Attachment and job stress in primary school leadership in the Netherlands : research of the potential mediating roles of Reappraisal, Suppression and Resilience and the potential moderating role of Empowering Leader Beh

(1)

Attachment and Job Stress

in Primary School Leadership

in The Netherlands

Research of the potential mediating roles of Reappraisal, Suppression and

Resilience and the potential moderating role of Empowering Leader

Behavior

Author : M.L. (Margriet Louise) van Ast Student Number : 10099387

Date of Submission : 30th_{of January 2018}

Master Thesis : MSc. EPMS – Leadership & Management Track Institution : Amsterdam Business School, UvA

(2)

Statement of Originality

This document is written by Student Margriet Louise van Ast who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Abstract

The aim of this study was to obtain better insight into the associations between Dutch primary school leaders’ attachment orientation (secure, anxious, pre-occupied and avoidant) and their experience of job stress, by testing the possible mediating roles of two emotion regulation strategies (reappraisal and suppression) and of resilience and by testing the possible moderating influence of the empowering behavior of the school leaders’ supervisor (board member) on the relation between school leaders’ attachment orientation and resilience. Research had a cross-sectional design and in order to gather data, in 2016, 2713 Dutch primary school leaders were asked to fill out a self-report questionnaire online. 224 cases were used for analysis. Findings showed that, first, secure attachment orientation was not directly related to job stress, but was associated with higher reappraisal, lower suppression and higher resilience of which only the latter showed a (inconsistent) mediation effect on job stress, meaning resilience acted as a suppressor variable (MacKinnon, Krull & Lockwood, 2000). Second, anxious attachment was not directly related to job stress and reappraisal, but was associated with higher suppression and lower resilience of which only the latter showed -again- a (inconsistent) mediation effect on job stress. Third, pre-occupied attachment was positively related to job stress, but showed no association with reappraisal or suppression. A full mediation effect on job stress was however found for pre-occupied attachment via resilience. Furthermore, avoidant attachment was not directly related to job stress, reappraisal and resilience, but was associated with higher suppression. No mediation effects were found. Overall, reappraisal and suppression showed no association with job stress and failed to function as a mediator between attachment and job stress. Finally, empowering leader behavior failed to function as a moderator of the relation between attachment and resilience. Implications of these findings are discussed.

(4)

List of tables and figures

Tables

Table 1 Means, Standard Deviations, Pearson Correlations and Reliability

of (Test) Variables p. 49

Table 2 Pearson Correlations of Test and Control Variables p. 50 Table 3 Exploratory Factor Analysis for Emotion Regulation

(Cognitive Reappraisal and Expressive Suppression) p. 69 Table 4 Exploratory Factor Analysis for Secure Attachment Orientation,

Anxious Attachment Orientation, Pre-occupied Attachment Orientation and Avoidant Attachment Orientation p. 69

Figures

Figure 1 A process model of emotion regulation p. 17

Figure 2 Research model for Attachment Orientation (IV) and Job Stress (DV) p. 37 Figure 3 Total effects (C), Direct effects (C’) and Indirect effects (C-C’) Secure

Attachment Orientation and Job Stress p. 54 Figure 4 Coefficients Multiple Regression Models Secure Attachment Orientation

and Job Stress p. 54

Figure 5 Total effects (C), Direct effects (C’) and Indirect effects (C-C’) Anxious

Attachment Orientation and Job Stress p. 57 Figure 6 Coefficients Multiple Regression Models Anxious Attachment Orientation

and Job Stress p. 58

Figure 7 Total effects (C), Direct effects (C’) and Indirect effects (C-C’)

Pre-occupied Attachment Orientation and Job Stress p. 61 Figure 8 Coefficients Multiple Regression Models Pre-occupied Attachment

Orientation and Job Stress p. 61

Figure 9 Total effects (C), Direct effects (C’) and Indirect effects (C-C’) Avoidant / Dismissive Attachment Orientation and Job Stress p. 64 Figure 10 Coefficients Multiple Regression Models Avoidant / Dismissive

(6)

1. Introduction

‘The school principal’s professional world is characterized by overwhelming responsibilities, information perplexities, and emotional anxiety’ (Friedman, 2002, p. 229). Whereas in the past schools were relatively stable organizations, nowadays, due to government policies aimed at decentralization and accountability, school leaders need to perform an increasingly complex and political task in a far more competitive environment (Hargreaves, 2000; Maroy, 2009). These developments may induce intense emotions leading to more strain and job stress (Maulding, Peters, Roberts, Leonard & Sparkman, 2012; Schmidt, 2010).

Chaplain (2001) analyzed perceived stress among a sample of primary head teachers in the east of the UK – who have similar responsibilities as Dutch primary school leaders - and identified that failing to manage oneself in terms of a perceived lack of personal control and self-efficacy is, among other factors, a major cause of job stress. Adopting a psychological framework for understanding a school principal’s job stress, acknowledgement of the presence of not only individual subjectivity in the perception of stress, but also of individual differences in (the perception of) the ability to cope with stress-eliciting events is vital (Chaplain, 2001; Gmelch, 1988; Lazarus, 1966).

These individual differences in perception can potentially stem from variations in individual attachment styles. ‘Adult attachment research has been based on Bowlby’s (1973, 1988) concept of ‘working model’ and on Ainsworth, Blehar, Waters and Wall’s (1978) tripartite typology of secure, avoidant, and anxious–ambivalent attachment styles. Working models are internal representations of the attachment relationships a person experiences throughout the life span. They organize cognition, affect, and behavior in adult relationships; guide affect regulation; shape self-image; and are the definitional component of a person's attachment style (Baldwin, Keelan, Fehr, Enns & Koh-Rangarajoo, 1996; Collins & Read, 1994; Shaver, Collins & Clark, 1996).’ (Mikulincer, 1998, p. 513). Attachment styles carry an individual’s inter- and intrapersonal dispositions, beliefs about self and others and

(7)

expectations about being able to cope with stressors (Bowlby, 1982; Mikulincer & Shaver, 2007). Through individual stress/threat appraisal patterns, that constitute the underlying mechanisms, attachment can determine individual mental health and well-being. This individual process of stress appraisal is enacted through individual resilience and coping / regulation strategies (Karreman & Vingerhoets, 2012; Mikulincer & Shaver, 2007). Karreman and Vingerhoets (2012) were the first to find mediating roles of emotion regulation and resilience in the relationship between individual attachment styles and well-being with regard to a community sample, and they suggested that future research should further examine the association of attachment orientation and these mediators on both positive and negative outcome variables (like job stress). In short, as individual differences in attachment style are associated with differences in personal threat appraisal (based on different working models about the self and others for each attachment style), with differences in self-regulation of negative or positive affect that stems from these personal threat appraisals (e.g. differences in preferred emotion regulation strategies for each attachment style) and with differences in resilience (a stress-resistant attitude that influences threat appraisal), these relations may result in different levels of perceived job stress. Still, especially with regard to educational leadership, as Berkovich and Eyal (2015) partly noted, little research has been conducted on the specific integral dynamics of acting school leaders’ individual traits (using a multivariate approach based on attachment theory), emotion regulation strategies, resilience and perceptions of job stress. Furthermore, as empowering leader behavior can increase follower’s feelings of self-efficacy and perceptions of control (Keller & Dansereau, 1995), which are constituents of resilience, it may also moderate the relation between follower’s attachment orientation and their perception of resilience – a potential antecedent of perceived job stress. However, no specific research has yet explored the impact of (empowering) leadership

(8)

behavior of board members of primary school foundations in relation to their follower’s (primary school leaders) resilience.

This study aims to fill these gaps by (i) examining to what extent attachment styles of primary school leaders predict their perception of job stress, (ii) investigating if connections between attachment styles and job stress are mediated by resilience and by two specific emotion regulation strategies (cognitive reappraisal and suppression), and (iii) exploring the possible moderating influence of the empowering leader behavior of school leaders’ supervisor on the relations between attachment orientations and resilience. The research had a cross-sectional design and data of 224 Dutch primary school leaders were collected and used for analysis. The findings contribute to the growing body of academic work on the influence of adult attachment dynamics on behavior and experiences in the workplace, with a specific focus on the level of leadership and may increase understanding of the factors that influence individual differences in the experience of job stress. Knowledge regarding the effects of individual differences in attachment orientation of school leaders, on their preferred strategies of emotion regulation, resilience and their perceptions of job stress, may be crucial for HR managers in educational organizations, in their efforts to select, train and guide school leaders in high emotional abilities or coping strategies in order to prevent negative job and organizational related outcomes like job stress, job dissatisfaction, burnout and turnover. Furthermore, the investigation into the moderating influence of empowering supervisor behavior may not only contribute to theory as little research has focused on the influence of leader behavior (board level) on middle management (school leader level) behavior and experiences in the workplace, but also to practice as it may result in knowledge advancement regarding the effectiveness of board member leadership and school leaders’ outcomes in the Dutch primary education sector in which independent individual schools largely become integrated in larger school foundations.

(9)

2. Literature review

2.1 Job stress

2.1.1 Definition

Lazarus (1966, p. 9) provided a definition of psychological stress, by stating that ‘it occurs when an individual perceives that the demands of an external situation are beyond his or her perceived ability to cope with them’. More specific, ‘job stress is a particular individual’s awareness or feeling of personal dysfunction as a result of perceived conditions or happenings in the work setting’ (Parker & DeCotiis, 1983, p. 161). At the core of these definitions is the idea that an emotion-eliciting situation is only stressful (or satisfying for that matter) if it is perceived as such by an individual; ‘pain and pleasure are transformed into emotional distress or satisfaction only as a result of appraisals of their significance’ (Lazarus, 1991b, p. 821).

2.1.2 Stressors and emotional demands in educational leadership

School principals face an unstable and competitive environment with frequent reforms due to economic, social, political, and technological developments (Hargreaves, 2000; Schmidt, 2010). Krüger, Van Eck and Vermeulen (2005, p. 242) described the context of school leader’s in The Netherlands: ‘Increase in scale, together with policies of decentralization and deregulation, defined as the reduction of central government regulations, have brought about an increase in school autonomy. These developments have multiplied the tasks and accountability of school principals.’ This abundance of responsibilities throughout the years has increased principal’s job stress (Maulding et al., 2012; Sodoma & Else, 2009), which in turn highly predicts burnout (Friesen & Sarros, 1989; Torelli & Gmelch, 1992). Although the majority of international research has focused on the experience of job satisfaction, job stress and burnout among teachers in primary and secondary schools, based on a summary structure

(10)

of Friedman (2002) which was further complemented, the following eight major stressors of school leaders are identified in literature: a) relations with colleagues, supervisors, parents or

students (Chaplain, 2001; Friedman, 2002), b) work overload (Chaplain, 2001; Friedman,

2002; Torelli & Gmelch, 1992), c) role in the organization, role conflict, and ambiguity (Mackler, 1996; Whitaker, 1996), d) organizational structure and climate (Burke, 1988; Knutton & Mycroft, 1986), e) outside agencies (e.g. family) (Burke, 1988; Cooper & Kelly, 1993), f) Career development (Burke, 1988; Whitaker, 1996), g) inadequate resources (Cooper & Kelly, 1993; Whitaker, 1996) and h) physical environment (Burke, 1988). These stressors may each elicit strong emotional reactions (affective experiences), which determine cognitions, motivations and behaviors in an occupational setting (George, 2000). Research has shown that the work of educational leaders may induce intense positive (e.g. passion, relief) and negative emotions (e.g. anger, anxiety, grief) (Ackerman & Maslin-Ostrowski, 2002; Loader, 1997). Friedman (1995) noted that exhaustion (caused by emotional demands that exceed emotional resources) and self-dissatisfaction (loss of self-esteem and efficacy) constitute the 'core' of a burned-out principal. As a (perceived) leader’s emotional state may serve as an indicator of (perceived) effectiveness (Lewis, 2000; Schmidt, 2010) it is important for educational leadership to deal with emotions in an optimal way. Within this respect, Blase and Blase (2004, p. 258) noted that ‘An awareness and understanding of emotions (even as they occur), the ability to manage one’s emotions, and the ability to express emotions in appropriate ways, given the context, are regarded as critical to effective school leadership (Beatty, 2000; Goleman, 1995).’

2.1.3 Managerial stress cycle and individual dispositions

Based on the work of Lazarus (1966), Gmelch (1988) presented a four stage ‘managerial stress cycle’ in his research of educational leadership stress. The first stage presents the

(11)

organizational and external stressors and demands placed on the individual school leader. The next stage consists of an individual’s perception and interpretation (appraisal) of these demands, stemming from personality or personal dispositions. Stress will occur if one perceives he or she cannot respond adequately to demands. The third stage presents choices: an individual (un)consciously selects a response in order to deal with or counteract the stressor(s). In the final, fourth, stage the outcome is presented in terms of stress and its long-term health effects or burnout. In addition, Gmelch (1988) noted the presence of filters that influence individual’s perceptions and responses and the interaction among the stages. These filters consist of two components: an individual’s disposition (the focus of this study) and personal factors like age and gender (control variables in this study). These filters may explain why, under the same stressors and responsibilities, some seem to (unconsciously) appraise and regulate intense emotions that potentially lead to more strain and job stress in a much more positive manner than others, consequently causing them to maintain higher levels of mental and physical health than others. With regard to the understanding of these individual differences in appraisals of and emotional responses to stressful workplace situations, Nelson and Sutton (1990) also stressed the recognition of individual dispositions, apart from situational effects. Especially the psychological function of emotion regulation shows a wide variety among individuals. Research has indicated that several personality variables are related to the appraisal and effects of stress and burnout, like esteem, self-consciousness, self-efficacy, locus of control, positive and negative affectivity (Alarcon, Eschleman & Bowling, 2009). A more fundamental framework that can account for differences in individual dispositions, responses and adaptation to positive or negative affect caused by stressors, may be provided by the psychodynamic approach of attachment theory (Bowlby, 1969, 1973, 1980).

(12)

Attachment Orientation refers to a theory in personality research, based on the work of John Bowlby (1969, 1973, 1980, 1982). It states that all individuals are born with an innate desire to seek proximity to others in times of need or distress in order to increase their chances of survival. Early positive or negative care giving (childhood) experiences are internalized and to the extent efforts to gain proximity to these care givers (‘attachment figures’) are successful, individuals develop a sense of security and mental working models of the self (as worthy or unworthy of love and attention from a care giver) and of others (as available and caring or unreliable and rejecting). Consequently, the sense of security (or lack of security) forms the basis of an individual’s attachment style.

Based on the work of Ainsworth, Behar, Waters and Wall (1978), who provided a primary framework of attachment styles, Bartholomew and Horowitz (1991) distinguished four attachment styles based on the models of the self and others: (1) secure attachment – reflecting a positive model of the self and of others, resulting in confidence in interpersonal interactions, (2) fearful / anxious attachment – reflecting a negative model of the self and of others, resulting in a lack of confidence in interpersonal interactions and a fear of getting hurt, (3) pre-occupied attachment – reflecting a negative model of the self and a positive model of others, resulting in a continuous need to get accepted by others) and (4) dismissive / avoidant attachment – reflecting a positive model of the self and a negative model of others, resulting in self-confidence, self-reliance and avoidance of intimacy in interactions with others). Weinfeld, Sroufe, Egeland and Carlson (2008) indicated associations between infant attachment security and later development of for example anger, empathy, esteem, self-efficacy, anxiety and interpersonal skills and Mikulincer & Shaver (2007) identified attachment as an important determinant of well-being. Attachment orientations remain

(13)

relatively stable over the lifespan of the individual and influence adult adaptive functioning and interpersonal experiences (Fraley & Brumbaugh, 2004; Fraley & Shaver, 2008; Mikulincer & Shaver, 2007). The last two decades increased research attention has been given to the integration of attachment theory into organizational behavior models, examining the role of attachment orientations with regard to individual differences in experiences and attitudes in the occupational setting (Harms, 2011).

2.2.2 Attachment and job stress

The work of Bowlby (1969, 1973, 1980, 1982) suggested that, in general life, individual attachment dynamics become more intensely activated when one is confronted with stressors. In line with this theory, Simpson, Rholes and Nelligan (1992) corroborated this theory in their findings that attachment related differences in support seeking increased when the perception of threat caused by an external event increased. Also, a well-known study by Mikulincer, Florian and Weller (1993) showed that, for subjects living in a dangerous region where war had just ended, individual attachment orientations were related to differences in reported levels of distress and in manifestations of stress. They found that anxiously attached individuals reported more distress than securely attached people and expressed a variety of stress manifestations (e.g. anxiety, depression, cognitive intrusions), whereas avoidant attachment people reported their distress in a more indirect way by expressing higher levels of somatization, hostility and avoidance than securely attached individuals, but structurally leaving anxiety and depression out of their responses. Differences in the experience of stress may stem from differences in stress appraisal patterns, which in turn are based on differences in attachment-related beliefs and expectations regarding the ability to cope with the situation (depending on the internal working models of the self and others) (Mikulincer & Shaver, 2007).

(14)

Although not comparable to life and death situations, occupational settings (especially educational contexts) may also place intense emotional demands on workers, which in turn may trigger individual attachment dynamics and consequently attachment related working models, stress appraisals, coping patterns and experiences (Harms, 2011; Mikulincer & Shaver, 2007). In the workplace securely attached individuals are equally comfortable with both closeness and separateness in social relationships and they value interdependence. Based on their internal working model of the self and others, when confronted with a stressor, they acknowledge their emotions and effectively seek support when needed, which prevents the compounding of distress (Mikulincer, Shaver, Pereg, 2003). They are also more inclined to provide others with support. Not surprisingly, relative to other attachment orientations, workers high on secure attachment have been found to have fewer work-related worries, to be more satisfied with received recognition and with their colleagues and to have fewer symptoms of illness (Hazan & Shaver, 1990). Also, a negative relation has been found between secure attachment and burnout (Pines, 2004). In the current study, a negative association between attachment security and perceived job stress is anticipated.

Anxiously and pre-occupied attached individuals experience intense feelings of job insecurity and fears of rejection as a result of weak job performance. In addition, they are hypervigilant to social cues from others and report a lack of received recognition and appreciation by colleagues (Fraley, Niedenthal, Marks, Brumbaugh & Vicary, 2006; Hazan & Shaver, 1990; Hardy & Barkham, 1994). Like pre-occupied adults, anxiously attached individuals fear separateness and rejection in social relationships, are overdependent and long for closeness, but they are more reluctant to seek social support. In general life, due to their models of the self and of others, pre-occupied and anxious attachment have been associated with appraising an external stressor as more threatening, which results in compounding distress (Mikulincer, Shaver & Pereg, 2003). Also, in an occupational setting, Schirmer and

(15)

Lopez (2001) indicated that pre-occupied and anxious attachment were positively correlated with work stress intensity and symptomatic distress. Later research also showed that anxious attachment was related to more job burnout (Ronen & Mikulincer, 2009) and that (based on a longitudinal design) anxious attachment predicted future perceptions of job stress and future job burnout (Ronen & Baldwin, 2010). For this reason, positive associations between anxious attachment and perceived job stress and pre-occupied attachment and perceived job stress are anticipated in this study.

Employees high on avoidant attachment style have been found to undervalue their job performance and to feel dissatisfied with their colleagues. In addition, they prefer to work alone and to do overwork in order to avoid social closeness, reflecting a counterdependent attitude (Hazan & Shaver, 1990; Hardy & Barkham, 1994). In general life, due to related models of the self and of others, avoidant attachment has been associated with appraising an external stressor as more threatening, which results in compounding distress (Mikulincer, Shaver & Pereg, 2003). They do however have a positive model of their own capability to deal with stressors, which may compensate their threat appraisal. As mentioned earlier, Mikulincer, Florian and Weller (1993) found avoidantly attached individuals to express their distress in more indirect ways (e.g. somatization, but no direct reports of anxiety or depression). Interestingly, Schirmer and Lopez (2001) found avoidant attachment to be significantly positively related to symptomatic distress, but the positive relation between avoidant attachment and perceived work stress intensity showed no significance. Later research indicated a positive relation between avoidant attachment and job burnout (Ronen & Mikulincer, 2009). Given these findings, the current study, explores the relation between individuals high on avoidance attachment and the perception of job stress.

Based on attachment orientations, individual differences may not only occur in stress appraisal, but also on the further enactment of this appraisal through individual resilience and

(16)

the preferred use of emotion regulation strategies (Bowlby, 1969, 1982; Karreman & Vingerhoets, 2012; Mikulincer & Shaver, 2007). Depending on the level of resilience and the use of specific emotion regulation strategies, intense emotions may be regulated and dealt with ineffectively (maintaining or increasing negative affect) or effectively (maintaining positive affect or reducing negative affect), which in turn may influence individual differences in the perceived level of job stress. Literature regarding the suggested mediating roles of emotion regulation and resilience on the relations between different attachment orientations and job stress is reviewed in the following paragraphs.

(17)

2.3 Emotion regulation

‘Emotion regulation refers to the processes by which individuals influence which emotions they have, when they have them, and how they experience and express these emotions. Emotion regulatory processes may be automatic or controlled, conscious or unconscious, and may have their effects at one or more points in the emotion generative process’ (Gross, 1998, p. 275). In line with his definition Gross (1998) modelled a general emotion regulation process (see Figure 1).

Figure 1 A process model of emotion regulation. According to this model, emotion may be regulated at five points in the emotion generative process: (1) selection of the situation, (2) modification of the situation, (3) deployment of attention, (4) change of cognitions, and (5) modulation of experiential, behavioral, or physiological responses. The first four of these processes are antecedent-focused, whereas the fifth is response-focused. The number of response options shown at each of these five points in the illustration is arbitrary, and the heavy lines indicate a particular option that might be selected. (Gross, 2001, p. 215)

Although, no empirically validated taxonomy has yet developed, several studies and meta-analyses suggest that people have numerous different strategies in order to regulate their (positive or negative) affect, which may differ in their effectiveness (Augustine & Hemenover,

(18)

2009; Bushman, Baumeister & Phillips, 2001; Koole, 2009). In general, Gross (2001) classifies emotion regulation strategies by the time at which they intervene in the emotion-generation process and makes a distinction between antecedent-focused emotion regulation and response-focused emotion regulation (see Figure 1). When an individual evaluates an event as threatening or as offering important opportunities, emotional response tendencies will develop. These tendencies may however be altered, beforehand (e.g. antecedent-focused emotion regulation) in order to alter the emotional impact of a situation, or afterwards (response-focused emotion regulation) in order to alter the behavioral response to an already felt emotion. As strategies activated early in the emotion-generative process are more adaptive than those later in the process, the way in which they were altered defines the character and intensity of the final emotional response (Gross, 2002). Apart from timing, Koole (2009) addresses two other higher orders that are useful in classifying emotion regulation strategies. First, emotion regulation may target three emotion-generating systems: (1) attention, (2) knowledge, or (3) the body. Second, emotion regulation may serve three psychological functions: (1) satisfying hedonic needs, (2) achieving an explicit goal, task or norm, or (3) improving personality functioning. With specific regard to (school) leadership and emotion regulation in the workplace it is important to emphasize that some emotion regulation strategies are not available: ‘(…) it seems that educational leaders, because of their role, cannot behaviorally disengage (Parkinson & Totterdell, 1999) from situations that elicit negative emotions (…); therefore, they choose either evasive or proactive strategies’ (Berkovich & Eyal, 2015, p. 150). Based on their review of empirical studies, five key strategies that educational leaders use to regulate their emotions in the workplace were identified: (1) emotional impression management (e.g. hiding and suppressing emotional expressions to promote cooperation, or masking fears in order to motivate employees), (2) cognitive disengagement (e.g. shutting down emotionally in order to deal with criticism), (3)

(19)

behavioral distraction (e.g. taking a break or a walk), (4) cognitive engagement (e.g. daily reflection or cognitively reframing emotion-eliciting events) and (5) behavioral engagement (e.g. sharing emotions with others) (Berkovich & Eyal, 2015). For the purpose of this research, the focus is on two of these strategies (in terms of Gross): cognitive reappraisal (4) and expressive suppression (1), as Gross (1998, 2001) identified both as two key strategies in the emotion regulation process and since are among the most commonly researched strategies in academic work. Furthermore, attachment orientations have been shown to be associated with both cognitive reappraisal and expressive suppression, reflecting individual differences in the preferred use of both emotion regulation strategies (Mikulincer & Shaver, 2007; Karreman & Vingerhoets, 2012).

2.3.2 Cognitive reappraisal

The emotion regulation strategy ‘cognitive reappraisal’ focuses on ‘changing the way one thinks about a potentially emotion-eliciting event’ (John & Gross, 2004, p. 1301) and is classified by Koole (2009) as a strategy that targets personal knowledge (cognition) and has a goal-oriented psychological function. It is an antecedent-focused strategy that is aimed at reframing the subjective evaluation of a situation beforehand (early in the emotion process, before emotion response is initiated) in order to reduce the potential negative interference of emotional impact: ‘a short circuit of threat’ (Lazarus, 1991a). In the domain of goal-oriented emotion regulation, cognitive reappraisal is one of the most effective strategies, in maintaining or enhancing positive affect, and especially in downregulating negative affect (Larsen & Prizmic, 2004). Individuals using reappraisal first of all acknowledge the stressor and may then reframe a potential threat by formulating it as a potential chance or as a challenging event and by doing so benefit from the adaptive qualities of emotions. With

(20)

regard to the use of this strategy in everyday life, cognitive reappraisal has been found to be negatively related to the experience and expression of negative emotions and positively related to the experience and expression of positive emotions (Gross & John, 2003, 2004). Although this strategy draws upon memory resources, it appears to do so in a rather efficient manner that does not negatively impact memory for social interactions (Richards & Gross, 2000). In this respect a positive relation was found between reappraisal and social support and closeness (Gross & John, 2003, 2004). In addition, various research has found positive relations for (frequent) use of reappraisal and well-being (e.g. Gross & John, 2003, 2004; Haga, Kraft & Corby, 2009; McRae, Jacobs, Ray, John & Gross, 2012; Nezlek & Cuppens, 2008).

2.3.3 Expressive suppression

The emotion regulation strategy ‘expressive suppression’ focuses on ‘changing the way one responds behaviorally to an emotion-eliciting event’ (Gross & John, 2004, p. 1301) and is classified by Koole (2009) as a strategy that targets (control of) bodily expressions of emotion (e.g. facial expressions or bodily postures) and, like cognitive reappraisal, has a goal-oriented psychological function. This strategy is however enacted at a later stage in the emotion process than cognitive reappraisal, when the experience of emotion has had their full dysfunctional effect, as the event was not regulated in more positive terms beforehand. Individuals using this strategy actively try to avoid/suppress their outer expressions of their inner emotional experience, creating a discrepancy between inner experience and outer expression. Although this strategy might be successful in keeping the body ‘in check’, in terms of reduction of the inner experience of negative or unwanted emotions, the (frequent) use of this strategy appears to be ineffective in terms of affective, cognitive and social outcomes (Gross & John, 2004). False self-representations may lead to negative feeling about

(21)

the self (Gross & John, 1998). With regard to the use of this strategy in everyday life, suppression has been found to be positively related to the experience of negative emotions and negatively related to the experience and expression of positive emotions (Gross & John, 2003, 2004). Physiological evidence of increased negative arousal due to suppression has also been provided (e.g. Demaree, Robinson, Pu & Allen, 2006). Cognitive and social costs are also in play, as memory of ongoing social interactions during the process of suppression is negatively affected and as suppression is linked to lesser social support (Gross & John, 2004). Finally, Gross & John (2004) reported that (frequent) suppression predicts lower well-being (Coté, 2005; Gross & John, 2004; Haga, Kraft & Corby, 2009; Niven, Totterdell & Holman, 2007).

2.3.4 Cognitive reappraisal, expressive suppression and job stress

With regard to the use of the emotion regulation strategies of cognitive reappraisal and expressive suppression in occupational settings, Grandey (2000) applied the emotion regulation theory of Gross (1998) to the concept of emotional labor and stated that ‘response-focused emotion regulation’ (like the strategy of ‘expressive suppression’, among others) corresponds with the process of ‘surface acting’ (p.99), as this involves the modification of expression (instead of feelings). Research showed strong positive associations between surface acting and emotional exhaustion, depersonalization and psychological strain and a negative association with job satisfaction (Hülsheger & Schewe, 2011). Surface acting has also been found to decrease self-authenticity (Erickson & Ritter, 2001), which in turn is associated with job stress (Erickson & Wharton, 1997). With respect to ‘cognitive reappraisal’, Grandey (2000) suggests this strategy to be one out of two components of ‘deep acting’ in a work context, the other one being another emotion regulation strategy: ‘attentional deployment’ (changing the focus of personal thoughts). However, whereas reappraisal (Gross, 1998) appeared to buffer stress reactions in several studies described above (Lazarus &

(22)

Folkman, 1991; Gross & John, 2004; Lazarus, 1991), research results with regard to the concept of ‘deep acting’ in reduction of stress in the occupational context (specifically focusing on client interactions in service industries) are mixed. To this date, also due to the complexity of emotion research, academic attention focuses on the (refinement of) the construct and measurement of emotional labor (e.g. Blau, Fertig, Tatum, Connaughton, Park & Marshall, 2010; Mesmer-Magnus, DeChurch & Wax, 2012). As the focus of this study is specifically on the use of reappraisal and suppression in relation to job stress, their measurement and the according hypotheses are therefore still based on the theory of Gross (1998).

2.3.5 Attachment, cognitive reappraisal and expressive suppression

Individual differences in attachment have been found to be related to the use of emotion regulation strategies (Karreman & Vingerhoets, 2012; Mikulincer & Shaver, 2007). As described, in the face of threatening emotional events, when internal support seeking tendencies arise, different individual attachment worries may arise. These worries are related to views of self (can I manage the threat?) and other (is someone available for support or will I be rejected?) one has developed and that influence the appraisal of threat. Securely attached individuals tend to perceive emotion-eliciting events as less threatening than insecurely attached individuals, due to a positive view of self and the support and availability of others. ‘Secure people can reappraise situations, construe events in relatively benign terms, symbolically transform threats into challenges, maintain an optimistic sense of self-efficacy, and attribute undesirable events to controllable, temporary, or context-dependent causes.’ (Mikulincer & Shaver, 2007, p. 191). As a result, they can have an open stance towards their emotions, express them freely and deal with the reality of the event. There is no automatic inclination to suppress any aspect of the emotion process, which makes them less likely to engage in expressive suppression (Mikulincer & Shaver, 2007). Consistent with these ideas,

(23)

Karreman and Vingerhoets (2012) found a significant positive effect of secure attachment on reappraisal and a significant negative effect on suppression.

People high on anxious attachment find it difficult to trust others and may regulate their emotions by intensifying their expressions in order to get attention. ‘(…) their doubts about support availability, coupled with fear of rejection, may make them hesitant at times to ask directly for assistance. As a result, they may be ambivalent about support seeking and may express their need for protection in indirect ways that seem less likely to provoke rejection (e.g. exaggerated facial expressions of sadness, (…)) (Mikulincer & Shaver, 2007, p. 194). Koole (2009), with respect to his categorization of various emotion regulation strategies, ordered this strategy of ‘response exaggeration’ (Schmeichel, Demaree, Robinson & Pu, 2006) in the same domain as he did for expressive suppression: a domain of strategies that target bodily expressions of emotion and have a goal-oriented psychological function. Although response exaggeration may seem like the opposite of expressive suppression, with regard to the specific emotion of anger, one must note the tendency of anxiously attached individuals to (bodily) suppress this emotion instead of expressing their anger directly in contact with an attachment figure, as they fear loss of contact as a result (Bowlby, 1973; Feeney, 1995, 1998). Feeney (1999) also found adult individuals high on attachment anxiety to be associated with greater self-perceived control of negative emotions (bottling up) or suppression (of anger). Remarkably, Karreman and Vingerhoets (2012) found no significant direct effect of anxious attachment on suppression and found a significant positive effect of anxious attachment on reappraisal, which was not hypothesized. The authors were reluctant in explaining this effect and stressed the importance of replicating the results. It is also important to bear in mind that the strategy of response exaggeration was not part of their –nor of the current- research. Still, given anxiously attached individuals’ generally untrusting views of others, their perceived gain of continuously maintaining negative affect and lower inclination to share or discuss

(24)

their feelings openly (Feeney, 1998, Main, 1990; Mikulincer & Shaver, 2005, 2007), in combination with the occupational setting that is the focus of this study, it is still primarily expected that anxious attachment has a positive effect on the use of (bodily) expressive suppression and a negative effect on the use of cognitive reappraisal.

Individuals high on pre-occupied attachment carry a negative view of self, but a positive perception of others (Bartholomew & Horowitz, 1991; Bowlby, 1973). They continuously worry about being accepted and appreciated and anxiously seek validation from others by reaching out to them. In order to get support, they frequently use a strategy that promotes emotion expression (e.g. crying) (Laan, Van Assen & Vingerhoets, 2012). Breuer & Freud (1955) defined the behavior of intentionally expressing emotions impulses freely, in order to control anger and aggression, as ‘venting’. Research has however indicated the ineffectiveness of this strategy in reducing internal affect (it is a response focused strategy focusing on change of expression instead of an antecedent focused strategy focusing on change of affect), as it results in an increase of anger and angry thoughts and behavior (Baumeister & Bushman, 2003; Geen & Quanty, 1977). Koole (2009) has categorized this strategy of venting in the same domain as he did for expressive suppression and response exaggeration: a domain of strategies that target bodily expressions of emotion and have a goal-oriented psychological function. This may also suggest an inclination towards the use of suppression over reappraisal. Although Karreman and Vingerhoets (2012) did find a significant negative effect of pre-occupied attachment on reappraisal, they did not find a significant positive effect of occupied attachment on suppression. This may be due to pre-occupied individual’s habitual use of the venting strategy, which was out of the scope of their research, or to their research focus on emotion regulation in general life (that may foster well-being). Venting is not part of this research either. Still, as the focus of this study is on emotion

(25)

regulation in the context of work, this may increase the insecurely attached individuals’ inclination towards the use of expressive suppression.

Avoidantly attached individuals are inclined to experience the environment as threatening, due to their experience of inconsistent support in times of stress; others are perceived as unavailable, unresponsive and punishing (Bartholomew & Horowitz, 1991; Bowlby, 1973; Mikulincer & Shaver, 2005). As a consequence, attachment and support seeking behavior has been shut down, affect has been repressed and self-reliance has increased. The emotion regulation efforts of avoidantly attached individuals are aimed at down-regulating any negative as well as positive emotions, as these emotions may activate attachment or induce social closeness. In doing so, they prevent social rejection or separation (Cassidy, 1994). ‘The avoidant approach to emotion regulation often interferes with support seeking, problem solving and reappraisal (…) because this requires recognizing threats and errors that avoidant people prefer to deny. Problem solving can also be blocked if it hints at the possibility of failure or requires a person to admit that some problems are unsolvable, which does not sit well with an avoidant person’s sense of, or wish for, autonomy and superiority.’ (Mikulincer & Shaver, 2007, p. 193). Remarkably, research indicated that avoidant attachment is not only positively related to suppression, but also to reappraisal (Karreman & Vingerhoets, 2012). This may be due to the self-rating method used and avoidant individuals’ tendency to evaluate their (in this case ‘emotional’) competence in an overly positive manner. It is however primarily expected that avoidant attachment has a positive effect on the use of suppression and a negative effect on the use of reappraisal.

With regard to the potential mediating role of reappraisal, Karreman and Vingerhoets (2012) found, based on their multiple mediation model, the positive relation between secure attachment and well-being to be partially mediated via reappraisal and resilience, the negative relation between pre-occupied and well-being and the positive relation between avoidant and

(26)

well-being to be fully mediated via reappraisal and resilience. No mediation role was found in relation to anxious attachment and well-being. With regard to the potential mediating role of suppression, suppression failed to function as a mediator between attachment and well-being, although Karreman and Vingerhoets (2012) did find suppression to be univariately associated with attachment styles and well-being. They suggested future research to focus on other outcomes. To this date no specific research is known with regard to the potential mediation role of these emotion regulation strategies on the relation between attachment and perceived job stress. Based on the literature review above, the following hypotheses are presented:

(H1.1) The negative relation between Secure Attachment Orientation and Job Stress is mediated via Reappraisal.

(H1.2) The negative relation between Secure Attachment Orientation and Job Stress is mediated via Suppression.

(H2.1) The positive relation between Anxious Attachment Orientation and Job Stress is mediated via Reappraisal.

(H2.2) The positive relation between Anxious Attachment Orientation and Job Stress is mediated via Suppression.

(H3.1) The positive relation between Pre-occupied Attachment Orientation and Job Stress is mediated via Reappraisal.

(H3.2) The positive relation between Pre-occupied Attachment Orientation and Job Stress is mediated via Suppression.

(H4.1) The positive relation between Avoidant Attachment Orientation and Job Stress is mediated via Reappraisal.

(H4.2) The positive relation between Avoidant Attachment Orientation and Job Stress is mediated via Suppression.

(27)

2.4 Resilience

The concept of resilience has been used in perspectives with regard to ecology (Holling, 1973), organizations (Powley, 2009; Bhamra, Dani & Burnard, 2011) and supply chains (Sheffi, 2005). The focus of this study is on individual resilience, defined by Luthans (2002, p. 702) as the ‘developable capacity to rebound or bounce back from adversity, conflict, failure, or even positive events, progress, and increased responsibility’ and later described as a dynamic process of adjustment (Youssef & Luthans, 2005). Masten (2001) stressed that the development of resilience is not a unique, but an ordinary process that stems from basic human adaptational (protective) systems. Resilience is seen as a multidimensional construct that consists of internal and external protective factors that enhance health by directly, positively influencing resilience and/or by buffering effects of stress- or emotion-eliciting events on resilience (Richardson, 2002; Friborg, Hjemdal, Rosenvinge, Martinussen, Aslaksen & Flaten, 2006). With regard to resilience in the workplace, the current study focuses on the work and measure of Wagnild and Young (2009a, 2009b), who identified five interrelated aspects of resilience, which they defined as the Resilience Core: (1) meaningfulness (having goals, values and a sense of purpose), (2) self-reliance (believing in oneself), (3) perseverance (persisting in the face of adversities), (4) equanimity (having a balanced view on life’s difficulties and opportunities and being able to deal with this reality in an optimistic way), and (5) existential aloneness (being comfortable in one’s own skin, following a unique path). Internal protective factors that were found to be associated with resilience are acceptance of reality, self-efficacy, perseverance, internal locus of control, hope, emotional intelligence, creative thinking, improvisation, coping / adaptation skills and a deep belief that life is meaningful (Coutu, 2002; Edward, 2005; Jackson, Firtko & Edenborough, 2007). External protective factors include (the quality of) social support networks (e.g.

(28)

familial and friend support) (Greeff & Holtzkamp, 2007; Hardy, Concato & Gill, 2004; Wilks, 2008). Furthermore, research has shown that resilience is state-like and open to personal development over a lifetime (Newman, 2005, Rutter, 2006; Youssef & Luthans, 2005). Still, the degree of individual resilience may differ, as a consequence of heterogeneity in people’s basic responses to environmental events that threaten or compromise their adaptational system. In this respect Karreman and Vingerhoets (2012) perceive resilience as a supplementary factor to emotion regulation in the process of stress appraisal: ‘Emotional regulation mainly involves the appraisal of stressful events as threatening or not, as well as the individual’s ways of dealing with stressors, whereas resilience particularly involves a stress-resistant attitude, related to the appraisal of oneself as able to cope with stressors’ (p. 821).

2.4.2 Resilience and job stress

With specific respect to the workplace, it may well be that resilient individuals are better equipped in conquering negative and positive (but potentially overwhelming) challenges and change. Research has demonstrated a positive relation between resilience and well-being (Gross & John, 2003; Mak, Ng & Wong, 2011; Wagnild & Young, 1993), employee performance (Luthans, Avolio, Avey & Norman, 2007), job satisfaction (Kim, Oh, & Park, 2011; Youssef & Luthans, 2007) and organizational commitment (Youssef & Luthans, 2007) and a negative relation between resilience and job stress (Kim, Oh, & Park, 2011). Avey, Luthans & Jensen (2009) combined resilience, hope, optimism and efficacy in the construct ‘PsyCap’ (Psychological Capital) and found a significant negative relationship with perceived job stress and stated that ‘resilience is arguably the most important positive resource to navigating a turbulent and stressful workplace’ (p. 682).

(29)

2.4.3 Attachment and resilience

As described earlier, attachment can be seen as a human adaptational system that encompas- ses different appraisals and working models of self and other that are instrumental to internal and external protective factors that have been found to be theoretically associated with resilience (Atwool, 2006; Masten & Obradovic, 2008). First, individuals high on secure attachment tend to perceive themselves as worthy and others as reliable, accepting, available and responsive (Hazan & Shaver, 1987; Bartholomew & Horowitz, 1991; Bowlby, 1973; Main, Kaplan & Cassidy, 1985). ‘The sense of self-cohesion (…) provides a subjective feeling of solidity, stability and permanence, and allows a person to feel coherent, consistent, and clear-minded even under threatening or unpredictable conditions’ (Mikulincer & Shaver, 2007, p. 153). Secure individuals have been found to have higher resilience (Karreman & Vingerhoets, 2012; Mikulincer & Florian, 1998; Mikulincer & Shaver, 2005). Second, individuals high on anxious attachment experience low self-efficacy and tend to perceive themselves as unworthy (Bartholomew & Horowitz, 1991; Bowlby, 1973). Negative relations were found between anxious attachment and stress-resistant attitudes (Birnbaum, Orr, Mikulincer & Florian, 1997; Mikulincer & Shaver, 2007; Neria et al., 2001), suggesting low resilience. Surprisingly, Karreman and Vingerhoets (2012) found anxious attachment to have a small, but direct positive effect on resilience, which was not hypothesized. Although they stress their results need replication, they explain this finding by arguing that the negative model that anxiously attached individuals hold with regard to others may compensate their model of the self that reflects the belief they are incapable of coping with stressors. Third, pre-occupied attachment also comes with a negative model of self and low coping skills, suggesting low resilience. Karreman and Vingerhoets (2012) also found a direct negative effect of this attachment orientation on resilience. Finally, individuals high on avoidance attachment tend to perceive themselves as worthy and appraise their coping capacity in the same positive manner as securely attached individuals do (Mikulincer & Shaver, 2007). They

(30)

do not focus on threats, block emotional experiences, are highly self-reliant and have a high feeling of self-efficacy and control. This may increase resilience, which in turn may serve as a buffer against stress. Karreman & Vingerhoets (2012) found avoidant attachment to have a direct positive effect on resilience. With regard to the potential mediating role of resilience, their research indicated that the strongest indirect effects of attachment on well-being ran through resilience. As described, they found resilience to be associated with reappraisal and the positive relation between secure attachment and well-being to be partially mediated via reappraisal and resilience and the negative relation between pre-occupied and well-being and the positive relation between avoidant and well-being to be fully mediated via reappraisal and resilience. No mediation role was found in relation to anxious attachment and well-being. To this date no specific research is known with regard to the potential mediation role of resilience on the relation between attachment and perceived job stress. Based on the literature review above, the following hypotheses are presented:

(H1.3) The negative relation between Secure Attachment Orientation and Job Stress is mediated via Resilience.

(H2.3) The positive relation between Anxious Attachment Orientation and Job Stress is mediated via Resilience.

(H3.3) The positive relation between Pre-occupied Attachment Orientation and Job Stress is mediated via Resilience.

(H4.3) The negative relation between Avoidant Attachment Orientation and Job Stress is mediated via Resilience.

Maulding et al. (2012) provided evidence that resilience is a significant predictor of educational leadership capacity and success (on the level of school leaders) and the

(31)

importance of (building) individual resilience in the workplace has become subject of a growing body of academic research. Within this respect, Harland, Harrison, Jones and Reiter-Palmon (2005) were the first to raise the question if leader behavior can make a difference in helping subordinates to become more resilient in the face of workplace adversity, focusing on the potential contributions of transformational, transactional and laissez-faire leadership. Results of their empirical research, based on a sample of MBA students, showed that four out of five dimensions of transformational leadership (attributed charisma, idealized influence, intellectual stimulation and individualized consideration) were positively associated with subordinate resilience and they suggested future research to examine the association of other leadership behaviors on followers’ resilience. With regard to building individual resilience in order to reduce the experience of stress, given the hypotheses of the current study, it is important to understand the conditions under which the potential relations between attachment orientations and resilience may be strengthened or weakened. In influencing these working conditions, this study focuses on the potential moderating role of school leaders’ supervisor leader behavior (board room level). More specific, the focus is on ‘empowering’ leader behavior, not only because ‘(…) many organizations today increasingly use leadership practices that give power, autonomy and responsibility to employees in order to enable and encourage them to be more receptive and adaptive to their work environment (…)’ (Gao, Janssen & Shi, 2011, p. 789), but also since empowering leadership has been theorised to be especially appropriate and effective for (amongst others) middle managers, in settings in which the value and uniqueness of human capital is high, tasks are complex and long term development is important (Houghton & Yoho, 2005; Liu, Lepak, Takeuchi & Sims, 2003). As the primary educational research context and the middle management level of the school leader sample reflect these characteristics, a focus on the potential moderating role of empowering leader behavior may therefore reflect current board room practice.

(32)

In general, ‘leadership styles refer to sets of behaviours that leaders employ to influence the behaviours of subordinates’ (Skakon, Nielsen, Borg & Guzmanc, 2010, p. 109). With regard to empowering leader behavior, Ahearne, Mathieu and Rapp (2005) identified four key aspects (which are the focus of our study): (1) enhancing the meaningfulness of employees’ work, (2) fostering employees’ participation in decision making, (3) expressing confidence in employees’ high performance, and (4) providing employees autonomy from bureaucratic constraints. These aspects may be theorized to reflect two important empowering leadership types: ‘encouraging self-development’ and ‘encouraging independent action’ (Van Dijke, De Cremer, Mayer & Van Quaquebeke, 2012). According to Cheong, Spain, Yammarino and Yun (2016) the concept of empowering leadership as defined by Ahearne, Mathieu and Rapp (2005) represents an ‘enabling’ process of empowering leadership, as it enhances employees’ self-efficacy and performance (as opposed to a theorized ‘burdening’ process of empowering leadership that may result in increased job-induced tension). Spreitzer (1996), Manojlovich (2005) and Zhang & Zhou (2014) also found empowering leadership to positively influence employees’ self-efficacy, next to additional research that provided evidence for positive associations between empowering leader behavior and intrinsic motivation (Zhang & Bartol, 2010), self-determination (Spreitzer, 1996; Yagil & Gal, 2002), creativity (Harris, Li, Boswell, Zhang & Xie, 2014; Zhang & Zhou, 2014), innovative behavior (Chen, Sharma, Edinger, Shapiro & Farh, 2011) and psychological empowerment (Auh, Menguc & Jung, 2014; Raub & Robert, 2013); all aspects that may reflect or positively influence the adaptive functioning of individuals in the face of workplace setbacks and stressors.

(33)

2.5.2 Attachment and resilience: the moderating role of empowering leader behavior

Surprisingly, little work on the moderating role of empowering leadership has been done (Sharma & Kirkman, 2015). However, especially when confronted with stressors, leaders can serve as attachment figures and can not only induce attachment dynamics by the degree to which they provide employees with perceived security and protection, but also influence enacted attachment related behaviors by employees by the degree to which they provide perceived support (Mayseless & Popper, 2007; Popper & Mayseless, 2003), suggesting a potential moderating role. As hypothesized in 2.4, attachment related individual differences, especially with respect to the model of self and one’s appraisal of one’s capability to deal with job stressors or challenges, are associated with differences in the individual level of resilience, which may serve as a buffer in the experience of job stress. On the one hand, these attachment related appraisals are suggested to be primarily based on the perception of one’s self-efficacy. Efficacy, as defined by Bandura (1997), is the belief that an individual has to successfully perform a specific task and as a factor of PsyCap having confidence to take on and put in the necessary effort to succeed at challenging tasks (Luthans et al., in press).’ (Luthans, Vogelgesang & Lester, 2006, p. 31). On the other hand, self-efficacy has been indicated to be highly relevant in influencing the level of personal resilience, self-efficacy (Avolio & Luthans, 2006; Bandura, 1997). Consequently, in relation to attachment orientations, high self-efficacy is suggested for secure and avoidant attachment, potentially resulting in high resilience, and low self-efficacy is suggested for anxious and pre-occupied attachment, potentially resulting in low resilience. In influencing the occupational context, by continuously encouraging self-efficacy beliefs and stressing and facilitating employees’ self-development, empowering leadership may moderate the relationship between attachment and resilience the following hypotheses are presented, in such a way that resilience will be higher when high empowering leader behavior is experienced. The following hypotheses are presented:

(34)

(H1.4) ‘Empowering Leader Behavior’ will moderate the relation between ‘Secure

Attachment Orientation’ and ‘Resilience’, such that ‘Resilience’ will be higher when high ‘Empowering Leader Behavior’ is experienced.

(H2.4) ‘Empowering Leader Behavior’ will moderate the relation between ‘Anxious

(H3.4) ‘Empowering Leader Behavior’ will moderate the relation between ‘Pre-occupied Attachment Orientation’ and ‘Resilience’, such that ‘Resilience’ will be higher when high ‘Empowering Leader Behavior’ is experienced.

(H4.4) ‘Empowering Leader Behavior’ will moderate the relation between ‘Avoidant Attachment Orientation’ and ‘Resilience’, such that ‘Resilience’ will be higher when high ‘Empowering Leader Behavior’ is experienced.

Alhough hypothesized is a moderating effect of empowering leadership on the relations between all attachment orientations and resilience, resulting in higher resilience, when high empowering leader behavior is experienced, differences in the strength of the moderating impact are anticipated, due to potential existing differences in preferred leadership styles based on attachment. Boatwright, Lopez, Sauer, VanDerWege and Huber (2010) for instance indicated (based on a sample of retail employees) that, on average and controlling for gender, anxious and pre-occupied followers preferred a relational leadership style and secure and avoidant followers preferred a task-oriented leadership style. Although to this date no specific research is known with regard to attachment related preferences for empowering leader behavior, it can be argued that elements of relational leadership like ‘(…)helping workers to feel comfortable with themselves (…) and empowering employees to contribute to the workplace (…)’(Boatwright et al., 2010, p. 2) are integral to the concept of empowering

(35)

leader behavior, which potentially could make secure and avoidant individuals less susceptible to the positive influences of their supervisor’s empowering leader behavior, which in turn could potentially reduce the moderating impact of empowering leadership on the relation between those attachment orientations and resilience.

(36)

Based on the literature review, the following research model is presented and hypotheses are summarized for each attachment orientation.

(H1) The effect of Secure Attachment Orientation on Job Stress is mediated via Reappraisal, Suppression and Resilience, whereby Empowering Leader Behavior moderates the relation between Secure Attachment Orientation and Resilience.

(H1.1) The negative relation between Secure Attachment Orientation and Job Stress is mediated via Reappraisal.

(H1.2) The negative relation between Secure Attachment Orientation and Job Stress is mediated via Suppression.

(H1.3) The negative relation between Secure Attachment Orientation and Job Stress is mediated via Resilience.

(H1.4) ‘Empowering Leader Behavior’ will moderate the relation between ‘Secure

(37)

(H2) The effect of Anxious Attachment Orientation on Job Stress is mediated via Reappraisal, Suppression and Resilience, whereby Empowering Leader Behavior moderates the relation between Anxious Attachment Orientation and Resilience.

(H2.1) The positive relation between Anxious Attachment Orientation and Job Stress is mediated via Reappraisal.

(H2.2) The positive relation between Anxious Attachment Orientation and Job Stress is mediated via Suppression.

(H2.3) The positive relation between Anxious Attachment Orientation and Job Stress is mediated via Resilience.

(H2.4) ‘Empowering Leader Behavior’ will moderate the relation between ‘Anxious

(H3) The effect of Pre-occupied Attachment Orientation on Job Stress is mediated via Reappraisal, Suppression and Resilience, whereby Empowering Leader Behavior moderates the relation between Pre-occupied Attachment Orientation and

Resilience.

(H3.1) The positive relation between Pre-occupied Attachment Orientation and Job Stress is mediated via Reappraisal.

(H3.2) The positive relation between Pre-occupied Attachment Orientation and Job Stress is mediated via Suppression.

(H3.3) The positive relation between Pre-occupied Attachment Orientation and Job Stress is mediated via Resilience.

(38)

(H3.4)‘Empowering Leader Behavior’ will moderate the relation between ‘Pre-occupied Attachment Orientation’ and ‘Resilience’, such that ‘Resilience’ will be higher when high ‘Empowering Leader Behavior’ is experienced.

(H4) The effect of Avoidant Attachment Orientation on Job Stress is mediated via Reappraisal, Suppression and Resilience, whereby Empowering Leader Behavior moderates the relation between Avoidant Attachment Orientation and Resilience.

(H4.1) The positive relation between Avoidant Attachment Orientation and Job Stress is mediated via Reappraisal.

(H4.2) The positive relation between Avoidant Attachment Orientation and Job Stress is mediated via Suppression.

(H4.3) The negative relation between Avoidant Attachment Orientation and Job Stress is mediated via Resilience.

(H4.4) ‘Empowering Leader Behavior’ will moderate the relation between ‘Avoidant Attachment Orientation’ and ‘Resilience’, such that ‘Resilience’ will be higher when high ‘Empowering Leader Behavior’ is experienced.

Attachment and job stress in primary school leadership in the Netherlands : research of the potential mediating roles of Reappraisal, Suppression and Resilience and the potential moderating role of Empowering Leader Beh