Age-related effects of job characteristics on burnout and work engagement

Transcription

1 Occupational Medicine 2016;66: Advance Access publication 24 January 2016 doi: /occmed/kqv172 Age-related effects of job characteristics on burnout and work engagement R. Ramos, G. Jenny and G. Bauer Division of Public and Organizational Health, Epidemiology, Biostatistics and Prevention Institute, University of Zürich, Hirschgraben 84, CH-8001 Zürich, Switzerland. Correspondence to: R. Ramos, Division of Public and Organizational Health, Epidemiology, Biostatistics and Prevention Institute, University of Zürich, Hirschgraben 84, CH-8001 Zürich, Switzerland. Tel: ; Background In light of an ageing and age-diverse workforce, it is imperative to understand how psychosocial aspects of work might influence health throughout working life. Recently, there has been an implicit call to differentiate job characteristics beyond the two factors of job demands and job resources. As needs, abilities and motivation fluctuate with age, different job characteristics might yield differential benefits. Additionally, markers beyond chronological age should be considered. Aims Methods Results To explore systematically interactions between different job characteristics, age and age covariates (i.e. job tenure and position type) and their relationship with work-related health outcomes. An online survey of workers in Switzerland, Austria and Germany, recruited through a panel data service provider. We excluded participants working fewer than 30 hours a week, trainees, self-employed people and senior managers. We assessed seven areas of psychosocial risks at work, burnout, work engagement and demographics. Of the 6000 workers contacted, 1916 responded (31%). After applying exclusion criteria, we analysed data from 1417 respondents. We found that age barely had a moderating effect between psychosocial factors and health outcomes, but its three-way interaction with age covariates had more explanatory potential. Young workers with high job tenure showed particular vulnerability to job demands and the lack of certain job resources. Older workers with managerial positions were more resilient. Conclusions Age and its covariates, such as job tenure and position type, should be considered in developing agesensitive occupational health models. Key words Age; age covariates; burnout; job characteristics; job demands; job resources; lifespan; Management Standards; psychosocial factors at work; work engagement. Introduction The demographic tapestry of the job market is changing. People are living and working longer, while the number of new workers stagnates. The ageing workforce poses challenges that call for an interdisciplinary approach. In occupational health psychology, age has turned from a mere statistical control variable into a central focus of study [1]. It is well established that psychosocial aspects of work are crucial determinants of health [2]. The Job Demands-Resources model (JD-R) [3] has dominated the field in recent years, particularly due to its broader assessment of predictors and outcomes than previous models. The model proposes two parallel processes: the energetic process, in which job demands deplete employees resources, eventually leading to burnout and health impairment, and the motivational process, in which job resources foster work engagement and well-being while buffering the deleterious effects of job demands [4]. Researchers have investigated the incidence of workrelated health outcomes in different age groups, finding inconsistent results. A meta-analysis yielded no conclusive picture on the relationship between age and burnout [5], whereas a large study with data from 10 countries has shown a small yet statistically significant positive association between age and work engagement [6]. The potential moderating effects of age in the relationship between job characteristics and health outcomes have been less studied. However, the limited evidence suggests that The Author Published by Oxford University Press on behalf of the Society of Occupational Medicine. All rights reserved. For Permissions, please journals.permissions@oup.com

2 RAMOS ET AL.: Age, job characteristics and mental health 231 the JD-R model remains invariant across different age groups [7]. While the contributions of the JD-R model are indisputable, we argue that the dualism between job demands and job resources (often conceived as latent variables representing an amalgamation of job characteristics) may not be adequate when it comes to considering the significance of a worker s age. It tacitly assumes that all job demands are equally wearing and all job resources are equally rewarding across a person s working life, thus ignoring age-related shifts in physical and cognitive abilities, motivation and decision-making processes. Lifespan development theories provide a better framework in this regard. For instance, the selective optimization with compensation theory [8] posits that with age people become increasingly selective in allocating their resources, attempting to optimize their use in order to achieve desired outcomes. People also develop compensation strategies to cope with the decay of certain resources. In a similar vein, the Socioemotional and Selectivity Theory [9] suggests that as time passes people seek to maximize positive emotional and social experiences and minimize their social and emotional risks. Some scholars have proved the need for differentiation between different job demands and job resources in age-related questions. For instance, Zaniboni and colleagues showed that task variety led to less burnout and lower turnover intentions for younger workers, whereas skill variety led to lower turnover intentions for older workers [10]. They concluded that task variety might be seen as a way of gaining experience for younger workers, whereas skill variety might enable older workers to showcase the experience they have gained throughout their career. Moreover, whereas younger workers are characterized by their fluid intelligence (required when performing novel tasks), older workers excel through their crystallized intelligence (needed when tasks require a certain level of expertise). In an extensive theory-driven review, Truxillo et al. [11] hypothesized how the availability of different job resources might have differential effects on the satisfaction, engagement and performance of younger and older workers. However, one major caveat of their proposition is that it does not consider the age-related effects of the lack of such resources. This study attempts to address this issue. Finally, some see chronological age as a mere umbrella variable, which subsumes myriad covarying factors. Such factors can represent subjective individual experiences (e.g. work centrality, future time perspective) or objective work-related variables (e.g. job tenure, position type), all of which correlate with age [12 14]. Based on the preliminary differential findings [10] and theory-based inferences [11], this study aimed to investigate the differential age effects of an array of job demands and resources on work-related health outcomes. Specifically, we wanted to observe the interaction between age, job characteristics and objective age covariates, namely job tenure and position type, and their relationship with burnout and work engagement. Methods We conducted an online survey through a Europe-wide panel data service provider, inviting a random sample of 6000 workers in Germany, Austria and Switzerland to participate. To simplify the analysis and reduce the impact of confounding through other variables, we excluded participants working fewer than 30 hours a week, trainees, self-employed people and senior managers. The sample therefore consisted mostly of full-time employees with or without management positions. Respondents represented a broad range of occupations and industries. The data service provider sent a link, based on our inclusion criteria, to potential respondents who were informed of the overall goal of the study and invited to complete the anonymous questionnaire online. This was then returned to us directly. Following University of Zurich ethical guidelines, the study did not require independent ethical review. We assessed psychosocial factors in the workplace with the German version of the UK Health & Safety Executive (HSE) Management Standards Tool [2]. This well-established battery of 35 items has proven to be suitable in different occupational contexts and countries [15]. It comprises seven key dimensions of work design that are crucial for health-related outcomes [16]: quantitative demands (e.g. I have unrealistic time pressures ), job control ( I have a choice in deciding what I do at work ), colleague support ( I get the help and support I need from colleagues ), supervisor support ( I can rely on my line manager to help me with a work problem ), role clarity ( I am clear what my duties and responsibilities are ), negative social relations (NSR) ( There is friction or anger between colleagues ) and transparency in times of change ( change ) ( I have sufficient opportunities to question managers about change at work ). Participants responded on a 5-point Likert scale from 1 (never) to 5 (always). We measured burnout with a 4-item scale from the Copenhagen Psychosocial Questionnaire (COPSOQ II) [17]. This scale gauges the frequency of physical and emotional exhaustion in the last four weeks (e.g. How often have you felt worn out? ). Participants answered on a 5-point Likert scale from never/almost never to always. We assessed work engagement with the Utrecht Work Engagement Scale [6]. This 9-item scale has three subdimensions (vigour, dedication and absorption), captured with three items each on a 7-point Likert scale from never to always. (e.g. I am enthusiastic about my job ). We collected information about participants age, gender, job tenure and position type (i.e. whether they held a management position or not). Hereafter, we use the term workers regardless of position and the terms employees and managers when distinction is necessary.

3 232 OCCUPATIONAL MEDICINE We first ran correlation analyses for all study variables. We then conducted two stepwise hierarchical regression analyses in order to assess the main effects of job characteristics on burnout and engagement. We included demographic and work-related variables in the first step of the regression, job demands and resources in the second and third steps accordingly. Finally, we ran 70 (2 7 5) moderation analyses using the Process macro for SPSS [18]. We regressed the two outcome variables on all seven key areas with five different interaction terms (see Table 3). We plotted graphs for significant interactions following the ±1 standard deviation (SD) convention and reported regions of significance (when needed), following Johnson-Neyman s technique [19]. Results Of the 6000 workers contacted, 1916 completed the survey (31%). After applying exclusion criteria, we analysed responses from 1417 participants, aged between 20 and 65 (mean = 42; SD = 11) and of whom 61% were men and 29% held a management position. The average job tenure was 7.83 years, but the distribution was positively skewed, with the longest tenure being 40 years (this was considered in the analyses, employing natural logarithms [20]). Table 1 shows correlations between demands, resources and outcome measures. Men were more likely to hold a management position and women reported higher levels of burnout. However, we did not elaborate on gender further as exploratory analyses showed no interaction effects. Position type and tenure correlated positively with age but did not correlate with each other. Table 2 shows that job demands (10%) and resources (12%), respectively, explained similar amounts of variance in burnout. Along the motivational axis, job resources accounted for 18% of the variance in work engagement after controlling for demographics (4%) and demands (11%). The main findings of this study are summarized in Table 3, and significant moderation effects are depicted in Figures 1 and 2. Two-way interactions (Figure 1) show that older and high-tenure workers (both represented as dashed lines) displayed steeper slopes than young and low-tenure workers, respectively, indicating a stronger sensitivity towards the lack or availability of specific job demands and resources. We found that older workers reported significantly higher levels of work engagement than younger workers when role clarity was high (Figure 1a; z > 0.19; P < 0.05). High-tenure workers reported higher levels of burnout than low-tenure workers when reporting high quantitative demands (Figure 1b; z > 1.33; P < 0.05). Conversely, they reported less burnout than low-tenure workers when quantitative demands were low (z < 1.81; P < 0.05). Although the interaction with colleague support (Figure 1c) was significant and symmetry was observable as scores deviated from the mean, no statistically significant transition point was found. Finally, high-tenure workers reported lower levels of work engagement when quantitative demands were high (Figure 1d; z > 1.11; P < 0.05). Three-way interactions with work-related age covariates cast further light on the relationships. Figure 2a and 2b reveals that the levels of burnout and engagement for older workers were contingent on NSR but not their tenure (as illustrated by the perfectly overlapping lines for older workers with low and high tenure, respectively). They also Table 1. Correlation table for study variables Variables Age 2. Gender 0.07* 3. Work position 0.13*** 0.13*** 4. Job tenure 0.45*** Quantitative demands * 0.00 (0.84) 6. Negative social relations * *** (0.83) 7. Job control 0.07* 0.09* 0.13*** 0.08* 0.28*** 0.34*** (0.84) 8. Role clarity 0.11*** *** 0.13*** 0.21*** 0.43*** 0.32*** (0.84) 9. Colleague support 0.09* * 0.27*** 0.56*** 0.33*** 0.42*** (0.83) 10. Managerial support *** *** 0.51*** 0.39*** 0.41*** 0.57*** (0.91) 11. Transparency in change *** *** 0.45*** 0.42*** 0.52*** 0.53*** 0.73 (0.62) 12. Burnout 0.12*** 0.14*** 0.07* *** 0.36*** 0.24*** 0.19*** 0.29*** 0.30*** 0.26*** (0.89) 13. Work engagement 0.10*** *** *** 0.34*** 0.34*** 0.41*** 0.37*** 0.49*** 0.43*** 0.31*** (0.96) Cronbach s alphas for scales are shown in parentheses along the diagonal. *P < 0.05; ***P <

4 RAMOS ET AL.: Age, job characteristics and mental health 233 Table 2. Stepwise hierarchical regression analysis for main effects Steps and predictor variables Burnout Work engagement β 95% CI β 95% CI Step 1: demographics Adjusted R 2 = 0.03** Adjusted R 2 = 0.04** Age 0.10** 0.16, ** 0.04, 0.14 Gender (male = 0, female = 1) 0.13** 0.08, * 0.01, 0.10 Management position (no = 0, yes = 1) 0.05* 0.10, ** 0.05, 0.14 Job tenure , , 0.01 Steps 2 and 3: job demands Adjusted R 2 = 0.10** Adjusted R 2 = 0.11** Quantitative demands 0.30** 0.25, ** 0.14, 0.04 Negative social relations 0.10** 0.04, * 0.12, 0.00 Steps 2 and 3: job resources Adjusted R 2 = 0.12** Adjusted R 2 = 0.18** Job control , ** 0.09, 0.19 Role clarity , ** 0.14, 0.24 Colleague support 0.11** 0.17, * 0.00, 0.12 Supervisor support 0.11** 0.18, ** 0.24, 0.38 Transparency in change , , 0.06 Total R Steps 2 and 3 were interchanged based on the outcome variable (i.e. job demands were entered second and job resources third when work engagement was the outcome variable, and job resources were entered second and job demands third when burnout was the outcome variable). *P < 0.05; **P < N = Table 3. Summary of significant two-way and three-way interactions between age, job tenure, position and job characteristics Burnout β 95% CI Job tenure quantitative demands 0.058* 0.011, Job tenure colleague support 0.050* 0.101, Job tenure age negative social relations 0.080* 0.142, Work engagement β 95% CI Age role clarity 0.052* 0.004, Job tenure quantitative demands 0.054* 0.106, Age position quantitative 0.056* 0.107, demands Age position job control 0.062* 0.012, Age position supervisor support 0.057* 0.010, Age job tenure negative social 0.074* 0.011, relations Age job tenure supervisor 0.072* 0.130, support Age job tenure transparency in change 0.060* 0.117, Standardized coefficients are shown. All coefficients were significant at the 0.05 significance level (*P < 0.05). We tested five interaction terms for each of the seven job characteristics: age job characteristic, job tenure job characteristic, position job characteristic, age job tenure job characteristic, and age position job characteristic. This was performed for both burnout and work engagement as outcome variables, adding up to a total of 70 regression analyses. reported significantly less burnout than low-tenure young workers (z < 0.09; P < 0.05) as NSR decreases. Finally, high-tenure young workers reported significant higher levels of burnout than older (z > 0.47; P < 0.05) and low-tenure young workers (z > 0.17; P < 0.05). The last two groups did not differ in burnout levels when NSR was high. The pattern is very similar for work engagement, where older workers scored highest when NSR was low and high-tenure young workers reported the lowest work engagement when NSR was high. In both analyses, lowtenure young workers showed the flattest slope, suggesting that levels of burnout and engagement did not vary as strongly as for the other groups. The pattern for hightenure young workers applied also to supervisor support (Figure 2c) and change (Figure 2d): whereas participants did not differ significantly when both resources were available, high-tenure young workers showed significantly less engagement when supervisor support and transparency regarding change were missing. The three-way interactions between job characteristics, age and position revealed that, in terms of quantitative demands (Figure 2e), all groups displayed similar decreasing slopes in engagement as demands increased. Young employees were the exception: their levels of engagement were unaffected by demands, scoring significantly less engagement than older employees (z < 0.03; P < 0.05), young managers (z < 1.14; P < 0.05) and older managers (z < 1.89; P < 0.05) when quantitative demands were low. In terms of job control (Figure 2f), we observed inverted interactions between young workers and older workers, respectively. Having job control was predictive of high work engagement for all older

5 234 OCCUPATIONAL MEDICINE supervisor support was predictive of high engagement but the effect was weaker for young employees. Figure 1. Plot graphs of significant two-way interactions. workers. However, the lack of job control was particularly disengaging for older employees (compared with managers; z < 0.02; P < 0.05). Younger workers were more disengaged when they lacked job control, yet the availability of the latter was particularly beneficial for managers (compared with employees; z > 0.30; P < 0.05). Lack of supervisor support (Figure 2g) was predictive of lower engagement for all groups except for older managers, who were least affected. By the same token, high Discussion In this study sample, we found that age alone barely moderated the relationship between job characteristics and work-related health outcomes but its interplay with age covariates, namely job tenure and position type, could have more explanatory potential. With regard to specific job characteristics, high-tenure workers (regardless of age) displayed a steeper drop in engagement (Figure 1d) and rise in burnout (Figure 1b) than low-tenure workers when demands were high. NSRs were straining (Figure 2a) and disengaging (Figure 2b) for older workers regardless of job tenure. However, we found the strongest effects of NSR for high-tenure young workers. Older managers were the most resilient to low job control but the work engagement of young employees benefited least from the availability of such resource. This supports the idea that older (experienced) workers value job control more than younger workers who are still developing and may need less autonomy [11]. Our data suggest that this might particularly apply to younger workers with nonmanagerial positions. We also found that lack of support was particularly disengaging for young high-tenure workers (Figure 2c), while older managers were more resilient to it (Figure 2g). Age moderated the relationship between role clarity and work engagement with higher levels being more beneficial for older workers (Figure 1a). Role clarity may be less important in younger workers since a quota of ambiguity may be inevitable and necessary in the process of gaining new knowledge and experience (c.f. with specialization [11]). Finally, transparency regarding changes at work was related to higher levels of work engagement for all workers. However, its absence was particularly disengaging for high-tenure young workers (Figure 2d). The study has several limitations. Self-reported crosssectional data do not allow conclusions about causality. We have no information about non-responders but the demographics of our sample were similar to official demographic figures from the working populations in Germany, Austria and Switzerland. Common method variance cannot be excluded, although moderated relationships are less vulnerable to this bias [21]. Studies with more objective measures of health, time lag and longitudinal designs would substantially strengthen our inferences. The interactions we found are very small, accounting for up to 1% of the variance. However, small effect sizes are common in moderation models that explore psychological phenomena, where the explained variance rarely exceeds 3% [22]. Given that age and job tenure were highly correlated, caution is needed when attributing differences to one or the other. The exploratory nature of the study relies on multiple testing, which can lead to type I errors. Thus, the results should be regarded as a guide to future in-depth research. Also,

6 RAMOS ET AL.: Age, job characteristics and mental health 235 Figure 2. Plot graphs for three-way interactions. while the observation of isolated psychosocial factors is legitimate, it should only be an initial step towards a better understanding of the interaction between different demands and resources across workers lifespans. Potential buffering or boosting effects for specific demands and resources (e.g. quantitative demands support) should be explored as they relate to age and its covariates. Finally, this study considered objective covariates of chronological age. Future research could investigate the interactive effects of psychological covariates such as work centrality and future time perspective [12 14]. Despite these limitations, the results have given new insights that deserve further attention. It would be interesting to observe mechanisms at the individual level, such as the psychological contract. It is known for instance that high-tenure workers are more likely to experience contract breach [23] which could explain the deleterious effects on engagement and burnout when quantitative demands are high (i.e. excessive demands might be perceived as contract violation). Flat regression lines, such as the ones observed for low-tenure workers and young employees for the effect of demands on work engagement, call for studies investigating curvilinear effects. It could be that young workers and those who are relatively new to an organization are particularly willing to learn and develop; hence very low demands might be seen as unchallenging [24]. However, very high demands in an incipient career might be perceived as overwhelming. The results have also shown a consistent pattern of vulnerability for high-tenure young workers (see Figure 2a 2d) and resilience for older managers (see Figure 2f and 2g). While the latter can be intuitively explained through their greater experience and, presumably, their better coping strategies, the vulnerability of young high-tenure workers deserves particular consideration. Decision-making processes in the work domain may offer a promising path for future research as it is known that older adults are more likely than younger

7 236 OCCUPATIONAL MEDICINE adults to disinvest in situations with adverse outcomes, hence avoiding the sunk cost bias [25]. In light of these findings, it would be interesting to understand the motivation of some young workers, a generation otherwise known for its high turn over rates [26], to remain in an organization where adverse job characteristics prevail. Finally, although the HSE management standards tool has proven to be a reliable measure of psychosocial risks at work [15], future studies should cover a broader range of job characteristics that might be more relevant to the lifespan approach (e.g. job complexity, job significance). This study highlights the need for differentiation when considering age in occupational health models. Rather than age group comparisons of the demand-resource dichotomy [7] analysis of each job characteristic individually may be more informative, as suggested by recent theory-based propositions [11]. Moreover, rather than focusing only on the presence of certain job characteristics, this study also explored their absence. It also shows the potential relevance of age-covarying factors (rather than chronological age as a stand-alone predictor) in explaining the purported impact of job characteristics on burnout and work engagement. This study offers a new perspective to understanding the link between job characteristics and health-related outcomes as we attempt to prepare organizations and individuals for an ageing workforce and increasing crossgenerational collaboration. Key points Age-sensitive occupational health models should be guided less by the categorization of job characteristics into job demands and job resources and more by the alignment of such job characteristics with age-related shifts in motivation, values and abilities. The relationship between job characteristics and work-related health outcomes (i.e. burnout and work engagement) seems to be more influenced by the interaction between chronological age and its work-related covariates (i.e. job tenure and position type) than by chronological age per se. Future research should consider not only objective (i.e. job tenure and position type) but also psychological age covariates (e.g. work centrality and future time perspective) in order better to assess their role in reducing or augmenting the effect of psychosocial factors at work on work-related health outcomes. Funding This project was funded by Pfizer AG (Switzerland). Conflicts of interest None declared. References 1. Truxillo DM, Fraccaroli F. Research themes on age and work: introduction to the Special Issue. Eur J Work Organ Psychol 2013;22: Cousins R, MacKay CJ, Clarke SD, Kelly C, Kelly PJ, McCaig RH. Management Standards work-related stress in the UK: practical development. Work Stress 2004;18: Demerouti E, Bakker AB, Nachreiner F, Schaufeli WB. The job demands-resources model of burnout. J Appl Psychol 2001;86: Schaufeli WB, Bakker AB. Job demands, job resources, and their relationship with burnout and engagement: a multisample study. J Organ Behav 2004;25: Brewer EW, Shapard L. Employee Burnout: a meta-analysis of the relationship between age or years of experience. Hum Resour Dev Rev 2004;3: Schaufeli WB, Bakker AB, Salanova M. The measurement of work engagement with a short questionnaire: a crossnational study. Educ Psychol Meas 2006;66: Korunka C, Kubicek B, Schaufeli WB, Hoonakker P. Work engagement and burnout: testing the robustness of the Job Demands-Resources model. J Posit Psychol 2009;4: Baltes PB, Baltes MM. Psychological perspectives on successful aging: the model of selective optimization with compensation. In: Baltes PM and Baltes MM, eds. Successful Aging: Perspectives from the Behavioral Sciences. New York, NY: Cambridge University Press, 1990; Carstensen LL. Social and emotional patterns in adulthood: support for socioemotional selectivity theory. Psychol Aging 1992;7: Zaniboni S, Truxillo DM, Fraccaroli F. Differential effects of task variety and skill variety on burnout and turnover intentions for older and younger workers. Eur J Work Organ Psychol 2013;22: Truxillo DM, Cadiz DM, Rineer JR, Zaniboni S, Fraccaroli F. A lifespan perspective on job design: fitting the job and the worker to promote job satisfaction, engagement, and performance. Organ Psychol Rev 2012;2: Bal PM, de Lange AH, Zacher H, Van der Heijden BIJM. A lifespan perspective on psychological contracts and their relations with organizational commitment. Eur J Work Organ Psychol 2013;22: Kanfer R, Beier ME, Ackerman PL. Goals and motivation related to work in later adulthood: An organizing framework. Eur J Work Organ Psychol 2013;22: Zacher H, Frese M. Remaining time and opportunities at work: relationships between age, work characteristics, and occupational future time perspective. Psychol Aging 2009;24: Brookes K, Limbert C, Deacy C, O Reilly A, Scott S, Thirlaway K. Work-related stress and the HSE Management Standards. Occup Med (Lond) 2013;63: Kerr R, McHugh M, McCrory M. HSE Management Standards and stress-related work outcomes. Occup Med (Lond) 2009;59: Pejtersen JH, Kristensen TS, Borg V, Bjorner JB. The second version of the Copenhagen Psychosocial Questionnaire. Scand J Public Health 2010;38:8 24.

8 RAMOS ET AL.: Age, job characteristics and mental health Hayes A. Introduction to Mediation, Moderation, and Conditional Process Analysis. New York, NY: Guilford, Aiken LS, West SG. Multiple Regression: Testing and Interpreting Interactions. Thousand Oaks, CA: Sage, Field A. Discovering Statistics Using SPSS. London: Sage, Siemsen E, Roth A, Oliveira P. Common method bias in regression models with linear, quadratic, and interaction effects. Organ Res Methods 2010;13: Fairchild AJ. A comparison of frameworks for the joint analysis of mediation and moderation effects. Unpublished doctoral dissertation, Arizona State University, Robinson SL, Morrison EW. The development of psychological contract breach and violation: a longitudinal study. J Organ Behav 2000;21: Jonge JD, Dollard MF, Dormann C, Le Blanc PM, Houtman ILD. The demand-control model: specific demands, specific control, and well-defined groups. Int J Stress Manag 2000;7: Strough J, Mehta CM, McFall JP, Schuller KL. Are older adults less subject to the sunk-cost fallacy than younger adults? Psychol Sci 2008;19: Twenge JM. A review of the empirical evidence on generational differences in work attitudes. J Bus Psychol 2010;25: