriptives und intercorrelations of study variables (Study 3).

To reach an identified model to judge its fit with the data, and following the principle of parsimony, we adapted the model by removing these two non-significant paths. The final path model (Figure 2; standardized solution) reached very satisfactory fit indices: $χ_{(2)}^{2}$ = 3.20, p = 0.20; χ²/df = 1.60; CFI = 0.98; RMSEA = 0.071. Although this model trimming is based on empirical consideration, it has to be noted that all three paths that were already significant in the original model were still significant in this solution. Both models partially supported Hypothesis 5, as we found contradicting relationships between WA GO and exhaustion. While, against our prediction (rejecting H5a), WA GO showed a positive direct relation with exhaustion, we found some support for the novel perspective of WA GO as a possible buffer in demanding work situations. In detail (supporting H5b), there was a lowering exhaustion effect via reduced quantitative demands (e.g., a decrease in workload). Based on bootstrap analysis in the final model (based on N = 200 bootstrap samples), these contradicting effects were supported, with an overall positive effect of WA GO on exhaustion (standardized total effect: +0.16), being obtained by a direct positive (standardized direct effect: +0.29), and an indirect negative one via quantitative demands (standardized indirect effect: −0.13; 90% confidence interval of [−0.21, −0.05]). However, we found no support (rejecting H5c) for a connection of WA GO with qualitative demands—the specific nature of the work demands appears to be decisive here. In this respect, our model also showed that qualitative demands are not only unrelated to WA GO but also less relevant for exhaustion. In total, our final model was able to explain 23% of the variance in exhaustion.

Figure 2

Figure 2. N = 121. ^**p < 0.01. Hypothesized, but non-significant paths in dashed lines. Quantitative and qualitative demands were allowed to correlate.

6 General discussion

In line with our study's aims, our findings provide three key insights that significantly broaden GO theory and its application at work. First, we successfully developed a scale to measure WA GO within the broader work context and validated it in three heterogeneous samples of German employees based on CTT and IRT criteria. Besides high reliability and stability, we demonstrated the construct's validity as our WA GO scale can clearly be separated from the three other GOs. Second, we found support for the predictive value of WA GO for absenteeism behaviors and, thus, detrimental consequences for the organization. Third, we also provided an individual and novel perspective of WA GO as a potential protecting factor from high quantitative work demands.

6.1 Theoretical and practical implications

We significantly extend the traditional GO framework to include a fourth distinct and individual dimension that can be clearly separated from the other three dimensions (especially from performance-avoidance goals that show the most considerable content-related overlap). Future studies can now rely on a measurement instrument that offers a substantial and new differentiation of goal orientations in the lower performance and motivational range and regarding expenditure during the working process as opposed to competence as the motivational focus.

Furthermore, we adopted an organizational and individual perspective and provided initial evidence for the dual nature of WA GO. Thus, we join the recent theoretical developments in GO theory that criticize the simplified division into adaptive and detrimental GO and argue for a more differentiated examination of their diverse relationships with outcomes (Vandewalle et al., 2019). Regarding detrimental consequences, WA GO significantly predicted subsequent absenteeism/withdrawal 4 weeks later, manifesting in, for example, exceeding break times, calling in sick even though being healthy, or working less without a supervisor present. However, we could not show this relationship for another, more digital variant of CWB, cyberloafing. Our findings align with similar ones from the educational context that indicate, for example, that individuals with a dominant WA GO show more superficial information processing (Nolen, 1988), worse grades (Spinath et al., 2002), and, therefore, a lack of achievement. Extending these findings, we illustrate that WA GO in the workplace can lead not only to lower performance or willingness to perform (King, 2014) but also to CWB, harming the organization's objectives in the end. Regarding an individual perspective on WA GO, prior findings stem mainly from the educational context. Most of these paint a predominantly negative picture, showing, for example, lessened positive affect (e.g., Daumiller et al., 2019) or increased shame and boredom (Rinas et al., 2022). In our third study, we tested and supported a new and contrary perspective: Our results indicate a negative association of WA GO with quantitative demands, as well as contradicting associations with exhaustion. While these findings must be interpreted with the limitations of the cross-sectional nature of the data in this study in mind, they nevertheless provide a novel and important indication of possible mitigating effects of WA GO in stressful work environments.

Considering these insights, practitioners find themselves between the poles of the dark and bright sides of WA GO. The associations of employee WA GO with CWB call for organizational strategies to manage and redirect WA GO and associated work behavior. Reduced quantitative demands and absenteeism/withdrawal of employees means that necessary tasks may no longer be completed. As a result, organizational demands are no longer met, leading to a possible decline in productivity. Furthermore, the avoided demands could be shifted to other employees, who would have to compensate for the loss of productivity and thus risk overwork, burnout, or interpersonal conflicts. Given these risks, the specific work behaviors prompted by WA GO might need to be managed or channeled toward adaptive coping strategies, such as conscious work or task design regarding work autonomy, workload, interdependency, or supervisory feedback and support to combat possible disengagement. Our scale could also be of added value in a selection context. Understanding the relationship between WA GO and CWB can help identify individuals more likely to engage in counterproductive behaviors. However, given the bright side of the construct and the potential problem of response distortion (cf. the rather undesirable work behaviors addressed in the items), our scale should not be used in such a context before future research has provided more insights into the nature of the construct and our developed measure.

Nonetheless, the contradicting and indirect protective effects regarding exhaustion could represent a practical value of a WA GO that is worth preserving under certain circumstances. This new angle appears particularly relevant against the backdrop of increasingly intensified work demands for most employees (Paškvan and Kubicek, 2017) that, in turn, significantly impair wellbeing and motivation (Mauno et al., 2023). In this sense, our findings are also in line with those of Daumiller and Dresel (2020), who adopt a coping perspective with a WA GO in relation to exceeding demands. As GO has also been found to predict stressor appraisal (Ma et al., 2021), WA GO could for example prompt a cognitive reappraisal of intensified work demands, presenting an adaptive coping strategy. Additionally, WA GO tendencies might provide support when used only for a short period or in a targeted way for specific projects, thus creating a stronger sense of priority. Such strategies that employees holding high WA GO use might also inspire interventional approaches to reduce overwork and burnout of employees in the future.

6.2 Limitations and future research directions

Besides the strengths of our studies, such as the three heterogeneous samples from diverse work contexts; cross-sectional, as well as cross-lagged, data; and the consideration of CTT and IRT criteria for scale development, we have to admit certain limitations. All data are based on self-reports, implying the risk of common method bias and distortions, especially in querying counterproductive and thus socially undesirable behaviors. However, we also agree with Sackett (2002), who noted particular restrictions for CWB ratings. Nevertheless, in future research, objective behavioral indicators (Ones and Dilchert, 2013), such as sick days, should also be considered to evaluate the utility of a WA GO at work. Furthermore, all variables show significantly skewed distributions with rather low means, yet we see sufficient variance in all variables.

Because, in the first step, we focused on the predictive value of WA GO for absenteeism and cyberloafing and associations with reduced demands, these findings naturally reflect only an excerpt of possible behaviors following or associated with WA GO. Future research should continue this examination and particularly differentiate antecedents, correlates, and consequences to clarify causal sequences. Meaningful directions for future studies could be explicit and implicit achievement motives and their subcomponents as antecedents of WA GO, disentangling the possible relationships between WA GO and related constructs, such as procrastination or quiet quitting, or investigating further consequences like whether an individual's WA GO might lead to interpersonal conflicts at work. Such explorations could also be enriched by connecting and integrating GO Theory with other theoretical streams. Daumiller et al. (2022), for example, combined GO theory with self-determination theory (SDT; Deci et al., 2017) to explore the interplay of teaching personnel needs and achievement motivations. Following this direction, SDT could contribute to explaining how an environment that fulfills or denies personal psychological needs contributes to developing or manifesting a WA GO and thus complement important processes that precede mechanisms within GO theory (Janke and Dickhäuser, 2018). In particular, our contradictory findings regarding possible protective effects must be examined in the future in terms of their specific mechanisms, for example, by distinguishing between short- and long-term effects. In this context, the stressor-detachment model (Sonnentag and Fritz, 2015), might serve as inspiration to identify possible coping strategies within WA GO.

Furthermore, we developed and validated our scale entirely within a specific cultural sample of German employees. While Daumiller et al. (2022) demonstrated a strict invariance of GO measurement instruments between German, Indian, and U.S. American participants in their international study, Lee et al. (2003), found differences in the perception of GOs between students from the United States and Hong Kong. Future research should explore such cultural characteristics in the application, further development, or translation of the WA GO scale.

To broaden the topic of cultural differences even further, integrating organizational context factors might provide valuable insights into how positive or negative effects of WA GOs occur. The majority of the existing literature on WA GO stems from school or university contexts, which have an inherent focus on learning and development. Although our three studies include broader samples from the general work environment, they are largely based on highly educated white-collar employees. Not only do individuals pursue different GOs, but workplaces can also have different goal structures, thereby influencing the salience or further effects of certain goals via situational factors (Van Dam, 2015). While evidence for the interaction of dispositional and workplace-related leaning GO already exists (Schelp et al., 2023), future research could investigate contexts in which employees' WA GOs might be particularly salient or detrimental, considering situational factors such as work autonomy or interdependence with colleagues.

7 Conclusion

In summary, we provided a validated measurement instrument to assess WA GOs in the workplace, introducing a motivational focus beyond competence and, thereby, improving differentiation in the lower range of performance and motivation. Thus, we extend GO theory at work by a fourth dimension and present insights into its construct and predictive validity for the work setting. In particular, we found potentially contradicting effects with, specific negative effects for the organization in the sense of counterproductive work behavior and, protective mechanisms for individuals in stressful work environments.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement

Ethical approval was not required for the study involving humans in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Author contributions

TE: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. RF: Formal analysis, Visualization, Writing – original draft, Writing – review & editing. TB: Conceptualization, Resources, Supervision, Writing – original draft, Writing – review & editing.

Funding

The author(s) declare financial support was received for the research, authorship, and/or publication of this article. For the publication fee the authors acknowledge the financial support of Heidelberg University.

Acknowledgments

We thank A. Weidner, S. Hofmann, M. Duda, and H. Klenk for their help in collecting the data. A prior version of this paper was presented at the Society for Industrial and Organizational Psychology Annual Conference (April 2022, Seattle, Washington).

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/forgp.2024.1445014/full#supplementary-material

References

Alarcon, G. M. (2011). A meta-analysis of burnout with job demands, resources, and attitudes. J. Vocat. Behav. 79, 549–562. doi: 10.1016/j.jvb.2011.03.007