Burnout Research 2 (2015) 25–35

Contents lists available at ScienceDirect

Burnout Research
journal homepage: www.elsevier.com/locate/burn

Research Article

Attachment styles at work: Measurement, collegial relationships,

and burnout夽
Michael P. Leiter a,∗ , Arla Day b , Lisa Price a
a
Psychology Department, Acadia University, Wolfville, NS, Canada B4P 2R6
b
Psychology Department, Saint Mary’s University, 923 Robie St., Halifax, NS, Canada B3H 3C3

a r t i c l e i n f o a b s t r a c t

Article history: Although the potential deleterious effects of negative social interactions at work have been well estab-
Received 7 September 2014 lished in the literature, the impact of personal factors in forming work relationships has been relatively
Received in revised form 29 January 2015 neglected. Therefore, using a survey of 1624 Canadian healthcare providers, we examined the extent to
Accepted 27 February 2015
which attachment styles at work were associated with the quality of social relationships. We found sup-
port for a new measure of attachment styles at work that differentiated between anxiety and avoidance
Keywords:
attachment. Avoidance was negatively correlated with positive social constructs (civility, psycholog-
Burnout
ical safety, and trust) and with the efficacy dimension of burnout. Overall, compared to attachment
Attachment styles
Civility
avoidance, attachment anxiety was more strongly correlated with experienced and instigated workplace
Incivility incivility, exhaustion, and cynicism. Attachment avoidance was negatively correlated with positive social
Workgroup constructs (civility, psychological safety, and trust) and with the efficacy dimension of burnout. Adding
Healthcare these two attachment dimensions to a model of burnout as a function of workload, value congruence, and
Attachment anxiety coworker incivility significantly improved its fit. This study suggests that employees with high attach-
Attachment avoidance ment anxiety tend to be more closely involved in work relationships and processes, but this closeness
comes at a cost in that they experience more strain when participating in social encounters.
© 2015 The Authors. Published by Elsevier GmbH. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction permit some employees to function, or even thrive, despite strained

social encounters. However, personal constraints, such as persis-
Much of contemporary work occurs in a social context. Health- tent mistrust, can negatively influence thoughts about cooperating
care work is especially social in that employees work in teams that with colleagues. In developmental and social psychology, one of
call for ongoing contact among colleagues, managers, and mem- the key theories of developing effective social relationships is
bers of other professions or workgroups. Furthermore, most work attachment theory. Although attachment has been used to explain
in healthcare settings directly or indirectly pertains to patient care, individual differences in emotional and physical reactions to stress,
often requiring interactions with patients and their families. Even if styles of coping, and thoughts, feelings, and behaviors in a variety of
this work generally goes smoothly, employees regularly encounter interpersonal relationship situations (Mikulincer & Florian, 1995;
strained social interactions with colleagues, other professionals, Miller, 2007), it has been only recently applied to understanding
managers, and patients (Pearson & Porath, 2009) that may lead to interpersonal relationships at work.
negative individual outcomes, such as strain and burnout (Leiter, This relative lack of attention in the organizational literature,
Day, Laschinger, & Gilin-Oore, 2012). To some extent, employees however, should not be misconstrued as it being unimportant
call upon their professional training and life experience to manage in our understanding of work relationships. In fact, because of
difficult social circumstances. Personal capabilities and resources their ability to influence the quality of adult relationships, attach-
ment styles must be considered as an important part of social
relationships at work (Collins & Read, 1990), and therefore, has
relevance to workplace relationships. For example, a recent arti-
夽 This research was conducted with support from the Canadian Institutes for cle published in Financial Times highlights how executives often
Health Research (Partnership in Health Services Research: 114118) and the Social become an ‘emotional dumping ground’ for employees due to the
Sciences and Humanities Research Council of Canada (Insight Grant: 435-2013- tendency of employees to implicitly recreate early relationships
0177).
∗ Corresponding author. in the workplace (Shragai, 2014, para. 18). Executives can be left
E-mail address: [email protected] (M.P. Leiter). feeling somewhat ill-informed in how to deal with the emotional

http://dx.doi.org/10.1016/j.burn.2015.02.003
2213-0586/© 2015 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).
26 M.P. Leiter et al. / Burnout Research 2 (2015) 25–35

spillover from employees (Shragai, 2014). As such, a more devel- research on the links of victimization with the big five personal-
oped understanding of attachment theory may be beneficial in ity characteristics was inconclusive and contradictory, but there
understanding social relationships, and the related emotions, at was some support for self-esteem having consistent negative rela-
work. Therefore, we explored the efficacy of using attachment the- tionships with victimization (Aquino & Thau, 2009). Information
ory in the workplace to help explain social relationships and predict about personal dispositions that are closely associated with the
individual outcomes. More specifically, we examined the potential way people perceive and cognitively process social relationships
of this theory to explain healthcare providers’ experience of the could provide more specific directions for developing a model of
social context of their workplace by developing and validating a workplace social behavior than the general construct of negative
measure of attachment at work, and examining the relationships affectivity.
between attachment and burnout and civility outcomes. That is, One reason for the inconsistent results regarding the connec-
we: (1) introduce a new measure of workplace attachment; (2) link tions of personal characteristics with experiences of mistreatment
attachment styles to workplace social encounters; (3) link attach- may be the general nature of the personal characteristics studied.
ment styles to job burnout; and (4) expand a model of job burnout For example, only one of the big five personality characteristics,
to encompass attachment styles. agreeableness, directly references social qualities. In contrast, the
core dimensions of attachment explicitly reference social percep-
1.1. Social relationships at work tion and social behavior. As such, they may have a greater potential
for establishing links with employees’ experience of their work-
There is convincing evidence that social relationships at work place social environments.
have a significant impact on individual health, strain, and burnout
(Day & Leiter, 2014; Leiter & Patterson, 2014). Not only is 1.2. Attachment theory
greater social support associated with encountering few distress-
ing demands, but social support buffers the stressful impact of Attachment theory suggests that individuals are innately pre-
demands when they are encountered (Lakey & Orehek, 2011). In disposed to seek out comfort and safety from an attachment figure
contrast, uncivil or abusive social encounters are exhausting in (Bowlby, 1969). Constructs from attachment theory may explain
themselves and may contribute to spirals of increasing distress how individuals perceive, react to, and cope with stress arising from
that is associated with further unpleasant social encounters at interpersonal relationships (Mikulincer & Florian, 1995). Depend-
work (Andersson & Pearson, 1999). Although research has focused ing on the consistency of care in times of stress, individuals develop
primarily on organizational factors associated with poor social internal working models of self and others and a relatively stable
relationships at work, it also has considered individual factors pattern of stress response known as attachment style. According
associated with displaying or receiving negative social behavior to attachment theory, individuals who have experienced consis-
at work (Cortina, 2008; Zapf & Einarsen, 2011). An incivility spiral tent and supportive care from an attachment figure develop a
encompasses processes in which the emotional impact of receiving secure attachment style (Hazan & Shaver, 1990). Those who expe-
incivility prompts people to exhibit incivility toward others. Models rience inconsistent availability or consistent unavailability from an
explaining spirals emphasize the social dynamics and workplace attachment figure are theorized to develop an anxious or avoidant
values pertaining to civil behavior (Andersson & Pearson, 1999). attachment style, respectively (Hazan & Shaver, 1990).
Fortunately, there also is a potential for positive spirals, in which Attachment styles can be conceptualized using a two-
receiving civil behavior prompts people to experience positive dimensional approach in terms of avoidance of intimacy and
emotions and to exhibit more civility in return (Andersson & anxiety over abandonment (Bowlby, 1969; Miller, 2007). Individ-
Pearson, 1999). uals who are on the lower end of both dimensions are described as
Less attention has been given to identifying personal char- more securely attached. Securely attached individuals have pos-
acteristics that may be associated with incivility. However, the itive internal working models of both self and others: They are
construct of incivility has special relevance on this point because comfortable in relationships, have high self-efficacy in dealing with
the subjective nature of assessing its occurrence. Because the for- stress, and believe that others will be available to provide support
mal definition of the construct acknowledges ambiguous intent, when needed. Securely attached individuals tend to have better
the characterization of a behavior as uncivil lies entirely with the mental and physical health than insecurely attached individuals
recipient of that behavior (Andersson & Pearson, 1999). Neither (Mikulincer & Florian, 1995).
the intention of the actor nor a standardized description of rude Individuals higher on anxiety about abandonment tend to have
social behaviors indicates whether a behavior is uncivil. In related a negative view of self (Mikulincer & Florian, 1995). They tend to
research, personal characteristics impact one’s perceptions and be hypersensitive to signs of rejection and they have a compulsive
experiences. For example, negative affectivity may increase recip- need to be close to others. Ironically, this persistent need for close-
ients’ perception of bullying (Zapf & Einarsen, 2011) and incivility ness often prompts distance-seeking in the other person, which, in a
(Penney & Spector, 2005). In a comprehensive meta-analysis of cyclical fashion, can make those higher on anxiety attachment even
workplace harassment (Bowling & Beehr, 2006), the only individual needier (Miller, 2007). Furthermore, Mikulincer and Florian (1995)
difference variable with a consistent relationship with harassment attributed perceived unavailability from attachment figures to an
was negative affectivity. The authors reflected that the research to individual’s perceived own unworthiness of positive regard. Addi-
that point was inconclusive regarding the extent to which negative tionally, individuals high on anxiety attachment are likely to avoid
affectivity predisposed employees to harassment or resulted from instigating and participating in conflicts, because it may increase
the experience of harassment. They also speculated on potential chances of abandonment. Mikulincer and Florian argued that indi-
connections of personality characteristics—conscientiousness and viduals who are anxious about abandonment consistently monitor
agreeableness specifically—to workplace mistreatment but found their social environment for cues that support their beliefs about
little research examining these possible links. Aquino and Thau themselves. Furthermore, support for their beliefs is consistently
(2009) found a similar pattern regarding victimization from work- sought after even if those beliefs are negative.
place aggression. They found the most enduring relationships to be Individuals higher on avoidance of intimacy typically have a
with negative affectivity and concurred that the extant research negative view of others. They are compulsively self-reliant because
offered little insight on the extent to which negative affectivity they do not trust that others will be available to them when needed,
was a precursor or consequence of experiencing aggression. The and to the same degree, they often do not want people to depend
 M.P. Leiter et al. / Burnout Research 2 (2015) 25–35 27

on them (Miller, 2007). They prefer to keep a safe emotional dis- attachment avoidance presents distinct challenges for healthcare
tance from others when stressed (Mikulincer & Florian, 1995). This occupations and other industries working closely with the clients.
preference for distance is not only motivated by a negative view of Individuals higher on attachment avoidance are compulsively self-
people but also by distrust (Collins & Read, 1990). Because individ- reliant, and thus, they prefer to avoid close friendships or emotional
uals who are high on avoidance often do not seek the company involvement with others (Bartholomew, 1990). Mistrust character-
of others, it is common for them not to experience supportive izes social encounters with their colleagues or service recipients.
relationships that would contradict their previous beliefs about The lack of emotional connection with other people would deprive
relationships (Miller, 2007). Without exposure to these relation- avoidant employees of the potential benefits of a full participa-
ships that challenge existing beliefs, the avoidant attachment style tion in their team or their workplace community. Bartholomew
can persist throughout one’s adult life (Fraley, 2002). argued that their social encounters with others may not be actively
unpleasant but they lack emotional closeness. Interacting with
1.2.1. Attachment at work others through the formal structure of professional roles that pro-
Kets De Vries (1980) strongly advocated for the application vides emotional distancing from their colleagues would be well
of clinical practices to better understand workplace behavior. In suited to individuals who are higher on attachment avoidance
his work as an organizational consultant, De Vries found that not (Bartholomew, 1990; Bowlby, 1969). We propose that in contrast
all workplace behaviors were easily explained, such that man- to personal relationships, in which avoidance would be associated
agers viewed certain aspects of employee behavior as paradoxical. with isolation or loneliness, workplace avoidance would permit
However, De Vries argued that the realities of these seemingly ongoing social contact within the welcomed constraints of profes-
contradictory behaviors can be better understood through clini- sional roles.
cal principles. Building on this notion, Hazan and Shaver (1990)
were among the first to apply attachment theory to the workplace. 1.3. Civility and incivility at work
In their seminal article, they found that securely attached individ-
uals reported being more satisfied with various facets of their jobs, Research on workplace incivility has identified the disrup-
such as feeling competent and challenged at work, feeling secure tive impact of low intensity social encounters of ambiguous
with the job, and liking their coworkers. Secure individuals also had intent (Andersson & Pearson, 1999; Cortina, Magley, Williams, &
fewer psychological, psychosomatic, or physical symptoms of ill- Langhout, 2001). Leiter (2013) proposed a model emphasizing risk
ness in comparison to insecurely attached individuals. According to perception as a dimension of incivility. When people experience
Hazan and Shaver, individuals categorized as anxious/ambivalent social encounters as uncivil, they feel a risk for additional harm. We
were more worried about being rejected by coworkers and con- propose that a persistent sense of mistrust in social encounters has
cerned about the approval of others in the workplace. Finally, the potential to aggravate this dynamic. A workplace that permits
avoidant individuals were most likely to prefer to work alone and disrespectful relationships cannot be trusted to protect employ-
to use work as a way to avoid socializing. ees from humiliation, abuse, or other forms of mistreatment. The
A fundamental issue for attachment theory is whether secure identification of an uncivil interaction and the power of its impact
attachment is adequately defined by the absence of attachment depend to some degree on how the participants make sense of
anxiety and attachment avoidance. One solution to this question social relationships. Some people dismiss some uncivil interactions
has been to explicitly define and assess secure attachment as a dis- as trivial; others often experience incivility as extremely distressing
tinct construct in addition to assessing indicators of attachment (Leiter, 2013). One element contributing to these variations in expe-
anxiety and attachment avoidance (e.g., Joplin, Nelson, & Quick, rience could be social cognitions. We explore the proposition that
1999). An alternative solution follows the work of Fraley and Waller the core dimensions of attachment theory—anxiety about abandon-
(1998) by using both positively worded and negatively worded ment and avoidance of intimacy—contribute to that sense-making
items to assess anxiety and avoidance on two dimensions, Fraley process.
and Waller found strong support for a two-dimensional model of In general, anxiety may be more closely tied to the quality of
adult attachment. Their work supports two latent dimensions that social encounters at work because anxiety reflects a broad spec-
map onto the avoidance and anxiety dimensions of Bartholomew’s trum of concerns, including fear of embarrassment when in the
model (Griffin & Bartholomew, 1994b). Employees higher on anxi- presence of other people as well as fear of being deserted when
ety about abandonment might be more likely to misperceive social alone (Rapee & Heimberg, 1997). The general tendency is that
situations due to their concerns with feeling judged, criticized, higher scores on anxiety or avoidance are associated with a more
and rejected. This misperception may be particularly pertinent negative experience of work and social relationships (Hazan &
when dealing with incivility: Because of the ambiguous nature of Shaver, 1990; Rom & Mikulincer, 2003). Anxiety may be more
incivility, individuals higher on attachment anxiety may be more closely tied to the quality of social encounters at work because
likely to perceive innocuous acts as being uncivil. Thus, the anxiety anxiety reflects a broad spectrum of concerns, including fear of
associated with pervasive social interaction would be exhaust- embarrassment when in the presence of other people as well as fear
ing (Sonnentag, Binnewies, & Mojza, 2010). These concerns have of being deserted when alone (Rapee & Heimberg, 1997). Avoidance
the potential of inhibiting more anxious employees from effec- is less tied to the nature of social encounters because avoidance
tive team participation or clinical responsiveness as their social behaviors escape emotional connection. We propose that individ-
behavior might reflect their internal fears rather than responding uals higher on attachment avoidance limit social encounters within
appropriately to concerns of colleagues, clients, or patients. These professional roles. In this context, violations of professional deco-
experiences would contribute to more strained relationships with rum would be a form of incivility; however the formal nature of
people at work. The burden of compensating for this mismatch of workplace relationships for employees with high levels of avoid-
personal inclinations with their work context would increase their ance would reduce the emotional impact of a social encounter even
work demands. when it is perceived as uncivil or disrespectful.
The two dimensions of attachment have implications for Harms (2011) described consistent patterns of social relation-
employees dealing with the public. For example, although people ships associated with attachment styles. He argued that in a social
may choose healthcare careers because of a preference to work situation, people who are more securely or anxiously attached
with people, Savickas (2001) argued that they may lack insight into aspire to closeness, but people characterized as high on avoid-
their actual capacity to find fulfillment in such work. Furthermore, ance of attachment, do not aspire to such closeness. When in
28 M.P. Leiter et al. / Burnout Research 2 (2015) 25–35

close relationships, secure individuals have the capacity to build how healthcare workers higher on insecure attachment have a
trusting relationships while those higher on attachment anxiety greater potential for experiencing job burnout.
maintain mistrust (Tidwell, Reis, & Shaver, 1996). Attachment anx-
iety prompts intense emotions in social encounters in contrast to 1.5. Job burnout
secure people who maintain emotional stability (Cooper, Shaver, &
Collins, 1998). Given that those higher on attachment anxiety are Job burnout is a syndrome of chronic exhaustion accompanied
hyper-vigilant to signs of rejection, they would be more apt to per- by psychological distancing from work in the form of cynicism
ceive incivility in all social exchanges and would experience more or depersonalization. A lack of professional efficacy or dimin-
distress when they experience incivility. Individuals high in attach- ished sense of accomplishment completes the syndrome (Maslach,
ment avoidance may be more sensitive to some forms of incivility, Jackson, & Leiter, 1996). Research on job burnout has consistently
such as perceived inappropriate intimacy (e.g., discussing personal linked the syndrome with strained social relationships at work with
matters at work; see the Workplace Incivility Scale; Cortina et al., distinct contributions for supervisory relationships and collegial
2001), but they may be more likely to overlook or miss cues of relationships (Day & Leiter, 2014). Recent research has confirmed
other forms of incivility, such as being ignored or other forms of that improvements in the quality of collegial relationships are fol-
inconsideration. lowed by improvements in exhaustion and cynicism (Leiter et al.,
In summary, the dimensions of anxiety and avoidance that 2011), and that these improvements are evident at a one-year
underlie attachment styles are likely associated with social follow-up assessment (Leiter, Nicholson, Patterson, & Laschinger,
behavior at work. Incivility among colleagues violates ideals for 2012).
professional conduct and contradicts practical considerations for Psychological distancing has a defining role in both job
high quality team performance. We proposed that dispositional burnout and attachment theory. The original conception of
tendencies reflected in attachment avoidance or attachment anx- human service burnout (Maslach & Jackson, 1981) proposed
iety may further clarify the processes that sustain poor social that depersonalization—removing the emotional connection from
encounters among members of workgroups. In line with prin- social encounters with service recipients—reflects attempts to cope
ciples of attachment theory, we expect attachment anxiety and with exhaustion. Emotional social encounters are energy-intensive
avoidance to be reflected in the level of civility and incivility (Bakker & Demerouti, 2007; Morris & Feldman, 1996). To stem
employees encounter at work. Because trust is a fundamental issue the pace at which energy is being depleted, service providers may
in this theory, attachment styles should be negatively associated develop an impersonal social style. Other perspectives have pro-
with employees’ level of trust and their sense of psychological posed that cynicism (the parallel aspect within the general version
safety. of the MBI (Maslach Burnout Inventory—General Scale; Leiter &
Schaufeli, 1996)) reflects an inability to invest energy in work when
1.4. Trust at work experiencing exhaustion. The general burnout concept broadens
the original human service burnout perspective to propose that
In management research, trust is defined as, “the extent to which therapeutic relationships with patients are not the sole, or even
one is willing to ascribe good intentions to and have confidence in major, source of emotional demands at work. Any type of work
the words and actions of other people” (Cook & Wall, 1980, p. 40). can prompt emotional exhaustion, but work that demands intense
In team-based work cultures, trust has been found to predict bet- concentration, social encounters, or creativity are especially apt to
ter performance and greater innovation (Gilson, 2006; Laschinger, drain emotional energy (de Jonge, Spoor, Sonnentag, Dormann, &
Finegan, Shamian, & Casier, 2000). Trust is closely related to civility van den Tooren, 2012; Gutnick, Walter, Nijstad, & De Dreu, 2012;
and has been shown to improve when civility is improved in a work- Maslach et al., 1996).
place intervention (Leiter, Laschinger, Day, & Gilin-Oore, 2011). Recent research has found relationships of both attachment
Improvement in trust has been identified as a primary indicator anxiety and avoidance with job burnout. For example, managers’
of the viability of a healthcare system both for effective patient attachment insecurity predicted greater burnout and less sat-
care and personal fulfillment for providers (Berwick, 2003). Fac- isfaction among employees within their workgroups (Ronen &
tors that inhibit the development of trusting relationships present Mikulincer, 2012). Littman-Ovadia, Oren, and Lavy (2013) found
a practical concern for healthcare leadership. that the relationships of attachment avoidance with work-related
Closely aligned with civility and trust is psychological safety, outcomes were stronger under conditions of high job autonomy.
“a shared belief held by members of a team that the team is Therefore, we predict that both types of attachment insecurity will
safe for interpersonal risk taking” (Edmondson, 1999, p. 45). The predict higher levels of exhaustion and cynicism.
balance of civility and incivility evident in workplace social inter- As noted, healthcare providers have a high level of ongoing
actions provides employees with clues for assessing the team’s social contact with service recipients and/or colleagues. For peo-
riskiness. Employees’ level of psychological safety reflects their ple higher on attachment anxiety, ongoing social contact will make
experience in the workplace (Leiter & Laschinger, 2011; Pearsall exceptional demands on their energy beyond the usual demands
& Ellis, 2011), providing an overview of their perception of the of such contact. Given that concerns about social rejection pre-
team’s trustworthiness. It reflects employees’ relationships with dict burnout in those who are more anxiously attached (Ronen &
colleagues and supervisors. Moreover, psychological safety may Baldwin, 2010), we predict that relative to attachment avoidance,
reflect employees’ propensity to form trusting relationships. That the relationship between attachment anxiety exhaustion and cyn-
is, employees’ assessment of low psychological safety may reflect icism will be especially strong. Attachment anxiety augments the
more than an objective evaluation of the risks in the social envi- demanding quality of social encounters because the anxious per-
ronment: it also may reflect the employees’ overall approach son must contend with both the experience of anxiety per se as
to perceiving relationship as reflected in attachment anxiety or well as attempts to cope with that experience. When experiencing
avoidance. stressful demands, reducing anxiety takes priority over attending
Social relationships are important in themselves but they are to job performance (Bohlmeijer, Prenger, Taal, & Cuijpers, 2010).
also instrumental to fulfilling professional roles and obligations. Attachment styles also have implications for the reduced
Attachment styles not only play a role in the perception of social efficacy component of burnout. The more problematic social
encounters but also have implications for employees’ experience of relationships associated with anxiety and the more distant
worklife through teamwork and patient contact. We now consider relationships associated with avoidance could interfere with
 M.P. Leiter et al. / Burnout Research 2 (2015) 25–35 29

employees’ capacity to find a sense of accomplishment or con- remain stable across personal and work domains, and it is likely
firmation of their professional efficacy through their work. Much that work relationships have distinct qualities from romantic or
of hospital care is a team effort in which a variety of healthcare family relationships. A thorough consideration of the stability and
providers interact with a given patient over a series of shifts. Most domain specificity of attachment styles would potentially increase
accomplishments are shared accomplishments. Without a founda- the relevance of the construct. For example, confirmation of fluidity
tion of trusting relationships with colleagues, insecurely attached in attachment styles would encourage professional development
employees may have difficulties sharing the team’s accomplish- initiatives designed to assist leaders to gain insight into their own
ments. Without strong therapeutic relationships with patients, attachment perspectives and those of their employees, and to adapt
they limit or do not trust the direct feedback available from service their perspective as situations warrant, A general sense of belong-
recipients. Therefore, we predict that more insecurely attached ing (Deci & Ryan, 1991) has relevance to both domains, but the
individuals will have a lower level of efficacy in the workplace. emotional quality or importance of belonging in one’s personal
Research on incivility at work has consistently noted its recip- life is likely distinct from one’s worklife. Therefore, an effective
rocal nature (Pearson & Porath, 2009): People who report more measure of workplace attachment qualities would be sensitive to
frequent received incivility also report more frequent instigated the distinct qualities of attachment anxiety in contrast to attach-
incivility. Consistency of received incivility with instigated inci- ment avoidance. That is, the two dimensions would not be highly
vility may reflect the social culture of the work settings. That is, correlated with one another and that they would differ in their
cultures vary in their tolerance for incivility or their active pro- correlations with other constructs in ways that are consistent with
motion of civility (Lim & Lee, 2011). Another contributing factor their distinct qualities.
may be that more frequent reciprocal incivility reflects a dearth
of social skills. Both attachment anxiety and attachment avoid- 1.7. Summary and hypotheses
ance have been associated with poor social skills (Cooper et al.,
1998), suggesting a mechanism through which both forms of inse- Social relationships play an integral part in many workplaces,
cure attachment may be related to greater frequency of incivility, especially in healthcare. Much research has examined attachment
both instigated and received. theory in a variety of social situations in order to understand
and improve patterns of social relationships (Kets De Vries, 1980;
1.6. Measurement of attachment Hazan & Shaver, 1990), yet little work has examined this theory
in the workplace. Therefore, we developed and validated a mea-
There has been a long-standing debate around how to con- sure of attachment at work to examine the extent to which this
ceptualize and measure adult attachment in peer and romantic theory may help us improve healthcare providers’ experience of
relationships. Historically, researchers have used a categorical the social context of their workplace. We first examined the extent
approach (e.g., The Adult Attachment Interview; Main, Kaplan, to which the new attachment at work scale accurately represented
& Cassidy, 1985; The Relationships Questionnaire; Bartholomew the attachment construct. Because research has supported two dis-
& Horowitz, 1991; Hazan & Shaver, 1990) or used a three- tinct dimensions of social relationships (anxiety and avoidance;
dimensional approach (i.e., interdependent, counterdependent, Mikulincer & Florian, 1995), we hypothesized that:
and overdependent; Joplin et al., 1999). Fraley and Waller (1998)
Hypothesis 1. The anxiety and avoidance dimensions of adult
found the two dimensions mapped onto the avoidance and anxi-
attachment styles will factor into two separate, yet correlated, com-
ety dimensions of Bartholomew’s model (Griffin & Bartholomew,
ponents (i.e., anxiety and avoidance).
1994b). Furthermore, they noted a number of limits to the taxo-
nomic approach, such as the kinds of research questions that can Both of these forms of insecure attachment reduce the poten-
be asked, statistical power, and the ability to assess the stability of tial for people to experience fulfilling social relationships at work
attachment over time. (Lakey & Orehek, 2011), and thus, should be incompatible with a
More recent self-report measures use a dimensional approach sense of accomplishment or efficacy at work, trust in others, and
(e.g., Experiences in Close Relationships-Revised; Fraley, Waller, feeling ‘safe’ at work. Moreover, these forms of insecure attachment
& Brennan, 2000). However, most adult measures of attachment have been associated with negative personal outcomes in social
tend to focus on relationships with the family of origin or intimate situations (Lakey & Orehek, 2011). Therefore, we propose that:
partners (Fraley et al., 2000; Main et al., 1985). To our knowledge,
Hypothesis 2. Both anxiety and avoidance will be negatively asso-
there are no measures of adult attachment that are specific to rela-
ciated with (a) professional efficacy, (b) trust (c) psychological
tionships within the workplace. Existing measures, such as The
safety, and (d) civility; and positively associated with (e) incivility,
Experience in Close Relationships scale, have limitations in their
(f) exhaustion, and (g) cynicism,
application to relationships at work (Wei, Russell, Mallinckrodt, &
Vogel, 2007) because many of the items refer to romantic relation- We also propose that attachment avoidance and anxiety will
ships. These items may limit a measures’ applicability, because it differ in the extent to which they are related to some of these con-
is likely that people experience work-based relationships differ- structs. That is, attachment anxiety prompts employees to seek
ently from romantic relationships prompting distinct anxiety and close social relationships in hopes of alleviating insecurity. How-
avoidance orientations. Another limitation is that the two dimen- ever, because of their tendency to look for signs of rejection and
sions are not orthogonal. For example, Ronen and Mikulincer (2009) mistrust signs of acceptance, individuals higher on anxiety are more
reported that anxiety and avoidance were correlated (r = .32, p < .01) likely to perceive unpleasant social encounters. In contrast, indi-
and that the two scales had nearly equivalent correlations with viduals higher on attachment avoidance are more likely to avoid
the three aspects of job burnout, organizational fairness, and work- close social contact, and thus be less likely to notice some forms of
group cohesion. incivility than individuals who are higher on attachment anxiety.
Finally, Fraley (2002) argued that the two-dimension approach
Hypothesis 3. Compared to attachment avoidance, attachment
is the most accepted approach in contemporary attachment the-
anxiety will be more highly correlated with all forms of experienced
ory research. Therefore, our study used a new Short Workplace
incivility and instigated incivility.
Attachment Measure (SWAM; Leiter, Price, & Day, 2013) that builds
on previous attachment measures and makes specific reference to Similarly, anxious employees’ participation in close workplace
relationships at work. There is no evidence that attachment styles relationships with colleagues and service recipients will make
30 M.P. Leiter et al. / Burnout Research 2 (2015) 25–35

exceptional demands upon their energy and their involvement with 2.3.3. Incivility
work. The less intense social involvement associated with avoid- Incivility was measured with the 12-item Straightforward Inci-
ance should result in less exhaustion or disengagement with work. vility Scale (Leiter & Day, 2013) that included three subscales:
supervisor, coworker, and instigated incivility. Participants used a
Hypothesis 4. Compared to avoidance, anxiety will be more
7-point frequency scale (ranging from 0 – never to 6 – daily) to indi-
highly correlated with exhaustion and cynicism.
cate the extent to which they experienced uncivil behaviors from
To move attachment research forward in a work context, it is supervisors and coworkers, (e.g., “Spoke rudely to you.”) and the
important that attachment concepts fit into a theoretical frame- extent to which they engaged in uncivil behaviors toward others
work commonly used in organizational research. Specifically, it is (e.g., “Spoke rudely to others”). Cronbach’s alphas were ˛ = .82 for
proposed that the attachment dimensions will make significant supervisor incivility; ˛ = .94 for coworker incivility; and ˛ = .90 for
contributions to an existing model of job burnout as a function of one’s own instigated incivility
workload and coworker incivility (Leiter, Nicholson, et al., 2012).
2.3.4. Workgroup civility
Hypothesis 5. Both anxiety and avoidance will uniquely predict Workgroup Civility was measured with the 8-item CREW civility
exhaustion, cynicism, and efficacy, after controlling for workload, scale (Osatuke et al., 2009). Participants used a 5-point Likert-type
value congruence, and coworker incivility. scale (ranging from 1 – strongly disagree to 5 – strongly agree) to
indicate the extent to which they endorsed each item (e.g., “People
2. Method treat each other with respect in my work group.”). Cronbach’s alpha
was ˛ = .89.
2.1. Participants
2.3.5. Psychological safety
Participating organizations were four healthcare districts in Three items from a scale developed by Edmondson (1999) mea-
Eastern Canada. Participants included first-line managers (FLMs; sured psychological safety. Participants used a 5-point Likert-type
N = 157) and frontline staff (N = 1624). The average age for staff scale (ranging from 1 – strongly disagree to 5 – strongly agree) to
members was 43.14 years (SD = 10.84) and for FLMs was 47.35 indicate the extent to which the felt safe in their group (e.g., “It
years (SD = 8.68) (t(1978) = 4.94, p < .001). Staff members had an aver- is safe to take a risk in this work group.”). Cronbach’s alpha was
age of 15.97 years of healthcare experience (SD = 12.61); FLMs had ˛ = .81, with inter-item correlations ranging from r = .50 to .70.
an average 16.98 years (SD = 11.42) (t(1978) = 1.02, p ≤ .310). Partic-
ipants were from a wide range of healthcare professions with the
2.3.6. Coworker and supervisor trust
largest single group from nursing. The analyses used only data from
Coworker trust and supervisor trust were each measured with
frontline staff because staff and FLMs differed on many measures
three items from the trust subscale developed by Cook and Wall
in the study in addition to the demographics.
(1980). Participants used a 5-point Likert-type scale (ranging from
1 – strongly disagree to 5 – strongly agree) to indicate the extent to
2.2. Procedure which they felt they could trust their coworker or supervisor (e.g.,
“I can trust my co-workers to lend me a hand if I needed it”; “My
After receiving ethics approval from all participating hospitals own supervisor is sincere in his/her attempts to meet the workers’
and universities, the research team distributed surveys to health- point of view”). Cronbach’s alpha was ˛ = .72 for coworker trust and
care providers in four hospital districts in eastern Canada. The ˛ = .88 for supervisor trust.
project was presented as focusing on first line managers in health-
care and their challenges in managing the social environment of
2.3.7. Manageable workload
their workgroups during major organizational change. Participants
Manageable workload was measured using the Areas of Work-
had the option of completing the survey online or on paper.
life Scale (AWS; Leiter & Maslach, 2004). The items are worded as
statements of perceived congruence or incongruence between one-
2.3. Measures self and the job. Higher scores on manageable workload indicate a
better fit. All items (e.g., “I do not have time to do the work that
2.3.1. Attachment styles must be done”) are rated on a 5-point Likert scale ranging from
Attachment styles were assessed with the 10-item Brief Attach- 1 = strongly disagree to 5 = strongly agree. In the current study, the
ment Questionnaire comprising two subscales: anxiety (5-items) internal consistency was: ˛ = .79.
and avoidance (5-items; Leiter et al., 2013). The scale makes spe- Value Congruence. Value Congruence was measured using the
cific reference to relationships at work. Using a 5-point Likert type Areas of Worklife Scale (AWS; Leiter & Maslach, 2004). The items
scale (1 – not at all like me; 3 – somewhat like me; 5 – very much are worded as statements of perceived congruence or incongruence
like me), respondents indicated the extent to which items described between oneself and the job. All items (e.g., “My values and the
them (e.g., anxiety – “I worry that others don’t value me as much as organization’s values are alike”) are rated on a 5-point Likert scale
I value them”; avoidance – “I don’t need close friendships at work”. ranging from 1 = strongly disagree to 5 = strongly agree. In the current
Cronbach’s alphas were ˛ = .78 (anxiety) and ˛ = .78 (avoidance). study, the internal consistency was: ˛ = .77.

2.3.2. Burnout 3. Results

Burnout was measured with the 16-item Maslach Burnout
Inventory—General Scale (MBI—GS; Maslach et al., 1996) assessing In order to examine the factor structure of the Short Workplace
exhaustion, cynicism, and efficacy. Participants used a 7-point fre- Attachment Measure (SWAM), we conducted both an exploratory
quency scale (ranging from 0-never to 6-daily) to indicate the and confirmatory factor analysis. Table 1 displays a principal com-
extent to which they experienced each item (e.g., “I feel emotionally ponents factor analysis with Varimax rotation on the 10 SWAM
drained from my work.”). Cronbach’s alphas were ˛ = .94 for Emo- items, indicating two clear factors labeled anxiety and avoidance.
tional Exhaustion; ˛ = .81 for Cynicism; and ˛ = .88 for Professional The two factors accounted for 54.54% of the variance, and all
Efficacy. assigned loadings were greater than .56 and cross-loadings were
 M.P. Leiter et al. / Burnout Research 2 (2015) 25–35 31

Table 1
Principal components factor analysis with varimax rotation of attachment.

# Item Anxiety Avoidance

6 I worry that others don’t value me as much as I value them. .78 .06
9 I worry that I won’t measure up to other people at work. .76 −.01
8 I fear that friends at work will let me down. .72 .08
4 Others are often reluctant to be as close as I would prefer at work. .66 .10
10 I’m afraid to reveal too much about myself to people at work. .62 −.23
3 I make close friendships at work. (R) −.06 .84
1 I like to have close personal relationships with people at work. (R) .04 .82
7 A close friendship is a necessary part of a good working relationship. (R) .10 .77
2 I work hard at developing close working relationships. (R) .05 .66
5 I don’t need close friendships at work. .15 −.56

Bold values indicate coefficients > .50 on the appropriate factor.

Table 2
Means, standard deviations, and correlations of all study variables.

Mean SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Anxiety 1.98 .75 .78

Avoidance 2.94 .85 .01 .78
Exhaustion 2.19 1.49 .19 .03 .94
Cynicism 1.42 1.22 .24 .05 .64 .81
Efficacy 4.18 1.04 −.20 −.13 −.22 −.35 .88
Supervisor incivility .64 1.11 .13 .07 .39 . 42 −.15 .82
Coworker incivility .69 1.04 .26 .09 .28 .29 −.13 .22 .94
Instigated incivility .28 .46 .17 .05 .26 .32 −.14 .29 .53 .90
Civility 3.73 .70 −.24 −.25 −.32 −.41 .33 −.39 −.53 −.34 .89
Psych. Safety 3.39 .89 −.23 −.19 −.35 −.42 .28 −.41 −.47 −.28 .77 .81
Coworker trust 4.11 .64 −.27 −.18 −.19 −.26 .26 −.22 −.45 −.30 .61 .52 .72
Supervisor trust 3.68 1.06 −.14 −.11 −.40 −.49 .27 −.70 −.17 −.24 .47 .52 .26 .88
Manageable workload 3.03 .90 −.10 −.01 −.58 −.28 .10 −.20 −.17 −.09 .16 .14 .10 .19 .79
Value congruence 3.43 .75 −.09 −.10 −.31 −.40 .33 −.28 −.16 −.18 .34 .31 .24 .37 .14 .77

N = 1624. All correlations r > .06 are significant at p < .01. Reliabilities are along the diagonal.
Bold values indicate coefficients > .50 on the appropriate factor.

.23 or less, supporting Hypothesis 1. The factors were weakly cor- Table 3
Test of dependent correlations of anxiety and avoidance with burnout and incivility.
related r = .02, n.s.
A confirmatory factor analysis (CFA) conducted with EQS Variable Attachment Attachment t of
(Bentler & Chou, 1987). In this analysis, no error covariances were anxiety avoidance difference
freed; the first item in each factor was fixed at 1.00 and the other Incivility
items within each factor were freed. The covariance of the two fac- Supervisor incivility .13 .07 1.78*
tors was fixed. The CFA also supported a two-factor model with Coworker incivility .26 .09 5.06***
Instigated incivility .17 .05 3.49**
an adequate fit (2(35) = 325.79, p < .001; CFI = .928, RMSEA = .071),
Workgroup civility −.24 −.25 .30
and which was a significantly better fit (2(1) = 1996.74, p < .001) Safety & trust
Psychological safety −.23 −.19 1.19
than a one-factor model (2(35) = 2321.53, p < .001; CFI = .434, Coworker trust −.27 −.18 2.72**
RMSEA = .196), thus supporting Hypothesis 1. Supervisor trust −.14 −.11 .87
Burnout
Exhaustion .19 .03 4.79***
Cynicism .24 .05 5.75***
3.1. Relationship with burnout and social variables
Efficacy −.20 −.13 2.11**
*
The relationships of the attachment dimensions with burnout p < .05.
**
p < .01.
and social variables are displayed in Table 2. Both attachment ***
p < .001.
anxiety and avoidance were significantly correlated with effi-
cacy (r = −.20, p < .001; r = −.13, p < .001; respectively), coworker
trust (r = −.27, p < .001; r = −.18, p < .001), supervisor trust (r = −.14, 4 and 5), we conducted tests of dependent correlations. Anxiety
p < .001; r = −.11, p < .001), and psychological safety (r = −.23, was more strongly correlated with all forms of incivility than was
p < .001; r = −.19, p < .001), thus providing support for Hypothesis avoidance (supervisor, t = 1.78, p < .001; coworker, t = 5.06, p < .01;
2a, b, and c. and instigated, t = 3.49, p < .001; see Table 3). Anxiety was more
Anxiety also was positively correlated with supervisor incivil- strongly correlated with exhaustion, t = 4.79, p < .001, and cynicism,
ity (r = .13, p < .001), coworker incivility (r = .26, p < .001), instigated t = 5.75, p < .001, and efficacy t = 2.11, p < .01) than was avoidance
incivility (r = .17, p < .001), exhaustion (r = .19, p < .001), and cyn- attachment (see Table 3). Therefore, both Hypotheses 4 and 5 were
icism (r = .24, p < .001), and negatively correlated with civility supported.
(r = −.24, p < .001). However, avoidance was significantly corre-
lated only with supervisor incivility (r = .07, p < .01), coworker 3.2. Attachment contribution to a model of burnout
incivility (r = .09, p < .001), and civility (r = −.25, p < .001). Therefore,
Hypothesis 3 was partially supported. Finally, to test Hypothesis 6, we conducted a structural equa-
In order to test whether anxiety was more strongly correlated tion analysis using EQS (Equations; Bentler & Chou, 1987; Satorra
with incivility and burnout than was avoidance (i.e., Hypotheses & Bentler, 1988) to examine the contribution of attachment
32 M.P. Leiter et al. / Burnout Research 2 (2015) 25–35

these two paths, resulting in a nearly identical fit (CFI = .946;

RMSEA = .046) and making a significant improvement over the
Foundation Model (2(4) = 212.16, p < .001; see Fig. 1).
Adding the attachment dimensions increased the explained
variance for all three aspects of burnout (Foundation Model:
exhaustion, R2 = .386; cynicism, R2 = .484; efficacy, R2 = .180;
Attachment Model: exhaustion, R2 = .391; cynicism, R2 = .494; effi-
cacy, R2 = .233; see Fig. 1). In the final Attachment Model, all paths
were statistically significant except for coworker incivility to effi-
cacy (2 = 2.17, p = .141).

4. Discussion

The goals of this paper were to develop and validate a

work-based attachment styles scale, examining attachment’s rela-
tionship with incivility, civility, trust, psychological safety, and
Fig. 1. Expanded model of burnout: the impact of workload, value congruence, burnout. This study supports the inclusion of personal attachment
coworker incivility, anxiety, and avoidance.
styles in the study of workplace constructs, and is particularly rel-
evant to job burnout.
dimensions to predicting burnout beyond measures of workload, The results provide encouraging evidence for the capacity of
value congruence, and coworker incivility. In testing this model, we attachment styles to provide a distinct and relevant dimension to
combined the five items for exhaustion, cynicism, efficacy, coworker comprehensive models of workplace experiences. Increasing atten-
incivility, attachment anxiety, and attachment avoidance into three tion on workplace social relationships has emphasized the benefits
parcels. We defined workload and values with their respective of teamwork and the devastating impact of abusive interactions. A
three items. This approach focuses the analysis on the structural persistent issue in this research area is developing clear explana-
level of the model that is the primary concern in addressing the tions for the large variance in employees’ reactions to a given social
research question. As noted by Little, Cunningham, Shahar, and environment: That is, some employees manage well, whereas oth-
Widaman (2002) in their discussion of parceling, reducing the ers feel distressed. The current study provides preliminary support
number of items is appropriate when the primary research ques- that attachment styles may help to explain variations in employees’
tion concerns the relationships among the constructs. Using three perception of, and participation in, the social context of work.
parcels to define each construct addresses construct definitions We developed a workplace-specific measure of attachment, and
by assuring that the items cluster appropriately and that each we found support for two orthogonal dimensions of anxiety and
construct makes a contribution to the overall model (Leiter & avoidance in the newly developed social attachment at work scale.
Shaughnessy, 2006). In this analysis, the first item or parcel of each As hypothesized, anxiety was associated with incivility, civility,
construct was assigned the value 1.00 and the other two items were trust, psychological safety, and burnout. However, avoidance was
freed. No error variances were freed. significantly related only to efficacy, civility, and supervisor and
As noted in Table 4, the Structural Null model that assigns coworker incivility, thus providing support for Hypothesis 2 and
items to factors but leaves the factors uncorrelated has a poor fit, partial support for Hypothesis 3.
although it is a significant improvement over the Independence As expected, these two components differed in the strength of
model that lacks factor assignment. The Foundation model included their correlations with most of the other variables. That is, com-
two established paths among the burnout subscales: exhaustion to pared to avoidance, anxiety was more strongly (and positively)
cynicism and cynicism to efficacy. In addition, paths were freed correlated with the negative aspects of worklife, in terms of experi-
from coworker incivility and from value congruence to each of the enced and instigated incivility, exhaustion, and cynicism, and it was
three aspects of burnout and from workload to exhaustion and cyn- more strongly (and negatively) correlated with the positive aspect
icism. Previous research has demonstrated weak and inconsistent of efficacy and coworker trust. Therefore, it appears that avoidance
relationships of workload with efficacy (Maslach et al., 1996). Anx- may be a strategy for minimizing distress by simply avoiding social
iety and avoidance were included in the model with no paths to interactions.
other constructs. This model provided an adequate fit (CFI = .942; Moreover, the correlations of anxiety and avoidance with posi-
RMSEA = .051). The Attachment Model added six paths: anxiety tive aspects of worklife (i.e., civility, supervisor trust) were similar
to each of the three aspects of burnout and avoidance to each of in magnitude (and significant). That is, both of these types of
the three aspects of burnout. This model provided an improved attachment styles were associated with less workgroup civility,
fit (CFI = .943; RMSEA = .047) that was a significant improvement psychological safety, and efficacy, suggesting that neither insecure
over the Foundation Model (2(6) = 214.34, p < .001), supporting attachment style may benefit from these types of positive aspects
Hypothesis 6. Two of the paths had very small coefficients (avoid- from work and colleagues.
ance to exhaustion, ˇ = −.004; and avoidance to cynicism, ˇ = .005). Finally, we then examined the impact of adding these inse-
Therefore, we examined another Attachment Model that deleted cure attachment styles to a model of burnout. Even though three

Table 4
Structural equation model of burnout, worklife areas, and attachment styles.

Model Chi-square df CFI RMSEA Difference df Sig

Independence 15,991.98 276

Structural mull 3344.77 252 .804 .085 12,647.21 24 .001
Foundation model 1306.55 242 .931 .051 2038.22 10 .001
Attachment model (6 paths) 1110.76 235 .943 .047 195.79 7 .001
Attachment model (4 paths) 1112.94 237 .943 .046 193.61 5 .001
 M.P. Leiter et al. / Burnout Research 2 (2015) 25–35 33

of the pathways were not significant, the additions of attachment first line managers, which reduced the motivation of non-managers
improved the model fit, and both avoidance and anxiety made sig- to participate. However, the sample size of over 1600 healthcare
nificant contributions to predicting the three aspects of burnout providers produces sufficient power to explore the constructs in the
beyond the contribution of workload and value congruence with model. Our explicit focus was on healthcare providers. However,
distinct paths for anxiety to exhaustion and cynicism and for avoid- future research should examine other workers to see if the patterns
ance to efficacy. of relationships are similar and whether the scale is appropriate for
The results of these analyses encourage further exploration those populations.
of attachment styles as a meaningful component of employees’ Another direction for future research is focusing on the stability
experience of workplace relationships. The relationships with both of the construct of attachment at work over time and across sett-
received and instigated incivility suggest a pervasive role of these ings. In light of the considerable amount of research on attachment
cognitive/emotional schemas for interpreting the social context of styles pertaining to personal relationships, it would be informa-
work. The relationship of anxiety with the exhaustion and cyni- tive to determine the extent to which attachment styles may differ
cism aspects of burnout, even after controlling for the effects of in one’s work and personal life. That is, perhaps some people
workload and value congruence, suggests a close association of may maintain an avoidant attachment style at work while having
attachment and employees’ experience of their careers. This rela- a secure attachment style in their personal relationships. More-
tionship may be especially relevant for employees such as the over, the quality of interactions at work—in terms of civility and
healthcare providers, whose work involves ongoing social relation- incivility—may contribute to employees developing distinct attach-
ships. ment styles as work in contrast to their personal lives. These issues
Anxiety’s stronger correlations (relative to avoidance) with inci- could be explored using longitudinal designs and by tracking social
vility and burnout may reflect their distinct qualities in the social perceptions and experiences during the initial years of employ-
context of healthcare work. Attachment anxiety motivates peo- ment. In a parallel fashion, attachment styles may change following
ple toward close relationships despite their inherent risks, such as promotion from team member to management as their role expec-
humiliation and unreasonable demands. For those high on attach- tations change at work.
ment anxiety, these strains are preferable to fear of rejection. A cost Another important line of research is to examine the correspon-
of this trade-off is that anxious people have more social encoun- dence of attachment styles as measured by the SWAM with an
ters that they consider uncivil. If they are more intensely involved existing measure of attachment styles in personal relationships. A
in the workflow on their units, which is characterized by ongoing longitudinal, repeated measures design could determine the extent
communication exchanges in the highly team-oriented work pat- to which both measures cover the same domain and the extent to
terns of contemporary healthcare, they may be more apt to view which attachment styles at work comprise a different set of social
interactions as being uncivil, experience less trust, and report more cognitions and perceptions than do attachment styles in personal
burnout. relationships. The relative stability of attachment style over time
In contrast to attachment anxiety, attachment avoidance may would be of interest, especially for people for whom major life
have practical ‘advantages’ for employees lacking secure attach- events—promotion at work; a new romantic relationship in per-
ment qualities. By avoiding social contact with colleagues, they may sonal life—occur during the course of the study.
reduce their overall rate of incivility. However, similar to those high A conceptual issue requiring exploration is whether secure
in anxiety, those high in avoidance still experience lower workplace attachment style is adequately captured by the absence of avoid-
civility and trust. Attachment avoidance was not associated with ance and anxiety inclinations. A subscale comprising both positive
greater risk for exhaustion and cynicism, but it was associated with and negatively worded items, such as the avoidance subscale of
lower efficacy. It may be that maintaining a social distance avoids the SWAM, includes both active endorsement of avoidance implies
some workplace demands that create burnout, but it also deprives that the desire for and rejection of close relationships at work define
them of some opportunities for significant workplace accomplish- opposite poles on a continuum. A more far-reaching approach to the
ments. Future research must look at these issues in more detail construct would include not only opposite terms but bring in addi-
and use a longitudinal design before making any conclusions about tional elements beyond direct expressions of avoidance or anxiety.
directionality. This type of analysis is necessarily theory-driven because bring-
ing in elements other than anxiety and avoidance (or their direct
4.1. Practical implications opposites) redefines the construct to some extent.
Although we did not examine a moderating effect in the cur-
Having a greater understanding of attachment styles, and poten- rent study, attachment styles also may have a potential to impact
tial benefits and outcomes may have several implications for employees’ management of challenging social encounters. That is,
organizations. It may benefit managers to recognize their own attachment styles may influence the extent to which difficult social
attachment styles to help them perceive social events within their encounters produce strain or weaken employees’ performance of
workgroups more accurately. Managers would also benefit from their professional roles. With high levels of attachment anxiety,
recognizing that an employees’ interpretation of social events may employees may be more likely to label a social exchange as uncivil
differ from those of other employees and from those of the manager. and to experience distress in response to such an incident. Alter-
natively, avoidance attachment may allow employees a way to
4.2. Limitations and future research disengage from difficult situations. That is, the superficial nature of
workplace relationships for employees with high levels of avoid-
The study is limited by its reliance on a single source of self- ance may reduce the emotional impact of a social encounter even
reported data from a cross-sectional questionnaire. The concern when it is perceived as uncivil or disrespectful, such that they expe-
with common method bias is alleviated to some extent by a Har- rience less burnout. Future research should address this issue.
mon single factor test that found that only 23% of the variance Finally, a more ambitious research agenda is to explore
across the items in the model could be attributed to a single fac- interventions or training to help managers and team members
tor. Another concern is that the response rate from staff members accommodate a range of attachment styles at work. It seems likely
was low, limiting the extent to which the sample could be con- that employees with attachment anxiety would prefer a different
sidered representative of the participating hospitals or the general kind of supervisory relationship than would employees with secure
population. The low participation rate reflected the study’s focus on or avoidant attachment styles. Managers’ effectiveness could be
34 M.P. Leiter et al. / Burnout Research 2 (2015) 25–35

improved if they could identify employees’ attachment style and Day, A., & Leiter, M. P. (2014). The good and bad of working relationships: Impli-
adapt their supervisory behavior accordingly. cations for burnout. In M. P. Leiter, A. B. Bakker, & C. Maslach (Eds.), Burnout at
work (pp. 56–79). London: Psychology Press.
Deci, E. L., & Ryan, R. M. (1991). A motivational approach to self: Integration in
personality. Nebraska Symposium on Motivation, 38, 237–288.
5. Conclusion de Jonge, J., Spoor, E., Sonnentag, S., Dormann, C., & van den Tooren, M. (2012). “Take
a break?!” Off-job recovery, job demands, and job resources as predictors of
The biases in perception and behavior associated with attach- health, active learning, and creativity. European Journal of Work and Organiza-
tional Psychology, 21(3), 321–348.
ment styles pertain to an important dimension of healthcare work Edmondson, A. (1999). Psychological safety and learning behavior in work teams.
in terms of collaborating with, and providing services to, people. Administrative Science Quarterly, 44, 350–383.
Insecure attachment patterns have the potential for undermining Fraley, R. C. (2002). Attachment stability from infancy to adulthood: Meta-analysis
and dynamic modeling of developmental mechanisms. Personality and Social
employees’ participation in teamwork with other providers and Psychology Review, 6(2), 123–151.
to establish therapeutic relationships with their service recipients. Fraley, R. C., & Waller, N. G. (1998). Adult attachment patterns. A test of the typo-
Based on our sample of healthcare providers, we found prelimi- logical model. In J. A. Simpson, & W. S. Rholes (Eds.), Attachment theory and close
relationships (pp. 77–114). New York: Guilford Press.
nary support for the measure of attachment styles at work. We Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis
found that employees who had high attachment anxiety tended of self-report measures of adult attachment. Journal of Personality and Social
to be more closely involved in work relationships. However, this Psychology, 78, 350–365. http://dx.doi.org/10.1037/0022-3514.78.2.350
Gilson, L. (2006). Trust in healthcare: Theoretical perspectives and research needs.
closeness came at a cost because they also reported more strain
Journal of Health Organization and Management, 20, 359–375.
when interacting with others. Our study provides support to show Griffin, D. W., & Bartholomew, K. (1994). Models of the self and other: Fundamental
that attachment styles have the potential to contribute to mod- dimensions underlying measures of adult attachment. Journal of Personality and
Social Psychology, 67, 430–445.
els of workplace relationships and of occupational distress, and
Gutnick, D., Walter, F., Nijstad, B. A., & De Dreu, C. K. (2012). Creative performance
it provides a solid basis for continued research in this area. The under pressure: An integrative conceptual framework. Organizational Psychology
contribution of attachment styles to a model of burnout based Review, 2(3), 189–207.
on workload and value congruence emphasizes the importance Harms, P. D. (2011). Adult attachment styles in the workplace. Human Resource
Management Review, 21, 285–296.
of considering employees’ understanding of their social context. Hazan, C., & Shaver, P. R. (1990). Love and work: An attachment-theoretical
Many sectors of the contemporary work world require active and perspective. Journal of Personality and Social Psychology, 59(2), 270–280.
sophisticated modes of social interaction in the course of contacts http://dx.doi.org/10.1037/0022-3514.59.2.270
Joplin, J. R. W., Nelson, D. L., & Quick, J. C. (1999). Attachment behavior and health:
with service recipients and colleagues. Models of burnout and work Relationships at work and home. Journal of Organizational Behavior, 20, 783–796.
engagement will be more complete and effective when including Kets De Vries, M. F. R. (1980). Organizational paradoxes. Oxfordshire: Travistock
the range of personal perspectives that employees bring to their Publications Limited.
Lakey, B., & Orehek, E. (2011). Relational regulation theory: A new approach to
social encounters at work. explain the link between perceived social support and mental health. Psycho-
logical Review, 118, 482.
Laschinger, H. K. S., Finegan, J., Shamian, J., & Casier, S. (2000). Organizational trust
Conflict of interest and empowerment in restructured healthcare settings: Effects on staff nurse
commitment. Journal of Nursing Administration, 30(9), 413–425.
Leiter, M. P. (2013). Analyzing and theorizing the dynamics of the workplace incivility
The authors declare that there are no conflicts of interest. crisis. Amsterdam: Springer.
Leiter, M. P., & Day, A. (2013). Straightforward incivility scale manual. Technical Doc-
ument: Centre for Organizational Research, Acadia University.
References Leiter, M. P., Day, A., Laschinger, H. K. S., & Gilin-Oore, D. (2012). Getting better
and staying better: Assessing civility, incivility, distress, and job attitudes one
Andersson, L. M., & Pearson, C. M. (1999). Tit for tat? The spiraling effect of incivility year after a civility intervention. Journal of Occupational Health Psychology, 17,
in the workplace. The Academy of Management Review, 24, 452–471. 425–434.
Aquino, K., & Thau, S. (2009). Workplace victimization: Aggression from the target’s Leiter, M. P., & Laschinger, H. K. S. (2011). Psychological safety, respect, and values:
perspective. Annual Review of Psychology, 60, 717–741. Foundations of a psychologically healthy workplace. In Research Presentation in
Bakker, A. B., & Demerouti, E. (2007). The job demands-resources model: State of the Second World Congress on Positive Psychology Philadelphia, PA.
the art. Journal of Managerial Psychology, 22(3), 309–328. Leiter, M. P., Laschinger, H. K. S., Day, A., & Gilin-Oore, D. (2011). The impact of civility
Bartholomew, K. (1990). Avoidance of intimacy: An attachment perspective. Journal interventions on employee social behavior, distress, and attitudes. Journal of
of Social and Personal Relationships, 7, 147–178. Applied Psychology, 96, 1258–1274.
Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: Leiter, M. P., & Maslach, C. (2004). Areas of worklife: A structured approach to organi-
A test of a four-category model. Journal of Personality and Social Psychology, 61, zational predictors of job burnout. In P. Perrewé, & D. C. Ganster (Eds.), Research
226. in occupational stress and well being: Vol. 3. Emotional and physiological processes
Bentler, P. M., & Chou, C. P. (1987). Practical issues in structural modeling. Sociological and positive intervention strategies (pp. 91–134). Oxford, UK: JAI Press/Elsevier.
Methods and Research, 16, 78–117. Leiter, M. P., Nicholson, R., Patterson, A., & Laschinger, H. K. S. (2012). Incivility,
Berwick, D. M. (2003). Improvement, trust, and the healthcare workforce. Quality burnout, and work engagement. Ciencia & Trabajo, 14, 22–29.
and Safety in Health Care, 12, 448–452. Leiter, M. P., Price, L., & Day, A. (2013). Short work attachment measure manual.
Bohlmeijer, E., Prenger, R., Taal, E., & Cuijpers, P. (2010). The effects of mindfulness- Technical Document: Centre for Organizational Research, Acadia University.
based stress reduction therapy on mental health of adults with a chronic medical Leiter, M. P., & Schaufeli, W. B. (1996). Consistency of the burnout construct across
disease: A meta-analysis. Journal of Psychosomatic Research, 68(6), 539–544. occupations. Anxiety, Stress, & Coping, 9, 229–243.
Bowlby, J. (1969). . Attachment and loss: Attachment (Vol. 1) New York: Basic Books. Leiter, M. P., & Patterson, A. (2014). Burnout and engagement at work: From theory
Bowling, N. A., & Beehr, T. A. (2006). Workplace harassment from the victim’s per- to intervention. In A. Day, E. K. Kelloway, & J. Hurrell (Eds.), Workplace well-being:
spective: A theoretical model and meta-analysis. Journal of Applied Psychology, How to build psychologically healthy workplaces (pp. 197–217). Oxford, UK: Wiley
91, 998. Blackwell.
Collins, N. L., & Read, S. J. (1990). Adult attachment, working models, and relation- Leiter, M. P., & Shaughnessy, K. (2006). The areas of worklife model of burnout: Tests
ship quality in dating couples. Journal of Personality and Social Psychology, 58(4), of mediation relationships. Ergonomia: An International Journal, 28, 327–341.
644–663. Lim, S., & Lee, A. (2011). Work and nonwork outcomes of workplace incivility:
Cook, J., & Wall, T. (1980). New work attitude measures of trust, organizational com- Does family support help? Journal of Occupational Health Psychology, 16, 95–111.
mitment and personal need non-fulfillment. Journal of Occupational Psychology, http://dx.doi.org/10.1037/a002172
53, 39–52. Little, T. D., Cunningham, W. A., Shahar, G., & Widaman, K. F. (2002). To parcel or
Cooper, M. L., Shaver, P., & Collins, N. (1998). Attachment styles, emotion regulation, not to parcel: Exploring the question, weighing the merits. Structural Equation
and adjustment in adolescence. Journal of Personality and Social Psychology, 74, Modeling, 9(2), 151–173.
1380–1397. Littman-Ovadia, H., Oren, L., & Lavy, S. (2013). Attachment and autonomy in the
Cortina, L. M. (2008). Unseen injustice: Incivility as modern discrimi- workplace: New insights. Journal of Career Assessment, 21, 502–518.
nation in organizations. Academy of Management Review, 33, 55–75. Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and adult-
http://dx.doi.org/10.5465/AMR.2008.27745097 hood: A move to the level of representation. Monographs of The Society For
Cortina, L. M., Magley, V. J., Williams, J. H., & Langhout, R. D. (2001). Incivility in the Research In Child Development, 66–104.
workplace: Incidence and impact. Journal of Occupational Health Psychology, 6, Maslach, C., & Jackson, S. E. (1981). The measurement of experienced burnout. Journal
64–80. of Organizational Behavior, 2, 99–113.
 M.P. Leiter et al. / Burnout Research 2 (2015) 25–35 35

Maslach, C., Jackson, S. E., & Leiter, M. P. (1996). Maslach burnout inventory manual Ronen, S., & Mikulincer, M. (2009). Attachment orientations and job burnout: The
(3rd ed.). Palo Alto, CA: Consulting Psychologists Press. mediating roles of team cohesion and organizational fairness. Journal of Social
Mikulincer, M., & Florian, V. (1995). Appraisal of and coping with a real-life stressful and Personal Relationships, 26(4), 549–567.
situation: The contribution of attachment styles. Personality and Social Psychol- Ronen, S., & Mikulincer, M. (2012). Predicting employees’ satisfaction and burnout
ogy Bulletin, 21, 406–414. from managers’ attachment and caregiving orientations. European Journal of
Miller, R. S. (2007). Intimate relationships. New York: McGraw-Hill. Work and Organizational Psychology, 21, 828–849.
Morris, J. A., & Feldman, D. C. (1996). The dimensions, antecedents, and consequences Satorra, A., & Bentler, P. M. (1988). Scaling corrections for chi square statistics in covari-
of emotional labor. Academy of Management Review, 21, 986–1010. ance structure analysis. Los Angeles, CA: (UCLA Statistics Series 2). University of
Osatuke, K., Mohr, D., Ward, C., Moore, S. C., Dyrenforth, S., & Belton, L. (2009). Civil- California.
ity, respect, engagement in the workforce (CREW): Nationwide organization Savickas, M. L. (2001). Toward a comprehensive theory of career development:
development intervention at veterans health administration. Journal of Applied Dispositions, concerns, and narratives. Contemporary Models in Vocational Psy-
Behavioral Science, 45, 384–410. chology, 295–320.
Pearsall, M. J., & Ellis, A. P. (2011). Thick as thieves: The effects of ethical orien- Shragai, N. (2014). Why we see bosses as parents. Financial Times,. Retrieved from:
tation and psychological safety on unethical team behavior. Journal of Applied http://www.ft.com/intl/cms/s/2/9cebd9d0-9eda-11e3-8663-00144feab7de.
Psychology, 96, 401. html#axzz2vxEddGVr
Pearson, C., & Porath, C. (2009). The cost of bad behavior: How incivility is damaging Sonnentag, S., Binnewies, C., & Mojza, E. J. (2010). Staying well and engaged when
your business and what to do about it. New York, NY: Penguin Books. demands are high: The role of psychological detachment. Journal of Applied
Penney, L. M., & Spector, P. E. (2005). Job stress, incivility, and counterproductive Psychology, 95(5), 965.
work behavior (CWB): The moderating role of negative affectivity. Journal of Tidwell, M., Reis, H., & Shaver, P. (1996). Attachment, attractiveness, and social
Organizational Behavior, 26, 777–796. interaction: A diary study. Journal of Personality and Social Psychology, 71,
Rapee, R. M., & Heimberg, R. G. (1997). A cognitive–behavioral model of anxiety in 729–745.
social phobia. Behavior Research and Therapy, 35, 741–756. Wei, M., Russell, D. W., Mallinckrodt, B., & Vogel, D. L. (2007). The Experiences in Close
Rom, E., & Mikulincer, M. (2003). Attachment theory and group processes: The Relationship Scale (ECR)-short form: Reliability, validity, and factor structure.
association between attachment style and group-related representations, goals, Journal of Personality Assessment, 88(2), 187–204.
memories, and functioning. Journal of Personality and Social Psychology, 84, Zapf, D., & Einarsen, S. (2011). Individual antecedents of bullying: Victims and per-
1220–1235. http://dx.doi.org/10.1037/0022-3514.84.6.1220 petrators. Bullying and harassment in the workplace. Developments in Theory,
Ronen, S., & Baldwin, M. W. (2010). Hypersensitivity to social rejection and perceived Research and Practice, 177–200.
stress as mediators between attachment anxiety and future burnout: A prospec-
tive analysis. Applied Psychology, 59, 380–403.