Dynamics Between Applied Work Demands and Related Competence Beliefs: A 4-Year Study With Scientists

Abstract

In the frameworks of Job Demands-Resources (JD-R) theory and concepts of competence beliefs, we investigated trajectories of and dynamics between demands and competence beliefs relevant to applied work fields. The study had a longitudinal panel design with eight measurement waves (overall study span of 4 years). Participants (38.1% female) were early career scientists from science, technology, engineering, and mathematics fields who either worked at a university (academia group, n = 1,205) or in industry after having previously worked in academia (industry group, n = 436). We conducted bivariate dual change score modeling and found demands to increase in both groups and competence beliefs to increase in the industry group. While demands accelerated change in competence beliefs in the academia group, competence beliefs accelerated change in demands in the industry group. Implications for JD-R theory and concepts of ability-related self-views as well as practice are discussed.

Keywords

occupational competence beliefs ability self-concept job demands applied work bivariate dual change score model

Ability-related self-views are expected to develop across the life span and are influenced by social comparisons, performance feedback, verbal persuasion, physiological/affective states, and other cues of mastery (Bandura, 1986; Marsh & Shavelson, 1985). Specifically, at early career stages, occupational capability judgments were found to be malleable (Tschannen-Moran & Hoy, 2007). Even though it is a crucial challenge to develop expectations about one’s capabilities, knowledge on the formation of work-related competence beliefs is scarce (Frese et al., 2007; McNatt & Judge, 2008; Tierney & Farmer, 2011). Empirical studies demonstrate feedback loops between occupational self-beliefs and career performance as well as positive relations between subjective capabilities and educational or occupational accomplishments, career-related goal setting, motivation, and persistence (e.g., Huang, 2016; Marsh et al., 2017; Spurk & Abele, 2014). Early career decisions (e.g., choosing a field of study) are already influenced by ability-related self-views (Parker et al., 2014). Later in an individual’s career, applying for a position or leaving a job is also predicted by competence beliefs (Sadri & Robertson, 1993). Although the development of capability beliefs has been investigated to some extent (e.g., children’s and adolescent’s competence beliefs: Jacobs et al., 2002), to our knowledge it has rarely been longitudinally studied in relation to job demands during early career phases (e.g., Dicke et al., 2018). Despite its centrality for vocational behavior and decision making, Woods et al. (2013) point out that there are only a few studies longitudinally investigating dynamic processes between occupational and person-related variables (e.g., Frese et al., 2007; Sutin & Costa, 2010). The present article seeks to fill this research gap by longitudinally analyzing the unfolding and bidirectional dynamics of demands and related competence beliefs in early career professionals.

According to the Job Demands-Resources (JD-R) model (Bakker & Demerouti, 2017), job characteristics can be viewed either as demands (i.e., job aspects that require sustained effort or skills) or as resources (i.e., aspects of the job or the person that are functional for work goal achievement and stimulate personal development). In that framework, beliefs about own competencies can be defined as “a task-specific construct related to one’s assessment of level of expertise on a specific task or in a specific setting” (Williams & Lillibridge, 1992, p. 157) and thereby can be seen as personal resources (Akkermans et al., 2013). The investigation of ability-related self-views is mostly done in the context of self-efficacy research (i.e., belief in one’s abilities to master tasks to achieve desired goals; Bandura, 1986) or self-concept research (i.e., perceptions about oneself, comprising own abilities; Marsh & Shavelson, 1985). We refer to both constructs because they address ability-related self-views and were found to largely overlap (e.g., Hughes et al., 2011).

Whereas some domains of ability-related self-views have already been investigated (e.g., academic self-concept or social self-efficacy), other critical occupational domains have not yet been studied. In this study, we address one specific occupational domain, namely “applied work,” relating to a set of tasks that characterize applied work settings (i.e., industry). Whereas applied work demands refer to demands that are frequently required in industrial work settings (e.g., customer focus), applied work competence beliefs can be defined as self-referent capability judgments relating to applied tasks.

Applied Work

For our study, we focused on three facets of applied work: economic focus, business organization, and customer focus. Although we do not believe these three dimensions to represent all potential tasks relevant to applied settings, we consider them essential to many applied tasks. We define the first facet—economic focus—as an emphasis on commercialization, time- and cost-effectiveness, and a striving for profitable goods. As an aspiration for economic efficiency lies at the core of industrial organizations, they need to closely focus on their profitability (Perkmann et al., 2013). The second facet of applied work—business organization—is characterized by a focus on organizational process management, comprising all relevant steps from outlining a project, to its implementation and evaluation. With its focus on the management of entire organizational processes, business organization resembles the concept of process orientation, which emphasizes the end-to-end management of “a whole set of activities” relevant for generating products or services (Khosravi, 2016, p. 116) and which is associated with positive business outcomes, such as customer satisfaction and cost-effectiveness (see Kohlbacher, 2010). The third facet of applied work—customer focus—concerns an aspiration to satisfy customer needs. Customer focus is a core goal for market-oriented companies (Blocker et al., 2011).

Empirical studies with scientists and professionals working in industry and academia further supported our conceptualization of applied work. In a study investigating aspirations of professionals with a background in the same study field but working either in academia or industry (Erez & Shneorson, 1980), the two groups differed regarding their vocational interests. In terms of Holland’s (1997) interest model, professionals working in industry reported stronger preferences for enterprising activities than academics (Erez & Shneorson, 1980). Enterprising-type individuals are considered to prefer activities in which they can lead and persuade others and to frequently work in sales or other customer-centered occupations. They aim at achieving personal and organizational goals, pursuing financial and material accomplishments (Holland, 1997). Compared to academics, industrially working professionals further reported a stronger preference to operate organizational processes (Erez & Shneorson, 1980). Moreover, in a study with PhD students from science and engineering, Roach and Sauermann (2010) found interest in industry to be associated with financial topics. A more recent study further underlined previous findings, revealing early career science, technology, engineering, and mathematics (STEM) scientists who aspired to a position in industry to exhibit a greater focus on economic issues and leadership than those aspiring to a position in academia (Burk & Wiese, 2018).

Early Career STEM Scientists

In science- and technology-based industries, professionals with graduate STEM education play an important role (National Academies of Sciences, Engineering, & Medicine, 2018). Policymakers have encouraged a strong focus on STEM education (Caprile et al., 2015; National Science and Technology Council, 2018). Of all doctorates awarded in the United States, the share of science and engineering doctorates has starkly increased over the past decades, making up 77% in 2018 (National Science Foundation, 2019). Only a minority of PhD graduates are becoming tenured, full-time university professors (see Powell, 2015), and the majority (i.e., 56.8%) are employed outside academia. The latter group comprises employees in profit and nonprofit organizations, with estimations for the United States for 2015 reporting 36.2% of all STEM PhD holders working in for-profit businesses, 12.1.% in other businesses (e.g., nonprofit and unincorporated businesses), and 8.5% in government (see National Academies of Sciences, Engineering, & Medicine, 2018).

In reaction to complaints about PhD graduates lacking nonacademic competencies, such as commercial acumen, communication with nonexperts, and management of business-related operations (Edge & Munro, 2015), universities have started to introduce programs and initiatives for PhD students and holders to explore and practically experience nonacademic career options (e.g., Borrell-Damian et al., 2010). A number of doctoral programs even require their students to take courses, for instance, in “project management and other business activities” (Borrell-Damian et al., 2010, p. 496). In our study, we largely focused on these applied work competencies.

Although practical work experiences are suggested to influence the development of professional competencies, systematic longitudinal research in this area is pending (see Webster-Wright, 2009). Considering the efforts made to help early career scientists develop nonacademic competencies, it is surprising that knowledge about the formation of nonacademic competencies in this group is mostly anecdotal (Nowell et al., 2019). Therefore, research on the development of professional, nonacademic competence beliefs in academia and industry is highly warranted.

Development of and Dynamics Between Applied Work Demands and Competence Beliefs

By definition, private industry is based on applied tasks. Upon entering a position in industry (after having previously worked at a university), early career STEM scientists can be expected to be confronted with an increase in applied work demands (μ_SD; see Figure 1 for a depiction of the structural path model underlying the analyses). In addition, we suggest applied work demands to increase for early career scientists continuously working in academia (μ_SD). Interests in academic careers in STEM PhD students were found to decline over the course of the doctoral training (Roach & Sauermann, 2017). Almost 50% of engineering sciences PhD students were found to collaborate with industry partners (Mangematin, 2000), suggesting applied work requirements to be relevant already during doctoral training. Requirements may shift, leading more experienced PhD students and holders in academia to deal with nonscientific demands, for instance, applying for funding from external sponsors and clearly depicting financial needs.

Figure 1.
Path diagram of bivariate dual change score model for applied work demands and related competence beliefs. Note. For reasons of clarity, observed variables and error terms are omitted in the path diagram. D = demands, C = competence beliefs, S = slope, I = intercept, Δ = change in variable between measurement points. Paths with no coefficient are fixed to one.

Hypothesis 1: Demands for applied work increase for early career scientists in industry (after having previously worked in academia) and for those in academia.

Although we expect applied work demands to increase in academia and industry, we assume demands to be higher and thereby more salient and important upon entering a position in industry. Individuals can be assumed to increase their capability beliefs if they gain task-specific knowledge and develop work routines. Furthermore, Tierney and Farmer (2011) postulate that if an employee is assigned to a specific task, the assignment serves as a “source of efficacy cue,” showing the employee that they have what it takes to master the task (p. 280). We, therefore, assume applied work competence beliefs to increase in industrially employed scientists, as we expect them to be particularly confronted with applied work tasks (μ_SC).
Hypothesis 2: Competence beliefs concerning applied work increase for early career scientists in industry (after having previously worked in academia).

As suggested by the self-efficacy theory (Bandura, 1986), task engagement contributes to the development of ability-related self-views. The conceptualization of demands-abilities fit also postulates that specific abilities “can grow with use” (Edwards, 1996, p. 296). In their cross-sectional study, Schyns and von Collani (2002) found a positive association between job demands and occupational self-efficacy and suggest that people with higher task demands are more likely to experience task-specific mastery (as a result of involvement with difficult tasks). Empirical findings support this suggestion, revealing task engagement/training to increase work-specific capability beliefs (e.g., Schwoerer et al., 2005). Although we consider applied work demands in academia to not be as high as in industry, we expect them to grow over time. For early career scientists in academia, we assume higher demands—through involvement and engagement—lead to some learning experiences, thus exhibiting a positive influence on the development of competence beliefs (γ_DC).
Hypothesis 3: Demands for applied work accelerate change in related competence beliefs at later time points for early career scientists in academia.

In the context of organizational socialization, a substantial number of organizational newcomers experience so-called reality shocks (e.g., Van Maanen & Schein, 1979). Reality shocks can occur if employees have experiences upon entering a new position that starkly differ from their expectations. This confrontation with unknown and challenging work requirements can result in the feeling of lacking appropriate abilities (e.g., Jones, 1983). Accordingly, we assume that early career scientists who enter a new position in industry—after previously having worked in academia—are confronted with applied work requirements that are much more salient and central than ever before. They might feel challenged or even overwhelmed by these new requirements. We, therefore, expect high levels of applied work demands upon entering a new position in industry to lead to a slower increase in competence beliefs (γ_DC).
Hypothesis 4: Demands for applied work decelerate immediate change in related competence beliefs for early career scientists in industry (after previously having worked in academia).

Based on empirical studies on the JD-R model, Bakker and Demerouti (2017) conclude that employees possessing numerous job resources can handle their job demands better. Individuals who perceive themselves as not being competent enough to accomplish their tasks might be hesitant and unwilling to engage in these tasks (see Williams & Lillibridge, 1992). In contrast, employees who master job demands are more likely to look for and create new opportunities to actually make use of their competencies (Tims et al., 2012). As job crafting processes are known to result in greater congruence between personal capabilities and work conditions/tasks, we suggest higher applied work competence beliefs to lead to a faster increase in applied work demands in early career scientists in industry (γ_CD).
Hypothesis 5: Competence beliefs concerning applied work accelerate change in demands at later time points for early career scientists in industry (after previously having worked in academia).

Method

Procedure and Sample

The analyses are based on data from a larger project on early career STEM scientists (i.e., doctoral students and holders; for further details on the project, see Alisic & Wiese, 2020; Burk & Wiese, 2018). Students who started their doctoral training less than 1 year ago were excluded because we wanted participants to have a minimum of research-related work experience. Participants were excluded if they received their doctorate degree more than 10 years ago since we focused on early careers.

The longitudinal panel study (online questionnaire) comprised of eight measurement points. Participants were invited to respond to the next questionnaire 6 months after completion of the previous questionnaire and were reminded up to 5 times, resulting in a total study span of roughly 4 years. Regarding the variables relevant to this study, response rates for the total sample at the eight measurement points are presented in Table 1, starting with N_T1 = 3,005 (38.1% female). The increased number of participants from T4 to T5 resulted from the launch of a consecutive study phase. In the total sample, the average longitudinal effective dropout rate between consecutive measurement points was 9%. As an incentive, participants had the opportunity to take part in several raffles, winning up to €2,000 (approximately US$2,220).

Table 1.
Descriptive Statistics of Applied Work Demands and Related Competence Beliefs for the Academia Group, the Industry Group, and the Total Sample.

Measure Academia Industry Total Sample

n M SD n M SD N M SD ω [95% CI]

Demands_T1 1,195 2.22 1.09 — — — 3,005 2.40 1.22 .73 [.71, .75]

Demands_T2 912 2.24 1.10 415 3.71 1.15 2,294 2.48 1.25 .74 [.72, .76]

Demands_T3 672 2.28 1.07 263 3.74 1.06 1,795 2.59 1.26 .76 [.74, .78]

Demands_T4 441 2.36 1.09 222 3.85 1.02 1,262 2.71 1.27 .76 [.73, .78]

Demands_T5 506 2.33 1.13 176 3.88 0.98 1,576 2.88 1.29 .75 [.72, .77]

Demands_T6 429 2.39 1.13 94 3.86 1.09 1,383 3.07 1.35 .75 [.73, .77]

Demands_T7 381 2.49 1.17 78 4.02 0.99 1,253 3.22 1.31 .74 [.72, .77]

Demands_T8 390 2.52 1.15 35 4.10 0.95 1,340 3.26 1.31 .73 [.70, .75]

Comp. beliefs_T1 1,195 3.47 0.84 — — — 3,005 3.54 0.85 .62 [.60, .65]

Comp. beliefs_T2 912 3.39 0.90 414 3.45 0.85 2,294 3.47 0.86 .65 [.62, .68]

Comp. beliefs_T3 672 3.37 0.81 262 3.48 0.75 1,795 3.46 0.81 .61 [.57, .64]

Comp. beliefs_T4 441 3.40 0.85 222 3.68 0.71 1,262 3.53 0.82 .64 [.60, .68]

Comp. beliefs_T5 506 3.39 0.84 176 3.74 0.76 1,575 3.50 0.84 .63 [.59, .66]

Comp. beliefs_T6 427 3.31 0.94 94 3.84 0.79 1,381 3.51 0.90 .67 [.64, .70]

Comp. beliefs_T7 381 3.39 0.97 78 3.92 0.64 1,252 3.57 0.88 .66 [.61, .69]

Comp. beliefs_T8 390 3.39 0.96 35 4.06 0.71 1,339 3.58 0.88 .66 [.62, .69]

Note. Demands = applied work demands; comp. beliefs = applied work competence beliefs; ω = McDonald’s omega; CI 95% = lower and upper 2.5% of 95% confidence interval of ω; academia = subsample of participants continuously working in academia; industry = subsample of participants working in industry after previously having worked in academia (only post-transition measurement points were used for analyses).

Most participants were approached directly via email. We identified potential study participants by browsing through web pages of universities, research institutes, and business-related social networking services. Furthermore, we sent out invitations via online mailing lists of associations and interest groups for STEM professionals and scientists. We further contacted executives of many business enterprises (comprising all major companies in Germany), asking to distribute the invitation email among employees. Participants were further requested to share the invitation email with coworkers.

Participants indicated their primary occupational fields (i.e., university, private industry, nonuniversity scientific employment, nonscience government organization, self-employment, or other). The present paper focuses on two groups of the larger sample, that is, participants who primarily worked at a university (academia group) and participants who worked at a university when first taking part in the study and changed to a position in private industry during the study span (industry group). As we were interested in the period following the transition from academia to industry for the industry group, the measurement points succeeding the transition were part of the analyses. The remaining sample of participants who worked in other occupational fields, changed to fields of work other than from university to industry, or could not be assigned to any of these categories were not included in our analyses.

The academia group consisted of n = 1,205 early career scientists (36.8% female, M_age = 30.9, SD = 3.8), belonging to different STEM fields: 38.2% natural sciences, 41.0% engineering sciences, 11.2% computer science, 7.5% mathematics, and 2.1% other STEM fields. The ratio of PhD students to holders reversed from first (74.9 vs. 25.1) to last measurement point (29.3 vs. 70.7). Whereas of those who started as PhD students, 71.1% did not obtain their degree during the study span, and 28.9% graduated and continued to work at a university thereafter. Of those who were already PhD holders at T1, 3.6% were professors, 10.9% had a permanent and 77.2% a temporary position; of PhD holders at T8, 8.9% were professors, 18.8% had a permanent and 60.7% a temporary position (for the remaining sample, no information on employment situation was available). Participants who were already PhD holders at T1 received their degree on average 3.0 years ago (SD = 2.4).

The industry group comprised of n = 436 early career scientists (36.7% female). When starting to work in industry, participants were on average 32.3 years old (SD = 2.7); 30.0% were doctoral students (47.3% of these graduated during the study span), and the remaining sample already obtained their degree (on average 1.7 years, SD = 2.0, prior to the first measurement point in industry). As expected, the share of PhD students decreased over time, with 9.8% students and 90.2% doctorate holders at the last measurement point. Most participants in the industry group had a background in natural sciences (44.7%) and engineering sciences (34.6%), followed by computer science (12.4%) and mathematics (7.6%; the remaining sample reported other STEM fields).

Measures

The measures for applied work demands and competence beliefs each consisted of three indicators relating to specific competence domains relevant for applied work (i.e., economic focus, business organization, and customer focus). We asked participants to state the extent to which the specific competence domains were demanded in their job during the previous 2 months (demands: 1 = not needed at all, 6 = very much needed) and how competent they feel in comparison to their coworkers regarding the different domains (competence beliefs: 1 = to a much lesser extent, 6 = to a much higher extent). Participants were asked to compare themselves to their coworkers because social comparison with peers commonly takes place when evaluating one’s abilities (Marsh et al., 2017). The items were developed by experts in organizational environments and occupational demands. The experts regarded the three indicators as representing central facets of many applied work fields, thereby confirming former findings from the literature (e.g., Edge & Munro, 2015; Erez & Shneorson, 1980).

Because the observed variables tap different facets of applied work, we calculated the congeneric reliability coefficient ω, thus taking into consideration heterogenous factor loadings (see Table 1). To establish the validity of our measures of applied work, we drew upon the larger survey project in which participants longitudinally stated their job demands regarding a total of 18 competence domains. Based on data from T1 (i.e., measurement point with most participants), we ran exploratory factor analyses (EFAs) with oblique rotation for work demands and competence beliefs to determine the number of factors underlying the observed variables. Examining fit indices of models with increasing numbers of factors, a five-factor solution revealed a satisfactory fit in the EFAs for demands and competence beliefs (root-mean-square residual < .05, root-mean-square error of approximation [RMSEA] < .05; see Barendse et al., 2015). Based on factor loadings λ > .40, in both analyses, one of the five factors consisted of the three indicators of applied work (demands: economic focus λ = .68, business organization λ = .50, customer focus λ =.75; competence beliefs: economic focus λ = .79, business organization λ = .47, customer focus λ = .62). The indicators of applied work did not substantially cross-load on any other factor (λ > .40). We next investigated longitudinal invariance of the three-indicator measurement structure of the two factors and found it to be supported for applied work demands (strict factorial invariance model: χ² = 1,549.22, df = 248, comparative fit index (CFI) > .90, Tucker-Lewis index (TLI) > .90, RMSEA < .08; standardized factor loadings ranging between λ = .64 and .68 for economic focus, λ = .76 and .80 for business organization, λ = .65 and .72 for customer focus) and applied work competence beliefs (strict factorial invariance model: χ² = 1,287.47, df = 248, CFI > .90, TLI > .90, RMSEA < .08; standardized factor loadings ranging between λ = .66 and .72 for economic focus, λ = .78 and .80 for business organization, λ = .47 and .49 for customer focus).

Analytical Approach

Because we expected the growth patterns of interest to be rather complex, we considered latent change score modeling to be the most appropriate approach (Wang et al., 2016). To test model fit, we first conducted separate univariate dual change score (DCS) models based on latent true scores of scale means. For the academia group, we used data from T1 to T8. For the industry group, we used up to seven available measurement points after the transition from academia to industry took place. We next conducted bivariate DCS modeling (see Figure 1 for a depiction of the structural path model). The bivariate DCS model allows investigating growth trajectories (constant change components) of the two variables over time in addition to time-lagged influences of one variable on the same variable’s change (within-domain coupling) and on the other variable’s change (cross-domain coupling; Grimm et al., 2017). More precisely, the cross-domain coupling indicates the level of change in one variable as a function of the level in the other variable. We separately conducted bivariate DCS models for the academia and the industry group. To account for missing data, full information maximum likelihood estimation was used. Modeling was performed using Mplus 8 (Muthén & Muthén, 2017).

Results

Results of the univariate DCS analyses revealed sufficient fit of the models (CFI > .90, TLI > .90, RMSEA < .08). Next, we conducted bivariate DCS analyses for the two groups and found them to fit the observed data acceptably well (see Table 2 for models’ fit indices). Regarding the developmental trajectories, our hypotheses were supported: Demands increased over time for the academia as well as the industry group (H1; μ_SD in path model illustrated in Figure 1), and competence beliefs increased over time in the industry group (H2; μ_SC). In terms of cross-domain couplings, as expected, high demands accelerated later change in competence beliefs in the academia group (H3; γ_DC). Relating to the industry group, the hypothesis suggesting high demands to decelerate change in later competence beliefs was not supported (H4; γ_DC). Finally, in accordance with our assumptions, high competence beliefs accelerated the change rate in demands in the industry group (H5; γ_CD).

Table 2.
Results of Bivariate Dual Change Score Models of Applied Work Demands and Related Competence Beliefs for the Academia Group and the Industry Group.

Parameter Academia Industry

b SE b SE

μ_ID 2.24 .03 3.71 .06

μ_IC 3.46 .02 3.42 .04

σ²_ID 0.86 .04 0.91 .10

σ²_IC 0.46 .03 0.41 .05

μ_SD 1.38** .41 1.32* .56

μ_SC 0.39 .29 0.75* .34

σ²_SD 0.09 .05 0.39 .20

σ²_SC 0.02 .01 0.03 .03

β_D 0.18 .09 −0.95** .23

γ_DC 0.15* .07 −0.19 .19

β_C −0.22 .11 0.02 .19

γ_CD −0.51** .16 0.67* .29

σ_{ID, IC} 0.39 .03 0.31 .05

σ_{SD, SC} 0.04 .02 0.08 .09

σ_{ID, SD} 0.02 .04 0.39 .12

σ_{ID, SC} −0.04 .03 0.09 .09

σ_{IC, SD} 0.15 .05 −0.01 .07

σ_{IC, SC} 0.03 .04 0.00 .03

Note. b = unstandardized parameter estimate; SE = standard error; μ_ID = intercept demands; μ_IC = intercept competence beliefs; μ_SD = slope demands; μ_SC = slope competence beliefs; σ² = variance; β_D = influence of demands on change in demands; γ_DC = influence of demands on change in competence beliefs; β_C = influence of competence beliefs on change in competence beliefs; γ_CD = influence of competence beliefs on change in demands; σ = covariate between variables. Academia (n = 1,205), χ² = 386.66, df = 132, RMSEA = .04, CFI = .96, TLI = .96. Industry (n = 436), χ² = 157.68, df = 99, RMSEA = .04, CFI = .93, TLI = .94.

*p < .05. **p < .01.

Discussion

General Discussion

In this study, we investigated trajectories of and dynamics between demands for applied work and related competence beliefs. While theoretical conceptualizations and empirical research indicate different antecedents of perceived capability beliefs (e.g., Bandura, 1986; Gist & Mitchell, 1992; Williams & Lillibridge, 1992), the reciprocal relationship between applied work demands and related capability beliefs has not been previously investigated.

Regarding longitudinal change trajectories, the results empirically supported our expectations: Demands increased over time in the academia group (see μ_SD in Table 2) and the industry group and competence beliefs increased in the industry group (μ_SC). Contributing to the literature on work experiences of early career STEM scientists (e.g., Mangematin, 2000; Roach & Sauermann, 2017), the results support the increasing relevance of applied work demands, not just in industry but also during academic career phases. In addition, an explorative examination of the academia group revealed no significant change in competence beliefs over time (μ_SC). An explanation of this finding relates to applied work demands. Although demands increased in academia, the starting values were comparably low; thus, applied work tasks were not particularly salient or important. Accordingly, cues of performance (which contribute to the development of capability judgments) might have been limited (e.g., Tierney & Farmer, 2011; Tolli & Schmidt, 2008).

As expected, in academia, cross-domain couplings revealed demands to accelerate later change in competence beliefs (γ_DC). This finding supports theoretical considerations (Bandura, 1986; Edwards, 1996) suggesting task engagement to influence the development of task routines and mastery experiences, thereby contributing to the perception of one’s capabilities (e.g., Webster-Wright, 2009). The cross-domain effect of demands on change in competence beliefs, however, was rather small, which we attribute to the fact that demands only increased modestly in academic employment situations.

Contrary to our expectations, demands did not significantly decelerate change in competence beliefs in the industry group (γ_DC). However, the direction of the effect was negative, which is in line with our expectations relating to the potential effects of reality shock occurring after changing from academia to industry. Initial levels of applied work demands were comparably high in the industry group. As Gist and Mitchell (1992) pointed out, people likely analyze their capability levels in more detail when tasks change or become more salient. In an empirical study, individuals considering formidable tasks indicated lower self-efficacy, while those focusing on more manageable tasks reported higher self-efficacy (Cervone, 1989). Moreover, a study on creative self-efficacy revealed higher work requirements to be associated with lower self-efficacy (Tierney & Farmer, 2011). Thus, if task requirements are high, they might decrease a person’s sense of efficacy even though we did not find a significant negative effect.

Supporting our hypothesis, the results of the industry group revealed individuals feeling competent in applied tasks to report a faster increase in related demands (γ_CD). In line with the conceptualization of job crafting, the results indicate that individuals with high capability levels tailor their work to match their abilities (e.g., Tims et al., 2012). An explorative examination of the academia group showed high levels of competence beliefs to decelerate subsequent change in demands (γ_CD). As expected from a job crafting perspective, high competence beliefs should accelerate demands over time, particularly when initial levels of demands are comparably low and initial levels of competence beliefs are comparably high (as was the case in the academia group). The empirical result, however, contradicts this line of reasoning. A potential explanation for this effect refers to the subjective nature of the assessment of demands. Individuals with higher competence beliefs might not have perceived applied work demands to increase as strongly as individuals with lower capability beliefs.

With its longitudinal bivariate perspective, the study contributes to the literature on dynamics between personal and occupational factors (Woods et al., 2013). In JD-R theory, for instance, job demands and personal resources are postulated to exert influences on one another (Bakker & Demerouti, 2017). However, as indicated by the results, reciprocal processes between job demands and related competence beliefs can be suggested to be related to contextual factors (e.g., Dicke et al., 2018; Marsh et al., 2017), such as the occupational setting a person encounters (e.g., entering a new position, holding a position for a long time). We expect saliency and importance of one’s task experiences and capability perceptions to affect the extent and direction of reciprocal influences. Thus, when studying reciprocal relations, we encourage the processes underlying bivariate reciprocity to be considered as complex and dependent upon occupational settings.

Practical Implications

In their study on core self-evaluations, Wu and Griffin (2012) advise organizations to not only select employees already having a high level of positive self-evaluation but also to support the increase of positive self-evaluations by providing opportunities for affirmative job experiences (see also Axtell & Parker, 2003; Williams & Lillibridge, 1992). Regarding universities’ career development strategies for PhD students and holders, policymakers have stressed the importance of experiential engagement activities as opportunities for gaining nonacademic work competencies, exploring different career options, and making better informed career decisions (Borrell-Damian et al., 2010; Edge & Munro, 2015; National Academies of Sciences, Engineering, & Medicine, 2018). In this study, we found applied work demands to positively influence the development of related competence beliefs in the academia group, thereby supporting the effectiveness of opportunities for experiential learning in higher education. In addition to other initiatives and programs for early career scientists’ skill development (e.g., seminars, workshops, or internships; see Nowell et al., 2019), we propose training assessment centers as a way of providing employees with opportunities to familiarize themselves with applied tasks while subsequently receiving performance feedback (e.g., Tolli & Schmidt, 2008). For example, training assessment center sessions could include role-plays in which participants navigate difficult customer negotiation scenarios. This would help them to (a) build capability beliefs and (b) make better informed career decisions.

Strengths, Limitations, and Future Research Perspectives

Our study has several strengths. One particular strength is the longitudinal study design encompassing an overall time span of more than 4 years with eight time points. Thereby, we were able to investigate not only correlational associations between the variables but also temporal directionality. Another strength of our study relates to the sample sizes of more than 1,200 early career STEM scientists working in academia and more than 400 working in industry after previously having worked in academia. By their definition, the two samples exhibited group-specific characteristics. Because our research questions were specifically directed at early career scientists in these different employment situations, the sampling procedure corresponded to our research aim. The results of this study, therefore, bear relevance for academia as well as industry, both of which are typical employers for STEM PhDs (see Auriol et al., 2013). The measurement approach for domain-specific capability beliefs is another strength of this study. We operationalized competence beliefs not as a “one-size-fits-all trait” (Bandura, 2012, p. 17) but in relation to one specific activity domain, thereby increasing the predictiveness of the construct. Furthermore, from a methodological perspective, applying the bivariate DCS model allowed us to simultaneously investigate longitudinal change trajectories and cross-domain influences of demands on subsequent change in competence beliefs and vice versa.

There are also limitations to the current study. In applying frequently used cutoff criteria for scale reliability, concerns may arise for some of the measures of competence beliefs. However, the interpretation of reliability coefficients depends on various parameters such as the number of items and dimensionality of the constructs (Cortina, 1993). Given that our measures consisted of only 3 items and that the items represented the conceptual breadth of the constructs, reliability coefficients cannot necessarily be expected to be particularly high (e.g., van Griethuijsen et al., 2015). For congeneric measures in which factor loadings are heterogeneous, reliabilities greater than .6 are suggested to be desirable (Bagozzi & Youjae, 1988). We, therefore, consider our measures appropriate for our research purpose. Another limitation concerns the potential influence of common method bias that can lead to artificial covariance between the measures (Podsakoff et al., 2003). Given that a response gathered at a prior measurement point is likely to no longer be salient or retrievable months later, our longitudinal design addressed this concern by using temporal separation between predictor and criterion variables.

Clearly, our results cannot be generalized to other activity domains or samples. Capability beliefs for one domain (e.g., writing a peer-reviewed article) cannot be simply translated to another domain (e.g., writing a business plan; e.g., Williams & Lillibridge, 1992). Thus, for future studies, we encourage the investigation of the dynamics between demands and competence beliefs in other work domains. Furthermore, it is important to examine whether similar results can be found in other samples, such as early career scientists from the humanities or business administration.

In summary, by demonstrating longitudinal interdependencies between task demands and competency beliefs, our study combined socio-cognitive conceptualizations of self-beliefs with current approaches on proactivity in work psychology, that is, job crafting. Methodologically, the necessity of longitudinal designs in career research has been acknowledged for decades. However, there is a need for modeling approaches that incorporate the longitudinal interplay of concepts of interest. Future studies need to account for the complex change patterns that depend on individual adaptation to both changing demands and demand-changing activities.

Measure	Academia	Industry	Total Sample
Demands_T1	1,195	2.22	1.09	—	—	—	3,005	2.40	1.22	.73 [.71, .75]
Demands_T2	912	2.24	1.10	415	3.71	1.15	2,294	2.48	1.25	.74 [.72, .76]
Demands_T3	672	2.28	1.07	263	3.74	1.06	1,795	2.59	1.26	.76 [.74, .78]
Demands_T4	441	2.36	1.09	222	3.85	1.02	1,262	2.71	1.27	.76 [.73, .78]
Demands_T5	506	2.33	1.13	176	3.88	0.98	1,576	2.88	1.29	.75 [.72, .77]
Demands_T6	429	2.39	1.13	94	3.86	1.09	1,383	3.07	1.35	.75 [.73, .77]
Demands_T7	381	2.49	1.17	78	4.02	0.99	1,253	3.22	1.31	.74 [.72, .77]
Demands_T8	390	2.52	1.15	35	4.10	0.95	1,340	3.26	1.31	.73 [.70, .75]
Comp. beliefs_T1	1,195	3.47	0.84	—	—	—	3,005	3.54	0.85	.62 [.60, .65]
Comp. beliefs_T2	912	3.39	0.90	414	3.45	0.85	2,294	3.47	0.86	.65 [.62, .68]
Comp. beliefs_T3	672	3.37	0.81	262	3.48	0.75	1,795	3.46	0.81	.61 [.57, .64]
Comp. beliefs_T4	441	3.40	0.85	222	3.68	0.71	1,262	3.53	0.82	.64 [.60, .68]
Comp. beliefs_T5	506	3.39	0.84	176	3.74	0.76	1,575	3.50	0.84	.63 [.59, .66]
Comp. beliefs_T6	427	3.31	0.94	94	3.84	0.79	1,381	3.51	0.90	.67 [.64, .70]
Comp. beliefs_T7	381	3.39	0.97	78	3.92	0.64	1,252	3.57	0.88	.66 [.61, .69]
Comp. beliefs_T8	390	3.39	0.96	35	4.06	0.71	1,339	3.58	0.88	.66 [.62, .69]

Parameter	Academia	Industry
μ_ID	2.24**	.03	3.71**	.06
μ_IC	3.46**	.02	3.42**	.04
σ²_ID	0.86**	.04	0.91**	.10
σ²_IC	0.46**	.03	0.41**	.05
μ_SD	1.38**	.41	1.32*	.56
μ_SC	0.39	.29	0.75*	.34
σ²_SD	0.09	.05	0.39	.20
σ²_SC	0.02	.01	0.03	.03
β_D	0.18	.09	−0.95**	.23
γ_DC	0.15*	.07	−0.19	.19
β_C	−0.22	.11	0.02	.19
γ_CD	−0.51**	.16	0.67*	.29
σ_{ID, IC}	0.39**	.03	0.31**	.05
σ_{SD, SC}	0.04	.02	0.08	.09
σ_{ID, SD}	0.02	.04	0.39**	.12
σ_{ID, SC}	−0.04	.03	0.09	.09
σ_{IC, SD}	0.15**	.05	−0.01	.07
σ_{IC, SC}	0.03	.04	0.00	.03

Footnotes

Authors’ Note

The Federal Ministry of Education and Research of Germany did not influence the study design, the collection, analysis and interpretation of the data, the writing of the report, and the decision to submit the article for publication.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Federal Ministry of Education and Research of Germany (Bundesministerium für Bildung und Forschung Deutschland; grant number 16FWN009).

ORCID iD

André D. S. Lerche

References

Akkermans

Schaufeli

W. B.

Brenninkmeijer

Blonk

R. W. B.

(2013). The role of career competencies in the Job Demands-Resources model. Journal of Vocational Behavior, 83, 356–366. https://doi.org/10.1016/j.jvb.2013.06.011

Alisic

Wiese

B. S.

(2020). Keeping an insecure career under control: The longitudinal interplay of career insecurity, self-management, and self-efficacy. Journal of Vocational Behavior, 120, 103431. https://doi.org/10.1016/j.jvb.2020.103431

Auriol

Misu

Freeman

R. A.

(2013). Careers of doctorate holders: Analysis of labour market and mobility indicators (OECD Science, Technology and Industry Working Papers, No. 2013/04). OECD Publishing. https://doi.org/10.1787/18151965

Axtell

C. M.

Parker

S. K.

(2003). Promoting role breadth self-efficacy through involvement, work redesign and training. Human Relations, 56, 113–131. https://doi.org/10.1177/0018726703056001452

Bagozzi

R. P.

Youjae

(1988). On the evaluation of structural equation models. Journal of the Academy of Marketing Science, 16, 74–94. https://doi.org/10.1007/BF02723327

Bakker

A. B.

Demerouti

(2017). Job Demands-Resources theory: Taking stock and looking forward. Journal of Occupational Health Psychology, 22, 273–285. https://doi.org/10.1037/ocp000005627732008

Bandura

(1986). Social foundations of thought and action: A social cognitive theory. Prentice Hall.

Bandura

(2012). On the functional properties of perceived self-efficacy revisited. Journal of Management, 38, 9–44. https://doi.org/10.1177/0149206311410606

Barendse

M. T.

Oort

F. J.

Timmerman

M. E.

(2015). Using exploratory factor analysis to determine the dimensionality of discrete responses. Structural Equation Modeling: A Multidisciplinary Journal, 22, 87–101. https://doi.org/10.1080/10705511.2014.934850

10.

Blocker

C. P.

Flint

D. J.

Myers

M. B.

Slater

S. F.

(2011). Proactive customer orientation and its role for creating customer value in global markets. Journal of the Academy of Marketing Science, 39, 216–233. https://doi.org/10.1007/s11747-010-0202-9

11.

Borrell-Damian

Brown

Dearing

Font

Hagen

Metcalfe

Smith

(2010). Collaborative doctoral education: University-industry partnerships for enhancing knowledge exchange. Higher Education Policy, 23, 493–514. https://doi.org/10.1057/hep.2010.20

12.

Burk

C. L.

Wiese

B. S.

(2018). Professor or manager? A model of motivational orientations applied to preferred career paths. Journal of Research in Personality, 75, 113–132. https://doi.org/10.1016/j.jrp.2018.06.002

13.

Caprile

Palmén

Sanz

Dente

(2015). Encouraging STEM studies: Labour market situation and comparison of practices targeted at young people in different member states. http://www.europarl.europa.eu/RegData/etudes/STUD/2015/542199/IPOL_STU(2015)542199_EN.pdf

14.

Cervone

(1989). Effects of envisioning future activities on self-efficacy judgments and motivation: An availability heuristic interpretation. Cognitive Therapy and Research, 13, 247–261. https://doi.org/10.1007/Bf01173406

15.

Cortina

J. M.

(1993). What is coefficient alpha? An examination of theory and applications. Journal of Applied Psychology, 78, 98–104. https://doi.org/10.1037//0021-9010.78.1.98

16.

Dicke

Stebner

Linninger

Kunter

Leutner

(2018). A longitudinal study of teachers’ occupational well-being: Applying the Job Demands-Resources model. Journal of Occupational Health Psychology, 23, 262–277. https://doi.org/10.1037/ocp0000070

17.

Edge

Munro

(2015). Inside and outside the academy: Valuing and preparing PhDs for careers. The Conference Board of Canada.

18.

Edwards

J. R.

(1996). An examination of competing versions of the person-environment fit approach to stress. Academy of Management Journal, 39, 292–339. https://doi.org/10.2307/256782

19.

Erez

Shneorson

(1980). Personality types and motivational characteristics of academics versus professionals in industry in the same occupational discipline. Journal of Vocational Behavior, 17, 95–105. https://doi.org/10.1016/0001-8791(80)90019-6

20.

Frese

Garst

Fay

(2007). Making things happen: Reciprocal relationships between work characteristics and personal initiative in a four-wave longitudinal structural equation model. Journal of Applied Psychology, 92, 1084–1102. https://doi.org/10.1037/0021-9010.92.4.1084

21.

Gist

M. E.

Mitchell

T. R.

(1992). Self-efficacy: A theoretical analysis of its determinants and malleability. Academy of Management Review, 17, 183–211. https://doi.org/10.2307/258770

22.

Grimm

K. J.

Ram

Estabrook

(2017). Growth modeling: Structural equation and multilevel modeling approaches. Guilford Press.

23.

Holland

J. L.

(1997). Making vocational choices: A theory of vocational personalities and work environments (3rd ed.). Psychological Assessment Resources.

24.

Huang

(2016). Achievement goals and self-efficacy: A meta-analysis. Educational Research Review, 19, 119–137. https://doi.org/10.1016/j.edurev.2016.07.002

25.

Hughes

Galbraith

White

(2011). Perceived competence: A common core for self-efficacy and self-concept? Journal of Personality Assessment, 93, 278–289. https://doi.org/10.1080/00223891.2011.559390

26.

Jacobs

J. E.

Lanza

Osgood

D. W.

Eccles

J. S.

Wigfield

(2002). Changes in children’s self-competence and values: Gender and domain differences across grades one through twelve. Child Development, 73, 509–527. https://doi.org/10.1111/1467-8624.00421

27.

Jones

G. R.

(1983). Psychological orientation and the process of organizational socialization: An interactionist perspective. Academy of Management Review, 8, 464–474. https://doi.org/10.2307/257835

28.

Khosravi

(2016). Business process rearrangement and renaming: A new approach to process orientation and improvement. Business Process Management Journal, 22, 116–139. https://doi.org/10.1108/Bpmj-02-2015-0012

29.

Kohlbacher

(2010). The effects of process orientation: A literature review. Business Process Management Journal, 16, 135–152. https://doi.org/10.1108/14637151011017985

30.

Mangematin

(2000). PhD job market: Professional trajectories and incentives during the PhD. Research Policy, 29, 741–756. https://doi.org/10.1016/S0048-7333(99)00047-5

31.

Marsh

H. W.

Martin

A. J.

Yeung

A. S.

Craven

R. G.

(2017). Competence self-perceptions. In Elliot

A. J.

Dweck

C. S.

Yeager

D. S.

(Eds.), Handbook of competence and motivation: Theory and application (2nd ed., pp. 85–115). Guilford Press.

32.

Marsh

H. W.

Shavelson

(1985). Self-concept: Its multifaceted, hierarchical structure. Educational Psychologist, 20, 107–123. https://doi.org/10.1207/s15326985ep2003_1

33.

McNatt

D. B.

Judge

T. A.

(2008). Self-efficacy intervention, job attitudes, and turnover: A field experiment with employees in role transition. Human Relations, 61, 783–810. https://doi.org/10.1177/0018726708092404

34.

Muthén

L. K.

Muthén

B. O.

(2017). Mplus user’s guide (8th ed.). https://www.statmodel.com/download/usersguide/MplusUserGuideVer_8.pdf

35.

National Academies of Sciences, Engineering, & Medicine. (2018). Graduate STEM education for the 21st century. The National Academies Press.

36.

National Science Foundation. (2019). Doctorate recipients from U.S. universities: 2018 (Report No. NSF 20-301). https://ncses.nsf.gov/pubs/nsf20301/report

37.

National Science and Technology Council. (2018). Charting a course for success: America’s strategy for STEM education. The Committee on STEM Education of the National Science & Technology Council.

38.

Nowell

Ovie

Kenny

Hayden

K. A.

Jacobsen

(2019). Professional learning and development initiatives for postdoctoral scholars. Studies in Graduate and Postdoctoral Education, 11, 35–55. https://doi.org/10.1108/SGPE-03-2019-0032

39.

Parker

P. D.

Marsh

H. W.

Ciarrochi

Marshall

Abduljabbar

A. S.

(2014). Juxtaposing math self-efficacy and self-concept as predictors of long-term achievement outcomes. Educational Psychology, 34, 29–48. https://doi.org/10.1080/01443410.2013.797339

40.

Perkmann

Tartari

McKelvey

Autio

Broström

D’Este

Fini

Geuna

Grimaldi

Hughes

Krabel

Kitson

Llerena

Lissoni

Salter

Sobrero

(2013). Academic engagement and commercialisation: A review of the literature on university-industry relations. Research Policy, 42, 423–442. https://doi.org/10.1016/j.respol.2012.09.007

41.

Podsakoff

P. M.

MacKenzie

S. B.

Lee

J. Y.

Podsakoff

N. P.

(2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879–903. https://doi.org/10.1037/0021-9101.88.5.879

42.

Powell

(2015). The future of the postdoc. Nature, 520, 144–147. https://doi.org/10.1038/520144a

43.

Roach

Sauermann

(2010). A taste for science? PhD scientists’ academic orientation and self-selection into research careers in industry. Research Policy, 39, 422–434. https://doi.org/10.1016/j.respol.2010.01.004

44.

Roach

Sauermann

(2017). The declining interest in an academic career. PLoS ONE, 12, 1–23. https://doi.org/10.1371/journal.pone.0184130

45.

Sadri

Robertson

I. T.

(1993). Self-efficacy and work-related behavior: A review and meta-analysis. Applied Psychology: An International Review, 42, 139–152. https://doi.org/10.1111/j.1464-0597.1993.tb00728.x

46.

Schwoerer

C. E.

May

D. R.

Hollensbe

E. C.

Mencl

(2005). General and specific self-efficacy in the context of a training intervention to enhance performance expectancy. Human Resource Development Quarterly, 16, 111–129. https://doi.org/10.1002/hrdq.1126

47.

Schyns

von Collani

(2002). A new occupational self-efficacy scale and its relation to personality constructs and organizational variables. European Journal of Work and Organizational Psychology, 11, 219–241. https://doi.org/10.1080/13594320244000148

48.

Spurk

Abele

A. E.

(2014). Synchronous and time-lagged effects between occupational self-efficacy and objective and subjective career success: Findings from a four-wave and 9-year longitudinal study. Journal of Vocational Behavior, 84, 119–132. https://doi.org/10.1016/j.jvb.2013.12.002

49.

Sutin

A. R.

Costa

P. T.

(2010). Reciprocal influences of personality and job characteristics across middle adulthood. Journal of Personality, 78, 257–288. https://doi.org/10.1111/j.1467-6494.2009.00615.x

50.

Tierney

Farmer

S. M.

(2011). Creative self-efficacy development and creative performance over time. Journal of Applied Psychology, 96, 277–293. https://doi.org/10.1037/a0020952

51.

Tims

Bakker

A. B.

Derks

(2012). Development and validation of the job crafting scale. Journal of Vocational Behavior, 80, 173–186. https://doi.org/10.1016/j.jvb.2011.05.009

52.

Tolli

A. P.

Schmidt

A. M.

(2008). The role of feedback, casual attributions, and self-efficacy in goal revision. Journal of Applied Psychology, 93, 692–701. https://doi.org/10.1037/0021-9010.93.3.692

53.

Tschannen-Moran

Hoy

A. W.

(2007). The differential antecedents of self-efficacy beliefs of novice and experienced teachers. Teaching and Teacher Education, 23, 944–956. https://doi.org/10.1016/j.tate.2006.05.003

54.

van Griethuijsen

R. A. L. F.

van Eijck

M. W.

Haste

den Brok

P. J.

Skinner

N. C.

Mansour

Gencer

A. S.

BouJaoude

(2015). Global patterns in students’ views of science and interest in science. Research in Science Education, 45, 581–603. https://doi.org/10.1007/s11165-014-9438-6

55.

Van Maanen

Schein

E. H.

(1979). Toward a theory of organizational socialization. In Staw

B. M.

(Ed.), Research in organizational behavior (pp. 209–264). JAI Press.

56.

Wang

Zhou

Zhang

(2016). Dynamic modeling. Annual Review of Organizational Psychology and Organizational Behavior, 3, 241–266. https://doi.org/10.1146/annurev-orgpsych-041015-062553

57.

Webster-Wright

(2009). Reframing professional development through understanding authentic professional learning. Review of Educational Research, 79, 702–739. https://doi.org/10.3102/0034654308330970

58.

Williams

K. J.

Lillibridge

J. R.

(1992). Perceived self-competence and organizational behavior. In Kelley

(Ed.), Issues, theory, and research in industrial/organizational psychology (Vol. 82, pp. 155–184). Elsevier.

59.

Woods

S. A.

Lievens

De Fruyt

Wille

(2013). Personality across working life: The longitudinal and reciprocal influences of personality on work. Journal of Organizational Behavior, 34, 7–25. https://doi.org/10.1002/job.1863

60.

C. H.

Griffin

M. A.

(2012). Longitudinal relationships between core self-evaluations and job satisfaction. Journal of Applied Psychology, 97, 331–342. https://doi.org/10.1037/a0025673

Measure	Academia			Industry			Total Sample
Measure	n	M	SD	n	M	SD	N	M	SD	ω [95% CI]
Demands_T1	1,195	2.22	1.09	—	—	—	3,005	2.40	1.22	.73 [.71, .75]
Demands_T2	912	2.24	1.10	415	3.71	1.15	2,294	2.48	1.25	.74 [.72, .76]
Demands_T3	672	2.28	1.07	263	3.74	1.06	1,795	2.59	1.26	.76 [.74, .78]
Demands_T4	441	2.36	1.09	222	3.85	1.02	1,262	2.71	1.27	.76 [.73, .78]
Demands_T5	506	2.33	1.13	176	3.88	0.98	1,576	2.88	1.29	.75 [.72, .77]
Demands_T6	429	2.39	1.13	94	3.86	1.09	1,383	3.07	1.35	.75 [.73, .77]
Demands_T7	381	2.49	1.17	78	4.02	0.99	1,253	3.22	1.31	.74 [.72, .77]
Demands_T8	390	2.52	1.15	35	4.10	0.95	1,340	3.26	1.31	.73 [.70, .75]
Comp. beliefs_T1	1,195	3.47	0.84	—	—	—	3,005	3.54	0.85	.62 [.60, .65]
Comp. beliefs_T2	912	3.39	0.90	414	3.45	0.85	2,294	3.47	0.86	.65 [.62, .68]
Comp. beliefs_T3	672	3.37	0.81	262	3.48	0.75	1,795	3.46	0.81	.61 [.57, .64]
Comp. beliefs_T4	441	3.40	0.85	222	3.68	0.71	1,262	3.53	0.82	.64 [.60, .68]
Comp. beliefs_T5	506	3.39	0.84	176	3.74	0.76	1,575	3.50	0.84	.63 [.59, .66]
Comp. beliefs_T6	427	3.31	0.94	94	3.84	0.79	1,381	3.51	0.90	.67 [.64, .70]
Comp. beliefs_T7	381	3.39	0.97	78	3.92	0.64	1,252	3.57	0.88	.66 [.61, .69]
Comp. beliefs_T8	390	3.39	0.96	35	4.06	0.71	1,339	3.58	0.88	.66 [.62, .69]