138 Journal of Occupational and Organizational Psychology (2014), 87, 138–157 © 2013 The British Psychological Society www.wileyonlinelibrary.com Daily transactional and transformational leadership and daily employee engagement Kimberley Breevaart1*, Arnold Bakker1, Jørn Hetland2, Evangelia Demerouti3, Olav K. Olsen4 and Roar Espevik2,4 1 Department of Work and Organizational Psychology, Erasmus University Rotterdam, The Netherlands 2 Department of Psychosocial Science, University of Bergen, Norway 3 Department of Industrial Engineering and Innovation Sciences Human Performance Group, Eindhoven University of Technology, The Netherlands 4 Department of Leadership Development, Royal Norwegian Naval Academy, Bergen, Norway This diary study adds to the leadership literature by examining the daily influence of transformational leadership, contingent reward, and active management-by-exception (MBE active) on followers’ daily work engagement. We compare the unique contribution of these leadership behaviours and focus on the work environment to examine how these leadership behaviours influence followers’ daily work engagement. While travelling by sail ship, 61 naval cadets filled out a diary questionnaire for 34 days. Multilevel regression analyses revealed that, after controlling for followers’ work engagement the previous day, cadets were more engaged on days that their leader showed more transformational leadership and provided contingent reward. MBE active was unrelated to followers’ work engagement. As predicted, transformational leadership and contingent reward contributed to a more favourable work environment (more autonomy and support), while MBE active resulted in a less favourable work environment (less autonomy) for the cadets. This study highlights the importance of daily leadership for followers’ daily work engagement. Practitioner points  Leaders’ daily behaviour influences followers’ daily work engagement.  Leaders’ daily behaviour shapes the daily work environment. A key proposition of Bass’ (1985) theory of transformational and transactional leadership is that transformational leadership explains unique variance in outcome variables after controlling for the influence of transactional leadership. Transactional leaders ensure that expectations are met, which is the foundation on which transformational leaders build to motivate their followers to perform beyond expectations. The few studies that examined the effect of transformational and transactional leadership simultaneously mainly used cross-sectional or longitudinal survey designs with long time intervals (Judge & Piccolo, 2004). However, these studies ignore the dynamic, short-term effects of leaders’ *Correspondence should be addressed to Kimberley Breevaart, Woudestein, T12-46, PO Box 1738, 3000 DR Rotterdam, The Netherlands (email: breevaart@fsw.eur.nl). DOI:10.1111/joop.12041 Daily leadership and follower engagement 139 behaviour, which may fluctuate from day-to-day and accordingly, differently impact follower outcomes depending on the day. Such a within-person approach to leadership has several advantages. For example, it allows us to examine leadership as a proximal predictor of follower outcomes. Furthermore, this approach reduces recall bias, focuses on leadership in its natural context, and brings us closer to the leadership process, because data are collected close to the event (i.e., at the end of each working day). Finally, adopting a microlevel perspective to leadership provides a more comprehensive understanding of leadership, because it tells us something about the extent to which traitlike leadership processes are similar to statelike leadership processes. The present study contributes to the literature in a number of important ways. First, we examine the unique contribution of transformational and transactional leadership to followers’ work engagement simultaneously. Previous research has shown that transformational leaders contribute to their followers’ work engagement (Tims, Bakker, & Xanthopoulou, 2011; Zhu, Avolio, & Walumbwa, 2009), but have ignored the role of transactional leadership in followers’ work engagement. We argue that some components of transactional leadership may be effective in stimulating followers’ work engagement, although less effective compared with transformational leadership. Second, we examine how these leadership behaviours influence followers’ work engagement by focusing particularly on the resourcefulness of the work environment. Studying the underlying process may provide answers as to why some leadership behaviours are more effective than others and advances our understanding of transformational and transactional leadership. Third, we follow a within-person approach to examine the abovementioned relationship on a daily basis, and fourth, we examine the abovementioned relationships in a unique context, that is, a sample of naval cadets who received leadership training during their 34-day stay on a sail ship. Theoretical background According to Bass (1985, 1999), transactional leaders motivate their followers to fulfil their leaders’ expectations, while transformational leaders motivate their followers to perform beyond what is expected of them. In other words, although transactional leaders can be effective (e.g., promote follower job performance), transformational leaders are even more effective (e.g., promote follower job performance beyond transactional leaders). Bass argued that every leader uses both transactional and transformational leadership to some extent, but the most effective leaders use transformational leadership more frequently than transactional leadership. Transactional leadership consists of multiple components that differ in their effectiveness. For example, some components lead to more committed, loyal, and satisfied followers than others (Bass & Riggio, 2006). First and most effective is contingent reward. Contingent reward means that followers receive incentives after they accomplish their tasks to stimulate followers’ task motivation. Contingent reward is transactional when these incentives are material (e.g., bonus), but can also be transformational when the incentive is psychological in nature (e.g., praise). More ineffective compared with contingent reward is management-by-exception (MBE). MBE active is about the anticipation of mistakes and the enforcement of rules that may prevent mistakes from happening. In contrast, MBE passive refers to confronting followers with their mistakes and expressing disapproval about the mistakes that have been made. As MBE passive is most likely to occur when leaders have a large span of control (Bass & Riggio, 2006) and is often considered to be passive-avoidant leadership behaviour, we focused on MBE active. 140 Kimberley Breevaart et al. Transformational leadership is characterized by the four I’s: Idealized influence, inspirational motivation, individual consideration, and intellectual stimulation. Idealized influence means that followers identify with their leaders and respect and trust them. Inspirational motivation refers to creating and communicating an appealing vision of the future and to the leaders’ own optimism about this future. Next, individual consideration means that leaders are mentors and acknowledge that every employee has his/her own needs and abilities. Finally, intellectual stimulation refers to challenging followers to rethink some of their ideas and to take a different perspective on the problems they face in their work. In their meta-analysis, Judge and Piccolo (2004) examined the unique contribution of transformational and transactional leadership in predicting different outcomes. The results indicated that transformational leadership and contingent reward were important predictors of several outcome variables (i.e., satisfaction with the leader, motivation, leader job performance, and leader effectiveness). MBE active had a positive, but very small impact on the outcome variables. State work engagement State work engagement is a transient, positive, fulfilling, and work-related state of mind that fluctuates within individuals over a short period of time (Sonnentag, Dormann, & Demerouti, 2010), and is characterized by vigour, dedication, and absorption (Breevaart, Bakker, Demerouti, & Hetland, 2012; Schaufeli & Bakker, 2004). Vigour refers to high levels of energy and mental resilience, dedication means being enthusiastic about work and inspired by the work tasks, and absorption refers to being fully concentrated on work and feeling like time flies when working. This dynamic approach allows us to examine how leaders influence followers’ work engagement in their natural work context. What happens on the days that leaders stimulate their followers’ work engagement? Furthermore, a dynamic day-to-day approach may reflect leaders’ behaviours more accurately, because followers only have to think back over several hours when they rate their leaders’ behaviours instead of having to think back over several weeks or months (cf. Ohly, Sonnentag, Niessen, & Zapf, 2010). Only a handful of studies examined the influence of leadership behaviour on followers’ work engagement. For example, Zhu et al. (2009) found that transformational leadership predicted followers’ work engagement, especially for individuals with positive characteristics (e.g., active learning). In an online experiment, Kovjanic, Schuh, and Jonas (2013) found that individuals who imagined their leader to be transformational by reading a vignette were more engaged because it fulfilled their need for relatedness and need for competence. We are only aware of one study that examined the relationship between leadership behaviour and employee work engagement on a daily basis (Tims et al., 2011). Tims and her colleagues found that followers were more engaged on days that their leader showed more transformational leadership behaviour, because followers were more optimistic on these days. The present study contributes to the literature beyond the aforementioned studies, because we focus on the unique contribution of transformational and transactional leadership to followers’ work engagement on a daily basis. Tims et al. (2011) argue that transactional leaders lack the ‘motivational power and inspirational appeal’ (p. 122) that is needed to stimulate followers’ work engagement. However, we argue that, in line with Bass’ (1985, 1999) theory, followers will be more engaged on days when their leader shows more transactional leadership, but to a lesser extent than daily transformational Daily leadership and follower engagement 141 leadership. Furthermore, while Tims and colleagues focused on personal resources as the underlying mechanism to explain the relationship between leadership behaviour and follower work engagement, we focus on job resources. Leadership process A question that often remains unanswered in leadership research is ‘How do leaders influence follower outcomes?’. This is an important question, because it contributes to the understanding of the processes underlying the influence of leadership behaviour and hereby advances leadership theory. Yukl (2010) therefore called for more concentrated efforts to understand mediators that link leadership behaviours to follower outcomes. We argue that leaders influence their followers’ work engagement through their influence on the work environment. First, leaders may affect the way in which followers perceive their work environment. According to Smircich and Morgan (1982), leaders define and shape their followers’ work environment. Leaders are an important part of followers’ immediate social environment, on which followers rely when making judgements about their work environment (Salancik & Pfeffer, 1978). In line with this reasoning, research has shown that transformational leaders provide meaning for the work followers perform (Arnold, Turner, Barling, Kelloway, & McKee, 2007; Nielsen, Yarker, Brenner, Randall, & Borg, 2008). Second, leaders may encourage followers to actively increase their own resources, for example, by stimulating followers to think on their own and make their own decisions. Tims and Bakker (2010) argue that control over work is an important condition for employees to actively change their work environment to their benefit. Finally and most importantly, we argue that specific transformational and transactional leadership behaviours influence the actual availability of followers’ job resources. We will discuss these behaviours more in depth hereafter. Although several studies examined job resources to explain how leaders influence follower outcomes (Nielsen & Daniels, 2012; Nielsen, Randall, Yarker, & Brenner, 2008; Piccolo & Colquitt, 2006; Purvanova, Bono, & Dzieweczynski, 2006), these studies focused exclusively on transformational leadership and did not examine leadership on a daily basis. Research has shown that job resources have motivating potential, leading to higher work engagement (for meta-analyses see Crawford, LePine, & Rich, 2010; Halbesleben, 2010). On a daily basis, this means that employees are more engaged on days when they have more job resources (Xanthopoulou, Bakker, Demerouti, & Schaufeli, 2009; Xanthopoulou, Bakker, Heuven, Demerouti, & Schaufeli, 2008). To our knowledge, only one (survey) study examined whether leaders influence their followers’ work engagement because of their impact on the work environment. In their study among volunteer firefighters, Tuckey, Bakker, and Dollard (2012) showed that empowering leaders influence followers’ work engagement because they positively influence their followers’ cognitive job demands and job resources. However, it is still unclear whether job resources can explain how transformational and transactional leaders influence their followers’ (general or state) work engagement. Daily leadership behaviour and state work engagement Engaged employees perform their work because they enjoy it and are pulled towards it – they are intrinsically motivated (Schaufeli & Bakker, 2010). Transformational leaders contribute to employees’ intrinsic motivation, because they provide a meaningful rationale for their followers’ work (Avolio & Yammarino, 2002). That is, transformational 142 Kimberley Breevaart et al. leaders communicate an appealing vision of the future and show confidence in their followers’ ability to contribute to the realization of this vision (Seibert, Wang, & Courtright, 2011). Besides, transformational leaders stimulate followers to prefer the interests of the group over their self-interests (Avolio & Yammarino, 2002). On the sail ship, this means, for example, that leaders emphasize how conducting daily deck services and maintenance work contributes to the shared objectives and mission. Furthermore, this means that leaders stimulate the cadets to help and learn from other cadets during their ‘off-duty’ period, in order to achieve the common goals. It also means that leaders stimulate cadets to perform to the best of their abilities and delegate tasks that match cadets’ seamanship skills. This makes it likely that followers feel energetic, are dedicated to, and immersed in their work. Although contingent reward lacks the inspirational appeal of transformational leadership, it does not lack motivational power. Leaders who use contingent reward set clear goals and communicate what followers can expect when they reach these goals, which motivates followers (Bass & Avolio, 1995). This is in line with the results of Judge and Piccolo’s (2004) meta-analysis, which showed that contingent reward contributes to followers’ work motivation. Leaders who use contingent reward acknowledge that the cadets performed well in sailing the ship and developing their seamanship skills. This may provide meaning to the work, which is likely to increase the cadets’ work engagement on that day. Furthermore, research has shown that constructive feedback is an important predictor of work engagement (Halbesleben, 2010). Leaders who use contingent reward do lack inspirational appeal, and we therefore expect that transformational leaders influence their followers’ work engagement after controlling for the influence of contingent reward. Bass and Avolio (1994) claim that MBE active is neither effective nor ineffective. It therefore seems unlikely that leaders who use MBE active are able to influence their followers’ work engagement. Clearly, leaders who use MBE active lack both inspirational appeal and motivational power. However, Judge and Piccolo (2004) showed that MBE active was positively related to followers’ work motivation, although not as strongly as transformational leadership and contingent reward. We therefore expect leaders who use contingent reward are able to influence their followers’ work engagement after controlling for leaders’ use of MBE active. These arguments lead to the following hypotheses: Hypothesis 1a: Daily transformational leadership is positively related to followers’ daily work engagement after controlling for daily transactional leadership (i.e., contingent reward and MBE active). Hypothesis 1b: Daily contingent reward is positively related to followers’ daily work engagement after controlling for daily transformational leadership and daily MBE active. Daily leadership and autonomy We argue that transformational leadership behaviour is positively related to followers’ daily autonomy, because on days that leaders show more transformational leadership, followers are allowed to approach their problems from different perspectives, even if these perspectives are different from their leader’s ideas. For example, when the cadets have difficulties planning an anchoring operation, they can decide to distance themselves Daily leadership and follower engagement 143 from the problem for a while, to ask another cadet for help or to spend some time trying different solutions. This makes it likely that followers will experience more autonomy on days that their leader uses more transformational leadership. Research indeed supports that transformational leadership is positively related to follower autonomy (Jung & Sosik, 2002; Piccolo & Colquitt, 2006). However, these studies did not employ a diary design and did not study whether autonomy can explain the relationship between transformational leadership and follower work engagement. Although leaders who use transformational leadership may stimulate followers’ autonomy more explicitly, we propose that leaders who use contingent reward also provide followers with more decision latitude to decide how and when to perform their tasks. Leaders who use contingent reward value the achievement of working goals. Autonomy contributes to the achievement of working goals, because it allows followers to perform their work in the most efficient way. Therefore, it seems likely that they provide the cadets with autonomy. For example, cadets can decide to switch the order of their deck duties when the weather suddenly turns and first perform those duties that can be done under the current weather conditions. Besides, leaders who use contingent reward communicate clear expectations. When these expectations are communicated at the start of the day, this may decrease the necessity to control what and how cadets perform their work during the day. Finally, Eisenberger, Rhoades, and Cameron (1999) argued and found that recipients of rewards that are contingent on performance are able to control their environment, because they can decide whether or not to accept the reward. This ability to control rewards provides recipients with autonomy. Finally, we argue that leaders who use MBE active influence followers’ autonomy in a negative way. It is likely that followers have less autonomy to perform their work when their behaviour is constantly being monitored. For example, when the cadets try new ways to perform their work, there is a possibility that these are less effective and this increases the likelihood that mistakes are being made. Leaders who use MBE active try to actively prevent mistakes from happening, and thus, it is likely that they do not provide followers’ with any latitude to decide when and how to perform their work. Based on these arguments and the aforementioned literature showing a positive relationship between autonomy and work engagement (Halbesleben, 2010; Xanthopoulou et al., 2009), we hypothesize: Hypothesis 2a: Daily transformational leadership is positively related to followers’ daily autonomy after controlling for daily transactional leadership (i.e., contingent reward and MBE active). Hypothesis 2b: Daily contingent reward is positively related to followers’ daily autonomy after controlling for daily transformational leadership and daily MBE active. Hypothesis 2c: Daily autonomy mediates the relationship between daily (1) transformational leadership, (2) contingent reward, and (3) MBE active on the one hand, and followers’ daily work engagement on the other hand. Daily leadership and social support We further argue that followers will also receive more social support from their leader on days their leaders use more transformational leadership. Transformational leaders pay 144 Kimberley Breevaart et al. attention to their followers and listen to them carefully. Hence, on days that leaders use more transformational leadership, leaders will pay more attention to the needs of each follower. For example, leaders will take some time to privately talk to the cadets who are home sick or mediate in an argument between two cadets. This makes it likely that transformational leaders contribute to followers’ daily social support. Research has shown that support provided by the transformational leader has positive implications for followers, because it protects followers from experiencing stress and burnout (for a review see Skakon, Nielsen, Borg, & Guzman, 2010). Furthermore, we argue that contingent reward positively influences followers’ work engagement, because followers receive more social support when their leader uses contingent reward. Leaders who use contingent reward pay attention to their cadets and praise them when they are, for example, able to take the sails up and down in ten minutes. That is, followers’ performance is recognized and appreciated by leaders, which are well-known types of social support (Carlson & Perrewe, 1999; Etzion, 1984). With regard to the supervisory support followers receive, again, we claim that MBE active is less effective compared with transformational leadership and contingent reward. We argued that recognition and appreciation are important indicators of social support and that leaders who use MBE active do not recognize or appreciate followers’ performance. They rather show what they do not appreciate, namely mistakes being made. Furthermore, social support implies that there is at least some form of exchange and mutuality (Baumeister & Leary, 1995). Whereas the contingent reward component of transactional leadership is based on mutuality (i.e., praise in exchange for high performance), followers of leaders who use MBE active are expected to perform well and not make any mistakes, but followers are not rewarded in any way when they perform well. Together with the aforementioned literature showing a positive relationship between social support and work engagement (Halbesleben, 2010; Xanthopoulou et al., 2009), these arguments lead to our final hypotheses (see Figure 1 for an overview of all hypotheses): .08* Daily transformational leadership .10* Daily autonomy .37*** .10** –.06* Daily work engagement Daily contingent reward .33*** .14*** Daily MBE-active –.03 Daily social support .08* .03 Figure 1. Final model of daily leadership and employee work engagement, standardized solution including direct and indirect pathways. Note. *p < .05; **p < .01; ***p < .001. Daily leadership and follower engagement 145 Hypothesis 3a: Daily transformational leadership is positively related to followers’ daily social support after controlling for daily transactional leadership (i.e., contingent reward and MBE active). Hypothesis 3b: Daily contingent reward is positively related to followers’ daily social support after controlling for daily transformational leadership and daily MBE active. Hypothesis 3c: Daily social support mediates the relationship between daily (1) transformational leadership, (2) contingent reward, and (3) MBE active on the one hand, and followers’ daily work engagement on the other hand. Methods Participants and procedure Sixty-one Norwegian naval cadets from a Military University College participated in our study. As part of their leadership training, they travelled from northern Europe to North America by sail ship. The cadets received a booklet with diary questionnaires for the 40 days of their stay on the sail ship. During 6 days of their travel, the cadets went ashore and enjoyed their free days. We checked for missing data during this period and found out that almost all cadets did not fill out the diary during these days. As these missing values are not at random, we removed these 6 days from our analyses. We requested the cadets to fill out the questionnaire at 5 PM on each day. The cadets were part of one of eight teams, and most teams had multiple leaders on most of the days, so we asked them to rate the daily leadership behaviour of their nearest leader. The sample consisted of 46 male participants (75.4%) and 7 female participants (11.5%). Eight participants did not fill in their gender (13.1%). The mean age of the participants was 23.8 years (SD = 3.15). Measures We used daily diaries to measure our study variables. All day-level questionnaires were adapted versions of existing scales. We adapted the time frame of the scales and the number of questions so the questions could be answered on a daily basis (cf. Ohly et al., 2010). Moreover, the questionnaires were reduced in length when possible, because we asked the cadets to fill out the diary on all 34 days of their stay on the boat. The appendix provides an overview of all questions that were included in the study. Day-level transformational leadership behaviour Day-level transformational leadership behaviour was measured with five items from the multifactor leadership questionnaire-form 5X (MLQ 5X; Bass & Avolio, 1995). Participants could respond to the items on a 5-point scale, ranging from 1 (totally disagree) to 5 (totally agree). The average internal consistency of the scale across the days was .79. Day-level transactional leadership behaviour Day-level transactional leadership behaviour was measured with six items from the MLQ 5X; (Bass & Avolio, 1995); contingent reward and MBE active were assessed with three 146 Kimberley Breevaart et al. items each. Participants could respond to the items on a 5-point scale, ranging from 1 (totally disagree) to 5 (totally agree). The average internal consistency of the contingent reward scale was .61. On average, the internal consistency of the MBE active scale was .74. Day-level job resources Daily autonomy and daily social support were measured with three items each. Both scales were based on scales from Bakker, Demerouti, and Verbeke (2004). The items could be answered on a 5-point scale, ranging from 1 (not at all) to 5 (a very large degree). The average internal consistency of autonomy was .78, and the average internal consistency of social support was .76. Day-level work engagement Day-level work engagement was measured with the state version (Breevaart et al., 2012) of the 9-item Utrecht Work Engagement Scale (Schaufeli, Bakker, & Salanova, 2006). The statements could be answered on a 5-point scale (1 = totally disagree, 5 = totally agree). Work engagement showed an average internal consistency of .90. Strategy of analysis We used Mplus (Muthen & Muthen, 1998–2010) to take into account the nested structure of the data (i.e., days nested within persons and persons nested within teams). We have a 3-level model with days at the first level (Level 1; N = 2,440), persons at the second level (Level 2; N = 61), and teams at the third level (Level 3; N = 8). According to Maas and Hox (2005), a minimum of 30 cases at the highest, team level of analysis is needed for adequate power in multilevel modelling. Following this rule of thumb, we do not have a sufficient amount of cases at the highest, third level (N = 8) required for robust estimations. Therefore, we used the ‘TYPE = COMPLEX’ procedure in Mplus to calculate standard errors and chi-square values while taking into account the non-independence of observations due to clustering at the person level. We centred all our variables around the group (i.e., person) mean, because we are interested in how daily fluctuations from the baseline in the predictor variables are related to daily fluctuations from the baseline in the outcome variable (Ohly et al., 2010). Most variance of the predictor variables (67– 80%) was explained by the day-level. To test for the significance of the mediation effects, we used the parametric bootstrap method recommended by Preacher, Zyphur, and Zhang (2010) to create confidence intervals. This method does not make any assumptions about the distribution of the indirect effect, while the delta method confidence intervals in Mplus do not take into account the asymmetric nature of the indirect effect. We used the online interactive tool developed by Selig and Preacher (2008) that generates an R code to obtain confidence intervals for the indirect effect. As this tool does not allow specification of more than two paths, we adjusted the generated R code to test our sequential mediation hypothesis by adding an extra path from the second mediator to the outcome variable. Results Descriptive statistics Table 1 shows the intercorrelations, means, standard deviations, and internal consistencies of the study variables averaged over 34 days. Daily leadership and follower engagement 147 Hypotheses testing The first hypothesis suggests that daily transformational leadership is positively related to followers’ daily work engagement after controlling for daily contingent reward and MBE active (1a) and that daily contingent reward is positively related to followers’ daily work engagement after controlling for daily transformational leadership and daily MBE active (1b). We tested a model including paths from all leadership styles to followers’ work engagement to examine the unique contribution of each leadership style. The path from daily transformational leadership to daily work engagement was .15 (p < .001, 95% CI: 0.07, 0.22), the path from daily contingent reward to daily work engagement was .07 (p < .05, 95% CI: 0.01, 0.13), and the path from daily MBE active to daily work engagement was .03 (n.s., 95% CI 0.01, 0.08) after controlling for followers’ work engagement the previous day (.28, p < .001, 95% CI: 0.19, 0.38). Furthermore, we tested a model including only the path from MBE active to followers’ work engagement. This model explained 0.3% of the variance in followers’ daily work engagement. Next, we added the path from contingent reward to followers’ work engagement, which increased the explained variance in followers’ daily work engagement with 2.1%. Finally, the path from transformational leadership to followers’ work engagement was included and explained an additional 2.2% in followers’ daily work engagement. Together with work engagement the previous day, leadership explained 12.1% of the variance in followers’ daily work engagement. This means that Hypotheses 1a and 1b were both supported. Job resources as mediators Hypothesis 2a states that daily transformational leadership is positively related to followers’ daily autonomy after controlling for daily transactional leadership (i.e., contingent reward and MBE active), and Hypothesis 2b states that daily contingent reward is positively related to followers’ daily autonomy after controlling for followers’ daily transformational leadership and daily MBE active. In a similar vein, Hypothesis 3a states that daily transformational leadership is positively related to followers’ daily social support after controlling for daily transactional leadership (i.e., contingent reward and MBE active), and Hypothesis 3b states that daily contingent reward is positively related to followers’ daily social support after controlling for followers’ daily transformational leadership and daily MBE active. In line with our hypotheses, transformational leadership was positively related to both autonomy (.10, p < .05, 95% CI: 0.03, 0.18) and social support (.33, p < .001, 95% CI: 0.26, 0.39). Table 1. Means, standard deviations, intercorrelations, and internal consistencies (Cronbach’s alphas on the diagonal) between the study variables, N = 61 persons, N = 2,440 days) 1. Transformational leadership 2. Contingent reward 3. MBE active 4. Autonomy 5. Social support 6. Work engagement *p < .01; **p < .001. M SD 3.42 3.43 2.49 2.78 3.56 3.05 0.26 0.28 0.44 0.40 0.29 0.38 1 2 3 4 5 6 (.79) .51** .05 .15** .39** .19** .14** .14** .30** .15** (.74) .04 .003 .05* (.78) .16** .43** (.76) .18** (.90) 148 Kimberley Breevaart et al. Contingent reward was also positively related to autonomy (.10, p < .01, 95% CI: 0.04, 0.16) and social support (.14, p < .001, 95% CI: 0.07, 0.20). MBE active was negatively related to autonomy ( .06, p < .05, 95% CI 0.11, 0.00), but unrelated to social support ( .03, n.s.). Furthermore, we tested a model including the paths from MBE active to autonomy and social support. This model explained no variance in social support and only 0.2% of the variance in daily autonomy. Next, we added the paths from contingent reward to autonomy and social support. This model explained 6.6% of the variance in daily social support and an additional 1.6% of the variance in daily autonomy. Finally, we added the paths from transformational leadership to autonomy and social support. Transformational leadership was able to explain an extra 1.1% in daily autonomy and 9.7% in daily social support. We continued testing our mediation hypotheses. Hypothesis 2c states that the relationship between daily (1) transformational leadership, (2) contingent reward, and (3) MBE active on the one hand and daily work engagement on the other hand is mediated by daily autonomy. Hypothesis 3c states that the relationship between daily (1) transformational leadership, (2) contingent reward, and (3) MBE active on the one hand and daily work engagement on the other hand is mediated by daily social support. We tested a model including the double mediation by autonomy and social support, controlling for previous day work engagement. Social support (.08, p < .05, 95% CI: 0.02, 0.14) and autonomy (.37, p < .001, 95% CI: 0.31, 0.42) were both positively related to work engagement. Table 2 shows the results of the mediation effects. Autonomy mediated the relationship between transformational leadership, contingent reward, and MBE active on the one hand and work engagement on the other hand. Social support mediated the relationship between transformational leadership and work engagement. Contrary to our expectations, the relationship between contingent reward and MBE active on the one hand and work engagement on the other hand was not mediated by social support. Next, we compared the fit of our hypothesized model to the fit of the partially mediated model including direct effects from transformational leadership and contingent reward to followers’ work engagement. Results showed a Table 2. Indirect pathways from leadership to work engagement Indirect effect x ? m ? y Transformational leadership ? Autonomy ? Work engagement Transformational leadership ? Social support ? Work engagement Contingent reward ? Autonomy ? Work engagement Contingent reward ? Social support ? Work engagement Management-by-exception ? Autonomy ? Work engagement Management-by-exception ? Social support ? Work engagement Unstandardized path coefficient Standard error .04 95% Confidence interval Lower Upper p 0.02 0.01 0.08 .02 .03 0.01 0.01 0.05 .01 .04 0.02 0.01 0.06 .01 .01 0.01 0.00 0.02 .06 .02 0.01 0.04 0.00 .04 .003 0.003 0.01 0.003 .20 Daily leadership and follower engagement 149 significant decrease in chi-square (Dv2 = 8.72 (2), p < .05), indicating that the alternative model including the direct effects fits better to the data. Although the relationship between contingent reward and work engagement is no longer significant (.03, n.s.) after including the mediators, there was a small, but significant direct effect from transformational leadership to work engagement (.08, p < .05, 95% CI: 0.02, 0.15). The final model (Figure 1) fitted well to the data (CFI = .99; RMSEA = .04; SRMR = .01) and explained 3.1% of the variance in daily autonomy, 16.7% of the variance in daily social support, and 25.7% of the variance in daily work engagement. Discussion The aim of our study was threefold. First, we wanted to examine the unique contribution of transformational and transactional leadership to followers’ work engagement. Second, we focused on two job resources (i.e., autonomy and social support) to explain how leaders influence their followers’ work engagement. Third, we examined these relationships on a day-to-day basis. Contributing to the uniqueness of this study is the specific context in which the proposed relationships were tested, allowing us to examine our study model on a daily basis. During their stay on the boat, the cadets were trained to sail the boat, practice seamanship skills, and use transformational and transactional leadership. This meant that the cadets switched leadership positions, which allowed us to examine the influence of deviations in leadership compared to the baseline (average transformational and transactional leadership over the 34 days) in a natural, but highly controlled work context. In general, the results were in line with our expectations. We will now discuss our findings in more detail. Daily leadership and work engagement In line with Bass’ (1985, 1999) theory, we found that after controlling for transactional leadership, transformational leadership contributes to followers’ work engagement on a daily basis. We showed that, despite their lack of inspirational appeal, leaders who use contingent reward are also able to influence their followers’ daily work engagement in a positive way after controlling for daily MBE active. MBE active was unrelated to followers’ work engagement, which is in line with Bass’ (1999) contention that MBE active is neither effective nor ineffective. Although all three types of leadership explained a small amount of variance in followers’ daily work engagement, contingent reward explained additional variance in followers work engagement after controlling for MBE active. In a similar vein, transformational leadership explained additional variance in followers’ work engagement after controlling for transactional leadership (i.e., contingent reward, MBE active). Job resources as mediators This study focused on job resources to examine how transformational and transactional leaders influence their followers’ daily work engagement. We showed that daily autonomy is a promising mechanism through which leaders influence their followers’ daily work engagement. Transformational leadership and contingent reward positively 150 Kimberley Breevaart et al. influenced followers’ daily autonomy, which consequently influenced followers’ work engagement. MBE active decreased followers’ daily autonomy and, consequently, reduced followers’ work engagement. Again, results were in line with Bass’ contention that transformational leadership adds to the influence of transactional leadership. Daily contingent reward explains additional variance in followers’ daily autonomy compared with daily MBE active. Furthermore, daily transformational leadership explains additional variance in followers’ daily autonomy after controlling for daily transactional leadership (contingent reward and MBE active). In line with results for autonomy, transformational leadership and contingent reward were both positively related to social support on a day-to-day basis, the former explaining additional variance over the latter, and MBE active was unrelated to social support. Surprisingly, the mediation of social support in the relationship between contingent reward and employee work engagement was only marginally significant (p = .06). This may be due to the combination of the relatively small direct relationships between contingent reward and social support, and between social support and work engagement. Theoretical implications This study contributes to the literature in a number of ways. First and most importantly, it is one of the first to examine the influence of daily fluctuations in leadership behaviour on followers’ work engagement. Diary studies are important because they bring us closer to the process through which leadership is related to followers’ work engagement and reduce the risk of recall bias (Ohly et al., 2010), because followers rate their leader’s behaviour much closer to when it happens. With diary studies, followers only have to think back over a few hours instead of weeks or months. Therefore, transformational and transactional leadership measured on a daily basis may be a more accurate reflection of the leadership behaviours shown by the leader compared with leadership behaviour measured at one point in time. Moreover, ratings of general leadership behaviour may be more a reflection of someone’s attitude instead of a true reflection of leadership behaviour (Sonnentag et al., 2010). The present study shows the importance of monitoring leadership behaviour more closely when studying its effects, because most of the variance in transformational and transactional leadership was explained at the day-level. This shows that the degree to which leaders vary in their leadership from day-to-day may differently influence employees depending on the day. For example, on some days, leaders may predominantly use transformational leadership, while on other days they combine transformational leadership with contingent reward or MBE active. Furthermore, this is one of the few studies that examined the influence of transformational leadership and different components of transactional leadership simultaneously and the first to examine the relationship between different forms of transactional leadership and work engagement. Tims et al. (2011) argued that transactional leaders are unable to influence followers’ work engagement, but we showed that some transactional leadership behaviours (i.e., contingent reward) are able to stimulate followers’ work engagement. In line with the augmentation effect, transformational leadership adds to the influence of transactional leadership. It has often been questioned whether contingent reward is different from transactional leadership behaviour (Avolio, Bass, & Jung, 1999). Although this study does not provide an answer to this question, it does suggest that it is worthwhile to study its effect on follower outcomes separately from transformational leadership and other transactional leadership behaviours (e.g., MBE active). Daily leadership and follower engagement 151 Finally, the present study responds to the call for more research on the underlying mechanisms of the relationship between leadership and work outcomes (Yukl, 2010) and, more specifically, for more research on the mediating role of mechanisms rooted in the job (Piccolo & Colquitt, 2006). We showed that daily autonomy and daily social support mediated the relationship between daily leadership and followers’ daily work engagement. Furthermore, we contribute to the leadership literature because we focus on the effect of transformational and transactional leadership on a positive affective, motivational employee outcome – work engagement – while most other studies focus on stress and burnout or organizational outcomes such as performance and organizational citizenship behaviour (for a review see Skakon et al., 2010). Practical implications The present study also has some important practical implications. First, we show the importance of day-to-day fluctuations in leadership behaviour. For example, on days that leaders actively monitor their followers’ behaviour for mistakes, they indirectly reduce followers’ work engagement on that day. Besides, when leaders have an ‘off-day’ and show less transformational leadership or contingent reward, they are unable to positively influence followers’ work environment and work engagement on that day. According to Bass and Riggio (2006), every leader uses both transformational and transactional leadership. Our results confirm this and show that, in terms of work engagement, leaders are most effective on days that they show transformational leadership most frequently, followed by contingent reward. Although daily MBE active does not directly undermine followers’ work engagement, it does so indirectly because it reduces followers’ daily autonomy. This implies that leaders should only use MBE active on days that it is extremely important for followers to do exactly as they are told, for example, in threatening or urgent situations. Because leaders may not always be aware of how their behaviour affects followers, it may prove useful to provide leaders with feedback about their behaviour. For example, the leaders on the sail ship were provided with feedback about their transformational and transactional leadership behaviours based on their followers’ ratings. Subsequently, this feedback was used to set up development plans, which were then implemented by the leaders. Leaders could also use this direct and positive effect of transformational leadership to their benefit. For example, it is especially important that leaders use transformational leadership when followers’ engagement is of high importance (e.g., when there is an important deadline). Research has shown that leaders can be trained in their transformational leadership (Barling, Weber, & Kelloway, 1996; Dvir, Eden, Avolio, & Shamir, 2002). Barling et al. (1996) designed a transformational leadership training consisting of five sessions. Managers in the training group were rated by their followers as more intellectually stimulating, charismatic, and individual considerate 5 months after the training sessions compared with 2 weeks before the training and compared with the no-training control group. However, in the light of our findings and Bass’ theory, it seems important to not only focus on transformational leadership but also train leaders how to use transactional leadership behaviours effectively. Limitations of the study and implications for future research Although the current study has clear strengths due to its research design, it is not without limitations. The use of self-reports may potentially increase the risk of common method 152 Kimberley Breevaart et al. variance (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). According to Conway and Lance (2010), self-reports may be appropriate and authors should always provide arguments for their decision to use self-reports. In our study, we wanted to examine how perceived leadership behaviour is related to followers’ job resources. According to the JD-R model, each job and even each individual have their own constellation of job demands and job resources. Therefore, followers are the best source of information regarding their own job resources. Furthermore, we examined how followers’ job resources are related to employees’ work engagement, which is a private experience. Besides, common method bias is rarely strong enough to bias results (Doty & Glick, 1998; Spector, 2006), which is likely to be true for the present study, because the relationships between the construct can be best qualified as ‘moderate’. Finally, we followed Podsakoff et al.’s (2003) remedy to reduce the impact of common method bias by ensuring participants’ anonymity. Another limitation is that for some scales, on some days, the internal consistency was quite low. This is an important issue, also for diary studies in general, that warrants further attention. In diary studies, the time frame of existing scales is often adapted. For example, state work engagement is measured by reframing the items used that measure general levels of work engagement in such a way that they refer to the day. However, it is possible that some items refer to experiences that cannot be answered every day, resulting in lower interitem correlations and, consequently, lower internal consistency for the scale on that day (Sonnentag et al., 2010). Although the internal consistencies for some scales were low on some days, on average, the internal consistencies in our study meet Nunnally’s (1967) internal consistency of .60 for early stages of research. Furthermore, unreliable measures attenuate relationships between predictors and outcomes in such a way that the relationships are underestimated. This may imply that the relationships found in the present study can be considered as conservative. However, future research on the psychometric properties of daily diary measures is necessary. Another limitation of our study may be that we did not include the MBE-passive component of transactional leadership. One of the main aims of our study was to show that it is important to differentiate between the different components of transactional leadership, because some transactional leadership behaviours may be positively related to follower outcomes, while others are not. We did not include MBE passive, because Bass and Riggio (2006) argue that leaders will use MBE passive when they have a large number of subordinates, because that makes it difficult to actively monitor mistakes. Considering the setting of our study, we expected that leaders would prefer the use of MBE active to the use of MBE passive. Furthermore, Bass and Riggio (2006) state that MBE active is required and effective in certain situations, such as when safety is important. Following this reasoning, it is more likely that leaders on a sailboat who are trained to lead use MBE active instead of MBE passive. As the only difference between MBE active and MBE passive is the moment of intervening, we expect results to be similar for MBE active and MBE passive. Future research could test this assumption by including transformational leadership and all categories of transactional leadership behaviour (i.e., contingent reward, MBE active, and MBE passive). The specific sample and context in which our proposed relationships were examined may restrict the generalizability of our findings. Although the findings were in line with theoretically derived hypotheses, the results need to be replicated in other samples of employees working under different conditions. Note, however, that because of its specificity (i.e., being trained on a sail ship for a longer period of time), the context of this study provided a very special opportunity to examine the proposed relationships in a Daily leadership and follower engagement 153 highly controlled (i.e., few influences from the outside), but at the same time dynamic (e.g., switching positions, learning new skills) work context. In this study, we only focused on the process of the leadership behaviour – follower work engagement relationship. However, it is also important to study possible contingencies of leadership behaviour. Under which circumstances are certain leadership behaviours more or less effective? For example, it is conceivable that leaders’ influence on followers’ work environment is reduced when followers craft their own resources (Tims, Bakker, & Derks, 2012; Wrzesniewski & Dutton, 2001). Another example is that MBE active may be more effective compared with contingent reward and transformational leadership in high-risk professions such as those of aircraft pilots or nuclear plant workers. Conclusion The present study contributes to the literature in several ways. To our knowledge, this is the first study to examine the impact of transformational leadership and different components of transactional leadership on followers’ work engagement simultaneously and on a daily basis. We showed that both transformational leadership and contingent reward are positively related to followers’ work engagement, the former explaining additional variance in followers’ work engagement over the latter. Moreover, we examined how leaders’ daily leadership behaviour is related to followers’ daily work engagement. It appeared that transformational leaders and leaders who use contingent reward contribute to a favourable work environment (i.e., higher autonomy and social support), while MBE active contributes to a less favourable work environment (i.e., lower autonomy). To conclude, the present study shows that different types of daily leadership behaviours have an important direct and indirect influence on followers’ work engagement. References Arnold, K. A., Turner, N., Barling, J., Kelloway, E. K., & McKee, M. C. (2007). Transformational leadership and psychological well-being: The mediating role of meaningful work. Journal of Occupational Health Psychology, 12, 193–203. doi:10.1037/1076-8998.12.3.193 Avolio, B. J., Bass, B. M., & Jung, D. I. (1999). Re-examining the components of transformational and transactional leadership using the Multifactor Leadership Questionnaire. Journal of Occupational and Organizational Psychology, 7, 441–462. doi:10.1348/096317999166789 Avolio, B. J., & Yammarino, F. J. (2002). Transformational and charismatic leadership: The road ahead. Oxford, UK: Elsevier Science. Bakker, A. B., Demerouti, E., & Verbeke, W. (2004). Using the job demands–resources model to predict burnout and performance. Human Resource Management, 43, 83–104. doi:10.1002/hrm Barling, J., Weber, T., & Kelloway, K. E. (1996). Effects of transformational leadership training on attitudinal and financial outcomes: A field experiment. Journal of Applied Psychology, 81, 827–832. doi:10.1037/0021-9010.81.6.827 Bass, B. M. (1985). Leadership performance beyond expectations. New York, NY: Academic Press. Bass, B. M. (1999). Two decades of research and development in transformational leadership. European Journal of Work and Organizational Psychology, 8, 9–32. doi:10.1080/ 135943299398410 Bass, B. M., & Avolio, B. J. (1994). Improving organizational effectiveness through transformational leadership. Thousand Oaks, CA: Sage. Bass, B. M., & Avolio, B. J. (1995). Transformational leadership development: Manual for the multifactor leadership questionnaire. Palo Alto, CA: Consulting Psychologists Press. 154 Kimberley Breevaart et al. Bass, B. M., & Riggio, R. E. (2006). Transformational leadership. Mahwah, NJ: Lawrence Erlbaum. Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117, 497–529. doi:10.1037/ 0033-2909.117.3.497 Breevaart, K., Bakker, A. B., Demerouti, E., & Hetland, J. (2012). The measurement of state work engagement: A multilevel factor analytic study. European Journal of Psychological Assessment, 28, 305–312. doi:10.1027/1015-5759/a000111 Carlson, D. S., & Perrewe, P. L. (1999). The role of social support in the stressor–strain relationship: An examination of work–family conflict. Journal of Management, 25, 513–540. doi:10.1177/ 014920639902500403 Conway, J. M., & Lance, C. E. (2010). What reviewers should expect from authors regarding common method bias in organizational research. Journal of Business and Psychology, 25, 325–334. doi:10.1007/s10869-010-9181-6 Crawford, E. R., LePine, J. A., & Rich, B. L. (2010). Linking job demands and resources to employee engagement and burnout: A theoretical extension and meta-analytic test. Journal of Applied Psychology, 95, 834–848. doi:10.1037/a0019364 Doty, D. H., & Glick, W. H. (1998). Does common methods variance really bias results? Organizational Research Methods, 1, 374–406. doi:10.1177/109442819814002 Dvir, T., Eden, D., Avolio, B. J., & Shamir, B. (2002). Impact of transformational leadership on follower development and performance: A field experiment. Academy of Management Journal, 45, 735–744. doi:10.2307/3069307 Eisenberger, R., Rhoades, L., & Cameron, J. (1999). Does pay for performance increase or decrease perceived self-determination and intrinsic motivation? Journal of Personality and Social Psychology, 77, 1026–1040. doi:10.1037/0022-3514.77.5.1026 Etzion, D. (1984). Moderating effects of social support on the stress–burnout relationship. Journal of Applied Psychology, 69, 615–622. doi:10.1037/0021-9010.69.4.615 Halbesleben, J. R. B. (2010). A meta-analysis of work engagement: Relationships with burnout, demands, resources, and consequences. In M. P. Leiter & A. B. Bakker (Eds.), Work engagement: A handbook of essential theory and research (pp. 102–117). New York, NY: Psychology Press. Judge, T. A., & Piccolo, R. F. (2004). Transformational and transactional leadership: A meta-analytic test of their relative validity. Journal of Applied Psychology, 89, 755–768. doi:10.1037/ 0021-9010.89.5.755 Jung, D. I., & Sosik, J. J. (2002). Transformational leadership in work groups: The role of empowerment, cohesiveness, and collective self-efficacy on perceived group performance. Small Group Research, 33, 313–336. doi:10.1177/10496402033003002 Kovjanic, S., Schuh, S. C., & Jonas, K. (2013). Transformational leadership and performance: An experimental investigation of the mediating effects of basic needs satisfaction and work engagement. Journal of Occupational and Organizational Psychology, 86, 543–555. doi:10. 1111/joop.12022 Maas, C. J. M., & Hox, J. J. (2005). Sufficient sample sizes for multilevel modeling. methodology. European Journal of Research Methods for the Behavioral and Social Sciences, 1, 85–91. doi:10.1027/1614-2241.1.3.86 Muthen, L. K., & Muthen, B. O. (1998–2010). Mplus user’s guide (6th ed.) Los Angeles, CA: The Authors. Nielsen, K., & Daniels, K. (2012). Does shared and differentiated transformational leadership predict followers’ working conditions and well-being? The Leadership Quarterly, 23, 383–397. doi:10.1016/j.leaqua.2011.09.001 Nielsen, K., Randall, R., Yarker, J., & Brenner, S. O. (2008). The effects of transformational leadership on followers’ perceived work characteristics and psychological well-being: A longitudinal study. Work & Stress, 22, 16–32. doi:10.1080/02678370801979430 Nielsen, K., Yarker, J., Brenner, S. O., Randall, R., & Borg, V. (2008). The importance of transformational leadership style for the well-being of employees working with older people. Journal of Advanced Nursing, 63, 465–475. doi:10.1111/j.1365-2648.2008.04701.x Daily leadership and follower engagement 155 Nunnally, J. (1967). Psychometric methods. New York, NY: McGraw Hill. Ohly, S., Sonnentag, S., Niessen, C., & Zapf, D. (2010). Diary studies in organizational research: An introduction and some practical recommendations. Journal of Personnel Psychology, 9, 79–93. doi:10.1027/1866-5888/a000009 Piccolo, R. F., & Colquitt, J. A. (2006). Transformational leadership and job behaviors: The mediating role of core job characteristics. Academy of Management Journal, 49, 327–340. doi:10.5465/AMJ.2006.20786079 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. doi:10.1037/0021-9010.88.5.879 Preacher, K. J., Zyphur, M. J., & Zhang, Z. (2010). A general multilevel SEM framework for assessing multilevel mediation. Psychological Methods, 15, 209–233. doi:10.1037/a0020141 Purvanova, R. K., Bono, J. E., & Dzieweczynski, J. (2006). Transformational leadership, job characteristics, and organizational citizenship performance. Human Performance, 19, 1–22. doi:10.1207/s15327043hup1901_1 Salancik, G. R., & Pfeffer, J. (1978). A social information processing approach to job attitudes and task design. Administrative Science Quarterly, 23, 224–253. Schaufeli, W. B., & Bakker, A. B. (2004). Work engagement: The measurement of a concept. Gedrag & Organisatie, 17, 89–112. Schaufeli, W. B., & Bakker, A. B. (2010). Defining and measuring work engagement: Bringing clarity to the concept. In M. P. Leiter & A. B. Bakker (Eds.), Work engagement: A handbook of essential theory and research (pp. 10–24). New York, NY: Psychology Press. Schaufeli, W. B., Bakker, A. B., & Salanova, M. (2006). The measurement of work engagement with a short questionnaire: A cross-national study. Educational and Psychological Measurement, 66, 701–716. doi:10.1177/0013164405282471 Seibert, S. E., Wang, G., & Courtright, S. H. (2011). Antecedents and consequences of psychological and team empowerment in organizations: A meta-analytic review. Journal of Applied Psychology, 96, 981–1003. doi:10.1037/a0022676 Selig, J. P., & Preacher, K. J. (2008). Monte Carlo method for assessing mediation: An interactive tool for creating confidence intervals for indirect effects [Computer software]. Retrieved from http://quantpsy.org/ Skakon, J., Nielsen, K., Borg, V., & Guzman, J. (2010). Are leaders’ well-being, behaviours and style associated with the affective well-being of their employees? A systematic review of three decades of research. Work & Stress, 24, 107–139. doi:10.1080/02678373.2010. 495262 Smircich, L., & Morgan, G. (1982). Leadership: The management of meaning. Journal of Applied Behavioral Science, 18, 257–273. doi:10.1177/002188638201800303 Sonnentag, S., Dormann, C., & Demerouti, E. (2010). Not all days are equal: The concept of SWE. In M. P. Leiter, & A. B. Bakker (Eds.), Work engagement: A handbook of essential theory and research (pp. 25–38). New York, NY: Psychology Press. Spector, P. E. (2006). Method variance in organizational research: Truth or urban legend? Organizational Research Methods, 9, 221–232. doi:10.1177/1094428105284955 Tims, M., & Bakker, A. B. (2010). Job crafting: Towards a new model of job redesign. South African Journal of Industrial Psychology, 36, 1–9. doi:10.4102/sajip.v36i2.841 Tims, M., Bakker, A. B., & Derks, D. (2012). Development and validation of the job crafting scale. Journal of Vocational Behavior, 80, 173–186. doi:10.1016/j.jvb.2011.05.009 Tims, M., Bakker, A. B., & Xanthopoulou, D. (2011). Do transformational leaders enhance their followers’ daily work engagement? The Leadership Quarterly, 22, 121–131. doi:10.1016/j. leaqua.2010.12.011 Tuckey, M. R., Bakker, A. B., & Dollard, M. F. (2012). Empowering leaders optimize working conditions for engagement: A multilevel study. Journal of Occupational Health Psychology, 17, 15–27. doi:10.1037/a0025942 156 Kimberley Breevaart et al. Wrzesniewski, A., & Dutton, J. E. (2001). Crafting a job: Revisioning employees as active crafters of their work. Academy of Management Review, 26, 179–201. doi:10.2307/ 259118 Xanthopoulou, D., Bakker, A. B., Demerouti, E., & Schaufeli, W. B. (2009). Reciprocal relationships between job resources, personal resources, and work engagement. Journal of Vocational Behavior, 74, 235–244. doi:10.1016/j.jvb.2008.11.003 Xanthopoulou, D., Bakker, A. B., Heuven, E., Demerouti, E., & Schaufeli, W. B. (2008). Working in the sky: A diary study on work engagement among flight attendants. Journal of Occupational Health Psychology, 13, 345–356. doi:10.1037/1076-8998.13.4.345 Yukl, G. (2010). Leadership in organizations (7th ed.) Englewood Cliffs, NJ: Prentice Hall. Zhu, W., Avolio, B. J., & Walumbwa, F. O. (2009). Moderating role of follower characteristics with transformational leadership and follower work engagement. Group & Organization Management, 34, 590–619. doi:10.1177/1059601108331242 Received 25 April 2013; revised version received 10 September 2013 Appendix: Items of all study variables Transformational leadership 1. Today, my leader talked enthusiastically about what needed to be accomplished. 2. Today, my leader got me to look at problems from different angles. 3. Today, my leader helped me to develop my strengths. 4. Today, my leader emphasized the importance of having a collective sense of mission. 5. Today, my leader expressed confidence that goals would be achieved. Contingent reward 1. Today, my leader discussed in specific terms who was responsible for achieving performance targets. 2. Today, my leader made clear what to expect when performance standards were met. 3. Today, my leader expressed satisfaction when I met expectations. MBE active 1. Today, my leader focused attention on irregularities, mistakes, exceptions and deviations from standards. 2. Today, my leader concentrated his/her full attention on dealing with mistakes, complaints, and failures. 3. Today, my leader kept track of all mistakes. Autonomy 1. Today, did you have flexibility in how to perform your work? 2. Today, did you have control over how to carry out your work? 3. Today, I could participate in decisions regarding my own work. Daily leadership and follower engagement Social support 1. Today, when necessary, I could ask my supervisor for help. 2. Today, I could count on my supervisor when things were getting difficult. 3. Today, my work was recognized by my supervisor. Work engagement 1. Today at work, I felt bursting with energy. 2. Today at work, I felt strong and vigorous. 3. Today, I was enthusiastic about my work. 4. Today, my job inspired me. 5. When I got up this morning, I felt like going to work. 6. Today, I felt happy when I was working intensively. 7. Today, I was proud of the work that I do. 8. Today, I was immersed in my work. 9. Today, I got carried away when I was working. 157 Copyright of Journal of Occupational & Organizational Psychology is the property of WileyBlackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. Work & Stress ISSN: 0267-8373 (Print) 1464-5335 (Online) Journal homepage: http://www.tandfonline.com/loi/twst20 Transformational leadership and optimal functioning at work: On the mediating role of employees' perceived job characteristics and motivation Claude Fernet, Sarah-Geneviève Trépanier, Stéphanie Austin, Marylène Gagné & Jacques Forest To cite this article: Claude Fernet, Sarah-Geneviève Trépanier, Stéphanie Austin, Marylène Gagné & Jacques Forest (2015) Transformational leadership and optimal functioning at work: On the mediating role of employees' perceived job characteristics and motivation, Work & Stress, 29:1, 11-31, DOI: 10.1080/02678373.2014.1003998 To link to this article: https://doi.org/10.1080/02678373.2014.1003998 Published online: 02 Mar 2015. Submit your article to this journal Article views: 2567 View related articles View Crossmark data Citing articles: 11 View citing articles Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=twst20 Work & Stress, 2015 Vol. 29, No. 1, 11–31, http://dx.doi.org/10.1080/02678373.2014.1003998 Transformational leadership and optimal functioning at work: On the mediating role of employees’ perceived job characteristics and motivation Claude Ferneta*, Sarah-Geneviève Trépanierb, Stéphanie Austina, Marylène Gagnéc and Jacques Forestd a Department of Management Sciences, Université du Québec à Trois-Rivières, Trois-Rivières, QC, Canada; bDepartment of Psychology, Université du Québec à Montréal, Montréal, QC, Canada; c Department of Psychology, University of Western Australia, Crawley, WA, Australia; dSchool of Management, Université du Québec à Montréal, Montréal, QC, Canada (Received 21 March 2013; accepted 16 June 2014) This study aimed to deepen our understanding of the motivational mechanisms involved in the relationship between transformational leadership (TFL) and employee functioning. Drawing on the TFL literature, the job demands–resources model and self-determination theory, we propose an integrative model that relates TFL to employee psychological health (burnout and psychological distress), attitudes (occupational commitment and turnover intention) and performance (professional efficacy, self-reported individual and objective organizational performance) through two explanatory mechanisms: perceived job characteristics (job demands and resources) and employee motivation (autonomous and controlled). This research was conducted in two occupational settings (nurses and school principals), using a distinct variable operationalization for each. Results of both studies provide support for the hypothesized model, suggesting that TFL relates to optimal job functioning (psychological health, job attitudes and performance) by contributing to favourable perceptions of job characteristics (more resources and less demands) and high-quality work motivation (more autonomous motivation and less controlled motivation) in employees. Theoretical contributions and managerial implications as well as directions for future research are presented. Keywords: transformational leadership; job demands–resources model; self-determination theory; work motivation; psychological strain; job attitudes; job performance Introduction To remain competitive and to ensure sustainability, today’s organizations must adopt practices that foster high-quality functioning in their employees. Of the proposed practices, studies largely agree on the virtues of transformational leadership (TFL; Judge & Piccolo, 2004). This leadership style involves behaviours that transform employees’ standards and values and mobilizes them to achieve organizational goals that transcend *Corresponding author. Email: claude.fernet@uqtr.ca © 2015 Taylor & Francis 12 C. Fernet et al. their individual interests (Bass, 1985). TFL is associated with multiple motivational outcomes in employees, including empowerment (Avolio, Zhu, Koh, & Bhatia, 2004), autonomous motivation (Wang & Gagne, 2013) and self-concordance (Bono & Judge, 2003). By fostering positive perceptions of job characteristics, leaders can also affect employees’ attitudes, performance (Piccolo & Colquitt, 2006) and psychological health (Nielsen & Daniels, 2012; Nielsen, Randall, Yarker, & Brenner, 2008). Although the latter studies have advanced the understanding of the motivational effect of TFL behaviours, more research is needed to better grasp the role of TFL in employee functioning. Most studies have investigated the relationship between TFL and positive manifestations of employee functioning. However, in addition to contributing to employees’ attitudes and performance, whether TFL simultaneously prevents employees’ psychological strain remains to be explored. Moreover, although scholars have begun to address psychological mechanisms (e.g. emotional reactions, meaning, trust) linking TFL to employee functioning (see Arnold & Connelly, 2013), little is known about the motivational processes involved. Could job characteristics explain the motivational role of TFL in job performance, attitudes and psychological strain through the same motivational processes? To address these issues, we tested a motivational TFL model which posits that leadership is simultaneously related to employees’ perceptions of job demands and resources, which are differentially associated with employee motivation (autonomous or controlled motivation) and which are in turn linked to employees’ performance, attitudes and psychological strain (see Figure 1). This study makes a unique contribution to management theory by uniting three predominant perspectives: TFL theory (Bass, 1985), self-determination theory (SDT; Deci & Ryan, 2000) and the job demands–resources (JD-R) model (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001). Figure 1. The hypothesized model. Work & Stress 13 Theory and hypotheses TFL and perceived job characteristics Bass (1985) proposed one of the most well-established conceptualizations of TFL, which encompasses four behaviour types: idealized influence, inspirational motivation, intellectual stimulation and individualized consideration (Bass, 1985; Burns, 1978). Managers use idealized influence when they serve as a model or an example, encouraging employees to emotionally identify with them. They use inspirational motivation to encourage collective enthusiasm by sharing their values and goals, clearly communicating the organizational mission and convincing employees of the purpose and necessity of embracing challenges. In this way, managers persuade employees to adhere to the vision. They use intellectual stimulation to foster employees’ creativity, innovation and ideas for improvements. Finally, they show their consideration for employees’ individual needs through coaching, mentoring and other supportive behaviours. In the literature, these types of behaviours are generally strongly correlated (Judge & Piccolo, 2004), suggesting that TFL is a unitary higher order construct. To gain a deeper understanding of how TFL acts to motivate employees, we draw on the JD-R model (Demerouti et al., 2001), which considers two categories of job characteristic, job demands and job resources, which play a key role in optimal employee functioning. Job demands are the different taxing physical, psychosocial and organizational aspects of the workplace (Demerouti et al., 2001). They can be sources of stress when they hinder task achievement and generate cognitive, physical and emotional costs for employees (see Bakker & Demerouti, 2007, for a review). Examples of job demands include role-related problems (e.g. overload, ambiguity), interpersonal conflicts and organizational constraints. Job resources are various physical, psychosocial and organizational aspects that support individuals in their work (Demerouti et al., 2001). They can be of different nature (emotional, cognitive and physical). For example, social support, job control, recognition and skill discretion help employees accomplish their tasks and at the same time enrich their work and contribute to their personal and/or professional development and well-being (Bakker & Demerouti, 2007; Schaufeli & Bakker, 2004). Scholars have recognized that managers may wield considerable influence over their employees. For instance, they may define and embody the reality in which employees must work (Smircich & Morgan, 1982). Accordingly, Piccolo and Colquitt (2006) found that employees’ perceptions of their superior’s TFL are related to their perceptions of basic job conditions (variety, identity, significance, autonomy and feedback). Nielsen et al. (2008) found similar results in terms of role clarity, meaningful work and opportunities for development. Other studies, which did not specifically address TFL, also support this perspective. For instance, Schaufeli and Bakker (2004) found that perceptions of leadership quality (based on trust, respect and obligations) were positively associated with perceptions of performance feedback and social support by colleagues. Thus, through their actions, transformational leaders can influence perceptions of job resources by creating an environment that is conducive to communication and sharing, autonomy, as well as individual recognition. Transformational leaders can also shape employees’ perceptions of job demands. For example, they can mitigate job demands by providing employees with a meaningful rationale for the necessity or usefulness of demanding tasks. They can clarify workrelated ambiguities (Nielsen et al., 2008) by informing, answering questions and providing guidance or assistance when needed. Despite the lack of empirical studies 14 C. Fernet et al. specifically addressing TFL in relation to job demands, evidence points to a relationship between certain leadership behaviours and perceived job demands. For example, Schaufeli and Bakker (2004) found that perceptions of leadership quality (based on trust, respect and obligations) were negatively associated with work overload and emotional demands. Based on the above-presented theoretical rationale and the empirical literature, we propose the following hypotheses: Hypothesis 1a (H1a). TFL will be positively associated with employees’ perceptions of job resources (cognitive, emotional and physical in Study 1, participation in decision-making, job recognition and quality of relationships in Study 2). Hypothesis 1b (H1b). TFL will be negatively associated with employees’ perceptions of job demands (cognitive, emotional and physical in Study 1, work overload in Study 2). Perceived job characteristics and employee motivation SDT (Deci & Ryan, 2000) provides a comprehensive framework to understand the relationship between perceived job characteristics and employee motivation. An important distinction is made concerning the nature of motivation: not only do people invest in an activity to varied degrees—a quantifiable aspect—they also do so for various reasons—a qualitative aspect. Specifically, two broad forms of motivation— autonomous and controlled—capture the underlying reasons (motives) for performing a job (see Deci & Ryan, 2000; Gagné & Deci, 2005, for reviews). Autonomous motivation refers to acting with volition, as when employees engage in their job for the inherent pleasure and satisfaction (intrinsic motivation) and/or because they personally endorse the importance or value of a task (identified regulation). Controlled motivation refers to behaviours enacted under internal or external pressure, as when employees perform their job to enhance their self-worth or to avoid anxiety and guilt (introjected regulation) and/or because they are pressured by demands, threats or rewards by an external agent (external regulation). According to SDT, employees are motivated largely by “aspects of the social environment, including both aspects of the job and the work climate” (Gagné & Deci, 2005, p. 340). Environmental conditions that foster employee growth and development also encourage autonomous motivation. Conversely, a controlling workplace fosters controlled motivation. Regarding leadership, Kovjanic, Schuh, Jonas, Van Quaquebeke, and Van Dick (2012) showed that TFL is related to job satisfaction, occupational selfefficacy and commitment to the leader through the satisfaction of employees’ basic psychological needs (autonomy, competence and relatedness). According to SDT, this satisfaction provides the fuel required for optimal motivation (high autonomous and low controlled motivation) and consequently adaptive outcomes (Deci & Ryan, 2000). Apart from TFL-focused studies, SDT-based research generally shows positive associations between managers’ autonomy-supportive practices and employee motivation (e.g. Eyal & Roth, 2011). Other studies have associated perceived job characteristics with employee motivation. For example, Parker, Jimmieson, and Amiot (2010) showed that work overload was positively related to controlled motivation, whereas job control was positively related to autonomous motivation. In summary, these results concur with the premises of the JD-R model in that perceptions of job demands and resources impact employees’ psychological energy and motivation (Demerouti et al., 2001; Schaufeli & Work & Stress 15 Bakker, 2004). In light of the available empirical data, we propose the following hypotheses: Hypothesis 2a (H2a). Employee perceptions of job demands will be positively associated with controlled motivation. Hypothesis 2b (H2b). Employee perceptions of job resources will be positively associated with autonomous motivation. Work motivation and optimal employee functioning SDT-based research underscores the importance of considering the different forms of motivation as they are differentially associated with affective, attitudinal and behavioural outcomes (Gagné & Deci, 2005). Autonomous motivation has been positively associated with psychological well-being, job satisfaction, work engagement and occupational commitment, whereas controlled motivation has been positively associated with negative consequences for employees, including workaholism, burnout and turnover intention (see Fernet, 2013). Despite the empirical evidence, however, few studies have concomitantly examined the differential role of autonomous and controlled motivation in relation to various outcomes. In a nine-month longitudinal study, Fernet, Austin, and Vallerand (2012) found that employees’ autonomous motivation at time 1 simultaneously predicted occupational commitment and strain reactions (emotional exhaustion) at time 2, whereas controlled motivation predicted exhaustion only. Emotional exhaustion is a core dimension of burnout. It refers to an affective strain reaction resulting from overtaxing work (Maslach, Schaufeli, & Leiter, 2001). In contrast, occupational commitment reflects employees’ emotional attachment, involvement and identification with the occupation (Meyer, Allen, & Smith, 1993). To deepen our understanding of optimal employee functioning, we also investigated motivation in relation to other manifestations of job attitudes (turnover intention) and psychological strain (psychological distress) as well as job performance. Turnover intention (i.e. employees’ thoughts of quitting their current job; O’Driscoll & Beehr, 1994) and psychological distress (i.e. nonspecific symptoms of mental health disorders, such as anxiety or depression; Ilfeld, 1976) have been consistently related to employee motivation (e.g. Blais, Brière, Lachance, Riddle, & Vallerand, 1993). Job performance is defined as all actions and behaviours that are directly involved in the accomplishment of core job tasks and that contribute to organizational effectiveness (Motowidlo, 2003). We focus here on two aspects of job performance: professional efficacy (i.e. feelings of competence and productivity at work; Maslach et al., 2001) and task performance (i.e. in-role behaviours). Professional efficacy has been positively related to autonomous motivation and negatively related, albeit more weakly, to controlled motivation (Blais et al., 1993). Although the link between work motivation and task performance has been largely ignored, Kuvaas (2006) found a positive relationship between intrinsic motivation (which is prototypically autonomous) and selfreported task performance. In view of these findings, we propose the following hypotheses. Hypothesis 3 (H3). Employees’ autonomous motivation will be negatively associated with (H3a) psychological strain (burnout in Study 1, psychological distress in Study 2) and positively associated with (H3b) job attitudes (high occupational commitment in Study 1, 16 C. Fernet et al. low turnover intention in Study 2) and (H3c) job performance (task performance in Study 1, professional efficacy in Study 2). Hypothesis 4 (H4). Employees’ controlled motivation will be positively associated with (H4a) psychological strain (burnout in Study 1, psychological distress in Study 1) and negatively associated with (H4b) job attitudes (high occupational commitment in Study 1, low turnover intention in Study 2) and (H4c) job performance (task performance in Study 1, professional efficacy in Study 2). In our hypothesized model, TFL is related to employee functioning (attitudes, strain and performance) through two explanatory mechanisms: employees’ perceived job characteristics and motivation. We therefore propose that the relationships between TFL and employee functioning are mediated by perceived job demands and resources as well as controlled and autonomous motivation. The following specific hypotheses are proposed: Hypothesis 5 (H5): TFL will be negatively related to controlled motivation through perceived job demands. Hypothesis 6 (H6): TFL will be positively related to autonomous motivation through perceived job resources. Hypothesis 7 (H7): Perceived demands will be positively related to (H7a) psychological strain and negatively related to (H7b) job attitudes and (H7c) performance through controlled motivation. Hypothesis 8 (H8): Perceived resources will be negatively related to (H8a) psychological strain and positively related to (H8b) job attitudes and (H8c) performance and through autonomous motivation. Overview of the present research We tested our hypotheses in two cross-sectional studies in school principals and nurses. The variables were operationalized differently for each study in an attempt to capture the reality of the two professional settings (Bakker & Demerouti, 2007) and to allow for some triangulation and generalization of the results (Campbell, 1969). Thus, we sought to show that the motivational processes through which TFL is related to employee functioning remain consistent across different measures. The variables reflecting employee functioning were selected because they represent significant concerns for the two professions. For example, it is generally recognized that nurses are particularly at risk for burnout (e.g. Schaufeli & Enzmann, 1998), and that in Canada (where the study was conducted), 29% of school principals regularly question their career choice (Fortin, 2006), suggesting low occupational commitment. STUDY 1 Method Participants and procedure Study 1 was conducted among novice nurses working in public health care in the province of Quebec, Canada. An informative letter was sent to a random sample of 3800 nurses obtained from Quebec’s nursing association, inviting them to complete an online Work & Stress 17 questionnaire addressing workplace factors associated with well-being in the nursing profession. A total of 637 nurses participated in the study, for a 17% response rate. Participants were mostly women (88.4%), and mean age was 29.63 years (standard deviation [SD] = 9.40). Average job tenure was 3.47 years (SD = 3.45), and 56% of participants worked full-time (i.e. job status). Nurses who took part in the study had different work schedules: 23.9% worked the day shift, 28.3% the evening shift, 21.4% the night shift and 25.9% worked various shifts (day, evening and night). Despite the relatively low response rate, the sample is fairly representative of the demographics of novice nurses (with less than five years’ experience) in the association (e.g. 47% of nurses worked full-time; 86% were women; mean age 28.3 years). Measures All measures were administered in French. Original English scales were translated into French and then back-translated into English. English-speaking judges verified the semantic correspondence between back-translated and original items (see Vallerand & Halliwell, 1983). Means, standard deviations, Cronbach’s alphas and correlations are presented in Table 1. Transformational leadership. TFL was assessed using the seven-item Global Transformational Leadership scale (GTL; Carless, Wearing, & Mann, 2000). On a one- to five-point scale ranging from 1 (never) to 5 (almost always), participants rated their perceptions of the leadership shown by the nursing supervisor. A sample item is “He/she encourages us and recognizes our work.” The GTL has shown high convergent validity with lengthier and well-established questionnaires such as the Multifactor Leadership Questionnaire (MLQ) and the Leadership Practice Inventory (LPI) (Carless et al., 2000). In the present study, we created three parcels by pairing higher with lower loading items (see Little, Cunningham, Shahar, & Widaman, 2002) to assess the latent construct TFL. Using parcelling for unidimensional constructs improves the parsimoniousness of the model by reducing the number of estimated parameters. Because we were focusing on paths rather than the measurement model, parcelling was deemed appropriate (Bandalos & Finney, 2001). Job demands and resources. Job demands and resources were assessed using an adapted version of the DISC 2.0 questionnaire (van de Ven, Vlerick, & De Jonge, 2008). It contains six subscales assessing cognitive, emotional and physical demands and resources. Sample items for job demands are “I have to display high levels of concentration and precision at work” (cognitive; four items), “I have to deal with people whose problems affect me emotionally” (emotional; four items) and “I have to perform a lot of physically strenuous tasks to carry out my job” (physical; four items). Sample items for job resources are “I have access to useful information that helps me carry out complex tasks” (cognitive; four items), “I get emotional support from others when a tough situation occurs at work” (emotional; four items) and “I can take a break when my work becomes too physically strenuous” (physical; four items). Participants were asked to rate on a five-point scale from 1 (never) to 5 (almost always) the frequency with which they experienced these situations. Mean scores on each demand and resource type (cognitive, emotional and physical) were used as indicators of the latent constructs job demands and job resources, respectively, as recommended when conceptualizing multidimensional variables (Kline, 2005). 18 Table 1. Means, standard deviations and latent correlations between variables. Study1 Study 2 Mean SD α Mean SD α 1 2 3 4 5 6 7 8 1. 2. 3. 4. 5. 6. 7. 8. 2.77 5.03 4.41 3.35 5.59 2.99 3.27 6.14 1.00 0.98 0.96 1.01 0.90 1.08 0.77 0.71 .94 .83 .76 .73 .84 .83 .77 .90 3.31 3.28 3.70 3.70 4.71 1.67 2.02 6.12 0.83 0.78 0.56 0.56 0.93 0.43 1.47 0.62 .93 .78 .84 .88 .84 .77 .88 .84 − −.32** .45** −.14* .23** −.40** .42** .09* −.21** − −.35** .20** .03 .54** −.13** .08 .43** −.55** − −.04 .30** −.50** .40** .16** −.06 .18* −.22* − −.12* .37** −.11* −.14* .22* −.22** .70** .01 − −.55** .35** .31** −.15 .65** −.57** .13 −.39** − .50** −.24** −.28** −.30** −.51** .08 −.50** .35** − .16** .16* −.36** .67** −.17 .56** −.39** −.28** − Transformational leadership Job demands Job resources Controlled motivation Autonomous motivation Psychological strain Job attitudes Job performance Note: Correlations for Study 1 are below the diagonal. Study 1: Psychological Strain = burnout; Job attitudes = affective commitment; Job performance = in-role performance; Study 2: Strain = psychological distress; Job attitudes = turnover intention; Performance = professional efficacy; α = Cronbach’s alphas. *p < .05; **p < .01. C. Fernet et al. Variable Work & Stress 19 Work motivation. The Multidimensional Work Motivation Scale (Gagné et al., in press) was used to assess autonomous and controlled motivation. Participants rated on a sevenpoint scale from 1 (not at all for this reason) to 7 (exactly for this reason) their primary reasons for performing their job. Four motivational dimensions were assessed: external regulation (three items; e.g. “To get others’ approval”), introjected regulation (two items; e.g. “Because otherwise, I would be ashamed of myself”), identified regulation (three items; e.g. “Because this job has a personal significance for me”) and intrinsic motivation (three items; e.g. “Because my work is stimulating”). Mean scores on identified regulation and intrinsic motivation were used as indicators of the latent construct autonomous motivation, and mean scores on external and introjected regulation were used for the latent construct controlled motivation. Psychological strain. Psychological strain was operationalized by burnout, assessed with the emotional exhaustion and cynicism subscales of the MBI-GS (Schaufeli, Leiter, Maslach, & Jackson, 1996). Emotional exhaustion (e.g. “I feel used up at the end of a work day”) and cynicism (e.g. “I doubt the significance of my work”) were assessed with five items each. All items were rated on a seven-point scale ranging from 1 (never) to 7 (every day). Mean scores on subscales were used as indicators for the latent construct burnout. Job attitudes. Job attitudes were operationalized by occupational commitment, assessed with the affective commitment subscale of the Occupational Commitment Questionnaire (Meyer et al., 1993). A sample item is “I feel emotionally attached to my occupation” (six items). Items were scored on a five-point scale ranging from 1 (strongly disagree) to 5 (strongly agree). Two parcels were created by pairing higher with lower loading items to assess the latent construct occupational commitment. Job performance. Job performance was assessed using a four-item self-report scale adapted from the in-role performance subscale of William and Anderson (1991). On a 1 (do not agree at all) to 7 (very strongly agree) scale, participants indicated the extent to which they agreed with the proposed statements regarding their work performance. A sample item is “I adequately complete the tasks that are assigned to me.” Two parcels were created by pairing higher with lower loading items to assess the latent construct job performance. Analyses The hypothesized model was tested with structural equation modelling in Mplus (Muthén & Muthén, 2012). All models were tested with standardized coefficients obtained by maximum likelihood estimation. Goodness-of-fit was assessed using four indices: the Comparative Fit Index (CFI), the Tucker-Lewis Index (TLI), the Root Mean Square Error of Approximation (RMSEA) and the Standardized Root Mean Square Residuals (SRMR). Values above .90 for the CFI and the TLI indicate good fit (Hoyle, 1995), while values of .08 or less for the RMSEA and SRMR are deemed acceptable (Browne & Cudeck, 1993). Results and discussion study 1 Preliminary analysis We initially tested a measurement model (χ2(124) = 433.741, CFI = .925, TLI = .896, SRMR = .067, RMSEA = .070 [confidence interval, CI = .063, .078]), which provided 20 C. Fernet et al. satisfactory fit to the data (factor loadings above .47, p < .01). However, because JD-R– based research has typically conceptualized leadership as a job resource (e.g. supervisory support; e.g. Demerouti et al., 2001), we used confirmatory factor analysis (CFA) to compare the hypothesized structure (i.e. TFL and job resources as distinct constructs) to a model with leadership as a job resource. Results revealed that the latter model (χ2(98) = 440.681, CFI = .881, TLI = .835, SRMR = .076, RMSEA = .083 [CI = .075, .091]) did not provide good data fit, indicating that leadership is best represented as a distinct construct from job resources. The originally proposed conceptualization was therefore retained. We then performed a multivariate analysis of variance (MANOVA) to detect differences in variables according to significant background variables in nurses (gender, age, job status and work schedule). As no significant differences were found, demographic characteristics were excluded from further analysis. Testing the hypothesized model In Study 1, the hypothesized model (χ2(137) = 567.532, CFI = .896, TLI = .870, SRMR = .092, RMSEA = .079 [CI = .072, .086]) provided acceptable fit to the data, although there was room for improvement. We then examined whether including additional links in the model would significantly improve the model fit. Using a stepwise approach, we added links from TFL to work motivation and from job characteristics to outcomes. Six separately considered links significantly increased the model fit of the hypothesized model (TFL to controlled motivation, autonomous motivation and commitment; job demands to burnout and commitment; and job resources to commitment). Because these paths were consistent with prior empirical findings stemming from both the JD-R model and the TFL literature (e.g. Hakanen, Bakker, & Schaufeli, 2006; Piccolo & Colquitt, 2006), we included them in the model to more adequately represent the data. When these six links were added simultaneously, only four remained significant (TFL to autonomous motivation and commitment, job demands to burnout and job resources to commitment). A subsequent model, with the four links added to the hypothesized model, provided satisfactory fit to the data (χ2(133) = 463.163, CFI = .920, TLI = .897, SRMR = .077, RMSEA = .070 [CI = .063, .077]) and significantly improved the fit (Δχ2(4) = 99.12; p < .01). This final model (Figure 2; for simplicity, covariances are not shown) therefore obtained the best fit. The results nevertheless provide support for the hypothesized model: TFL is negatively related to job demands (supporting H1a) and positively related to job resources (supporting H1b). In addition, job demands are positively associated with controlled motivation (supporting H2a) and job resources are positively associated with autonomous motivation (supporting H2b). Moreover, autonomous motivation are negatively related to burnout and positively related to commitment and job performance (supporting H3a, H3b and H3c). As for controlled motivation, it is positively associated with burnout and negatively associated with job performance, but not commitment. These results provide support for H4a and H4c but not H4b. The final model explains 71% of the variance in burnout, 26% in job performance and 11% in occupational commitment. To more formally test the final model, we used bootstrapping to determine whether (1) job characteristics mediated the paths between TFL and employee motivation and (2) employee motivation mediated the paths between perceived job characteristics and the outcomes. Bias-corrected bootstrap 95% confidence intervals were computed from 1000 bootstrap samples. Confidence intervals indicate significant mediation when they exclude Work & Stress 21 Figure 2. Final model (standardized path coefficients). Top: coefficients in the nurses sample in Study 1 (psychological strain = burnout; job performance = in-role performance and job attitudes = occupational commitment). Bottom: coefficients on the left are based on the full sample of school principals in Study 2 (psychological strain = psychological distress; job performance = professional efficacy and job attitudes = turnover intention) and coefficients on the right are based on the supplementary analyses (job performance = school performance). Dotted arrows refer to the links included in Study 1 only. *p < .05; **p < .01. zero. As presented in Table 2, results indicated significant indirect relationships of TFL on controlled motivation through job demands and on autonomous motivation through job resources (supporting H5 and H6). Results also indicated significant indirect relationships of job demands on burnout and occupational commitment through controlled motivation (supporting H7a and H7b), and of job resources on burnout, commitment and job performance through autonomous motivation (supporting H8a, H8b and H8c). Note that H7c was not supported, because the indirect relationship of job demands on performance through controlled motivation was nonsignificant. A...
Purchase answer to see full attachment