The Relationship Between Organizational Culture and Work Engagement: A Multilevel Investigation
The Relationship Between Organizational Culture and Work Engagement: A Multilevel Investigation
The Relationship Between Organizational Culture and Work Engagement: A Multilevel Investigation
A Multilevel Investigation
Anja Krog
Department of Psychology
University of Oslo
May 2014
Acknowledgements
First, I want to thank my supervisor Sabine Raeder, Associate Professor at the
Department of Psychology at the University of Oslo. Raeder introduced me to her research
project, and helped me develop the framework for my thesis. Also, I want to thank Raeder for
constructive feedback and guidance in the writing of this thesis. The time and energy she has
devoted towards me and this paper has been of significant help throughout this process.
I would also like to extend my gratitude to all the companies who participated in this
project. Thank you for the time and effort you put into helping with the data collection. A
special thanks also to all individual participants who kindly took the time to answer the
questionnaire.
I also want to thank my fellow students Katrine Håkensen and Felix Anker Klein for
great teamwork throughout the months of writing our papers. Cooperating with you has been
greatly rewarding and motivating.
Last I want to thank my family for their support and encouragement. A special thank
to my sister, for always challenging me and pushing me to do my best.
Anja Krog
Table of Contents
Abstract ......................................................................................................................... 1
Introduction .................................................................................................................. 2
Theory............................................................................................................................ 3
Work engagement....................................................................................................... 3
Drivers of work engagement .................................................................................. 4
Organizational culture ................................................................................................ 6
The competing values framework ........................................................................... 7
Culture's relationship with employee attitudes and behavior ................................ 9
The culture-engagement link .................................................................................... 10
Method......................................................................................................................... 13
Sample and procedure .............................................................................................. 13
Measures................................................................................................................... 14
Statistical analyses.................................................................................................... 15
Results.......................................................................................................................... 17
Descriptive statistics ................................................................................................. 17
Multilevel analysis.................................................................................................... 19
Discussion .................................................................................................................... 23
Limitations and future research ................................................................................ 28
Implications for practice........................................................................................... 30
Conclusion ................................................................................................................... 30
References ................................................................................................................... 32
1
Abstract
Introduction
In the context of work, human capital is the resource that is involved in making the
important decisions, developing the innovative products, building customer relationships, and
winning contracts. Being a highly dynamic and non replicable resource, most managers
would agree that employees can make a critical difference when it comes to innovation,
competitive power, delivering high quality service, and ultimately effect the overall success of
a business (Bakker & Schaufeli, 2008; Goodman, Zammuto, & Gifford, 2001). As such,
organizations are depending on having employees who take initiative, are creative, proactive,
and willing to go the extra mile. This requires having vigorous employees who are dedicated
and absorbed by their work; that is, employees who are engaged (Bakker & Schaufeli, 2008).
To foster and facilitate employee engagement, organizations depend on knowledge of
what drives work engagement. While job resources are considered the most important
predictors of work engagement (Bakker, Demerouti, & Sanz-Vergel, 2014), few studies have
considered a broader contextual factor measured at a higher organizational level in the
prediction of work engagement (Bakker, Albrecht, & Leiter, 2011). This paper takes a closer
look at how the organizational context influences work engagement, with the aim to further
our understanding of the broader organizational factors that impact on work engagement.
Accordingly, this paper will approach organizational culture as a direct higher-level predictor
of work engagement, where the relationship between different culture types and employee
engagement is investigated. More specifically, organizational culture and work engagement
are investigated under the assumption that different culture types have differing effects on the
work engagement of employees.
While the concept of organizational culture has received great attention from scholars,
attention has been confined to investigating performance outcomes at the organizational level
(Denison & Mishra, 1995). Only a moderate amount of empirical investigations have focused
on phenomena such as the relationship between organizational culture and employee level
outcomes (see Denison & Mishra, 1995; Glisson & James, 2002; Hartnell, Ou, & Angelo,
2011; Lok & Crawford, 2001; Sheridan, 1992). Thus, although organizational culture is
believed to influence both organizational, group and individual constructs, research trying to
identify the employee behaviors and attitudes affected by organizational culture is limited.
Similarly, the work engagement research literature calls for investigating broader contextual
factors that impact on work engagement (Bakker et al., 2011). Thus, there are two
complimentary and concurrent research avenues that deserve further attention.
3
This paper considers the four organizational types within the Competing Values
Framework (CVF) forwarded by Quinn and Rohrbaugh (1983), in order to investigate the
cultures’ relationship with work engagement. The CVF is one of the most widely used culture
frameworks, and is applied here because of its simple structure and logic differentiation
between distinct culture types. A multilevel framework is applied in a model where culture is
featured as a contextual variable measured at the organizational level, while work engagement
is featured as an individual level construct measured at the employee level.
In summary, the present study contributes to research and literature in three ways.
First, a new avenue of research is explored where a broader contextual organizational variable
is considered as a direct influencing factor on work engagement. Thus far, this has not been
empirically investigated, and will contribute to moving the study of work engagement to a
higher level in the organizational context. Second, knowledge on organizational culture’s
relationship with employee level outcomes is extended. Third, a multilevel approach is used
to allow for studying the interaction between a variable characterizing individuals and
variables characterizing groups. Thus, this paper contributes to a more appropriate and
powerful approach to the study of a nested structure evident in organizational life.
This paper is divided into four sections: (1) a review of the work engagement
construct, organizational culture literature, and a presentation of theoretically derived
hypotheses (2) a description of the methodological procedure and statistical analyses, (3) a
review of the analytical results, (4) a discussion of results, contributions, study limitations,
recommendations for future research, and implications for practice.
Theory
Work engagement
In the last decades, research has witnessed a focus shift from the negative to the
positive pole of workers’ well being (Schaufeli, 2004; Seligman & Csikszentmihalyi, 2000).
Within this trend the concept of work engagement has emerged as an antipode of burnout,
considered to be “a positive, fulfilling, work-related state of mind that is characterized by
vigor, dedication and absorption” (Schaufeli, Salanova, González-Romá, & Bakker, 2002, p.
74). Among scholars and practitioners an increased interest on the concept of employee
engagement has developed (Bakker, Schaufeli, Leiter, & Taris, 2008; Crawford, LePine, &
Rich, 2010), leading to a sharp increase in scientific studies on engagement (Bakker et al.,
2011).
4
outcomes (Hartnell et al., 2011; Schneider, Ehrhart, & Macey, 2013). A brief background on
drivers of work engagement can contribute to understand how culture as a broader contextual
organizational factor can influence employees’ engagement.
Recent studies have shown that work related resources are the most consistent and
powerful drivers of work engagement (Bakker et al., 2011; Bakker et al., 2008). Job
resources are situational factors, and refer to the physical, psychological, social or
organizational aspects of the job that in some way are functional to the employee (Bakker &
Demerouti, 2007). Examples of job resources are autonomy, innovativeness, appreciation,
coaching, feedback, and personal and professional development (Bakker & Demerouti, 2007;
Crawford et al., 2010; Demerouti, Bakker, Nachreiner, & Schaufeli, 2001). Job resources are
functional in either achieving work goals, reducing job demands or stimulating growth,
learning and development (Bakker & Demerouti, 2007). As such, they will instigate a
motivational process where both extrinsic and intrinsic motivation is activated (Bakker &
Demerouti, 2008).
Resources have been recognized for their motivational potential in several theories,
such as the Job Characteristics Model (Hackman & Oldham, 1976), the Conservation of
Resources Theory (Hobfoll & Freedy, 1993), and now the job demands-resources model
(Demerouti et al., 2001). The job demands-resources model (JD-R) predicts that job resources
will have a direct positive relationship with work engagement: “the presence of job resources
leads to engagement, whereas their absence evokes a cynical attitude towards work” (Bakker
& Demerouti, 2007, p. 314). Supporting this statement, a meta analysis by Halbesleben
(2010) showed that the commonly studied resources of social support, autonomy, feedback
and a positive organizational climate hold a significant and strong positive relationship with
work engagement. Both self-report and observational analyses have provided support for the
job demands-resources model (Bakker, Demerouti, Hakanen, & Xanthopoulou, 2007;
Demerouti et al., 2001; Schaufeli & Bakker, 2004).
Recently, a few investigations have taken advantage of multilevel frameworks in order
to investigate higher-level precursors to conducive work environments preceding employee
engagement. For example, Dollard and Bakker (2010) found support for the hypothesis that
psychosocial safety climate had a positive effect on work engagement through its positive
relationship with job resources. Tuckey, Bakker, and Dollard (2012) similarly found support
for the hypothesis that the relationship between empowering leadership and follower
engagement was partially mediated by resources. Additionally, a direct relationship was
found between empowering leadership and follower engagement. Noting on the implications
6
of their findings, Tuckey et al. (2012) concluded: “Potential influences on work engagement
also exist within the […] wider organization.” (p. 22).
According to Bakker and Demerouti (2007), job resources are to be found at several
organizational levels. At the individual level, you will find job resources such as skill variety,
task identity or task significance. At the group level, interpersonal and social relations create
job resources such as co-worker support. The management creates job resources through
stimulating participation and sharing information. Last, the organization creates job resources
through job security or career opportunities. Thus, we look at job resources as something that
is external to the employee; a situational factor that is to be found in the broader
organizational context. As job resources are considered significant predictors of work
engagement, understanding what explanatory factor accounts for the variances in resource
presence and availability is fundamental. This way, the variances in level of engagement
expressed by the workforce can be explained for at a higher contextual level. As Hobfoll
(2001) states “Resources are not individually determined, but are both transcultural and
products of any given culture.” (p. 341). Expecting that culture will precede contextual
resources and ultimately working conditions, organizational culture becomes relevant to
consider in the prediction of work engagement. In effect, organizational culture will work as a
purposeful variable in setting the criteria for variances in resource availability between
organizations, thus impacting to what extent employees are engaged at work.
Organizational culture
According to Hobfoll (2001) individuals’ actions and cognitions are guided by the
availability of resources and the constraints of a culture. Thus, behavior and attitudes become
a product of the shared values, beliefs and norms – that is, the organizational culture of a firm
(Deshpandé & Webster, 1989; Hobfoll, 2001). Looking at organizational culture as a
contextual variable, some organizations provide their employees with more resources than
others, depending on the characteristics or type of the existent culture. To illustrate this:
engagement research considers support, appreciation and opportunities for being innovative to
be job resources (see Bakker et al., 2007; Crawford et al., 2010). As such, when an
organizational culture is perceived as supportive, appreciative, and innovative, in essence the
organizational culture is offering a fair amount of resources. Employees will likely interpret
the work environment accordingly. In turn, they are more likely to respond by dedicating their
effort and abilities, investing time and energy, and be psychologically connected and involved
in their work and organization (Bakker et al., 2011; Bakker & Demerouti, 2007). Ultimately,
the organizational culture of a firm can by itself have the ability of foster the work
7
the axis is oriented towards differentiation and rivalry (Denison & Spreitzer, 1991).
According to the competing values framework, cultures of organizations differ with respect to
the two sets of opposite values of focus and structure, ranging from internal to external, and
from flexibility to stability respectively (Cameron & Quinn, 2006). Combining these two sets
of competing values result in four unique culture types as depicted in Figure 1.
With an internal focus and emphasis on control, hierarchy cultures are to be found in
workplaces where formalized and structured procedures set the guideline for what people do
(Cameron & Quinn, 2006). In hierarchy cultures, rules and policies are said to hold the
organization together, so that stability and predictability will foster efficiency (Hartnell et al.,
2011). Clan cultures emphasize flexibility with an internal focus. Teamwork and employee
development, as well as participation and empowerment of employees hold the organization
together, while concern for people is part of defining success (Quinn & Kimberly, 1984).
Adhocracy cultures have an external focus, supported by a flexible organizational structure. In
adhocracy cultures emphasis is put on individuality and innovation, so that the organizations
becomes a dynamic and creative place to work (Cameron & Quinn, 2006). Last, market
cultures emphasize stability and control with an external focus. Market cultures are result-
oriented, and hold the organization together by focusing on productivity, profits, and on
winning (Quinn & Kimberly, 1984).
Figure 1. The competing values framework (adapted from Figure 3.1 in Cameron &
Quinn, 2006, p. 35).
9
The four culture types can be clearly distinguished through their placement in the CVF
framework. While organizations usually inhabit characteristics of several culture types,
Cameron and Quinn (2006) argue that most organizations develop a dominant culture style;
meaning that an organization usually can be described through one of the four culture types.
Culture’s relationship with employee attitudes and behaviors. Organizational
culture has been investigated in a variety of different settings in an effort to explore its many
influences on organizational life. While the relationship between organizational culture and
organizational effectiveness has been difficult to establish (Schneider et al., 2013), only a
moderate amount of investigations have turned the focus to culture’s relationship with
employee attitudes and behavior (Odom, Boxx, & Dunn, 1990). As human capital is the most
dynamic and invaluable resource of a firm, understanding what employee behaviors and
attitudes are affected by the organizational environment is critical.
Employee perceptions of the organizational environment has been linked to several
constructs. One of them is job involvement. Job involvement, similar to work engagement,
characterizes an employee who is intrinsically motivated, and who is committed to their work
(Hallberg & Schaufeli, 2006). In their investigation, Brown and Leigh (1996) found that a
work environment perceived to be safe, supportive, and meaningful was related to greater job
involvement, exerted work effort, and commitment of time and energy towards the
organization. When employees were given support, it facilitated their experienced
meaningfulness, roles became more clear, and work was perceived challenging (Brown &
Leigh, 1996). Thus, the contextual surroundings at work seem to have a significant effect on
employees’ attitudes and in what way they are connected to their work.
Other researchers have similarly investigated constructs related to employees’
experiences of the cultural context. In their meta analysis, Hartnell et al. (2011) found that
organizational culture was significantly related to job satisfaction. The relationship depended
on type of culture, with clan cultures having a significantly stronger relationship with job
satisfaction then other culture types. In addition organizational culture was related to
organizational commitment, again with clan cultures having the strongest positive relationship
(Hartnell et al., 2011). These findings were also supported by Odom et al. (1990) when
concluding that higher levels of commitment, satisfaction and cohesion amongst employees is
attained by providing a cultural context that is people oriented and supportive, such as clan
cultures. Similarly, Goodman et al. (2001) found that clan culture values were positively
related to organizational commitment, job involvement, empowerment and job satisfaction.
Clan cultures have also been found to have a significant positive relationship with affective
10
between engagement and characteristics of clan culture have been found, such as
appreciation, supervisor and coworker support, and information. For example, Bakker et al.
(2007) found that the job resources of supervisor support, appreciation and information
sharing had significant positive relationships with the three dimensions of work engagement.
Similarly, Crawford et al. (2010) concluded that the combined effects of several job
resources, including access to information, and support from coworkers, supervisors and the
organization, had significant positive relationships with work engagement. Ultimately, the
focus in clan cultures is to provide job resources to employees, as the presence and
availability of job resources is clearly evident in their values, focus and orientation. In effect,
the clan culture inherently provides employees with several important job resources, which in
turn will facilitate work engagement. In consequence, I propose:
contingent only on achievement may limit personal development, and lead to a lack of proper
feedback, whereas the expression of aggressiveness may be experienced as emotionally
demanding to the employee. Interpreting the characteristics of market cultures, the orientation
towards aggressively competing may lead employees in market cultures to experience high
work pressure, which in turn can be psychologically and physically demanding. Ultimately,
market cultures provide employees only with a limited set of positive and functional job
resources. With a culture predominantly focused on winning through stimulating employees’
competitiveness and aggressiveness, the presence and availability of job resources is
restricted, and consequently market cultures will not be conducive of work engagement.
Therefore, I hypothesize:
Method
Sample and procedure
Data for this study were gathered and used in a team of three master students, all under
the supervision of an associate professor. The sample of organizations was selected on the
basis of compiling a sufficiently heterogeneous sample regarding the independent variable,
size and industry. Two questionnaires were created. One questionnaire was to be completed
by an HR-representative and one to be completed by a group of employees in the
organization. As such, organizational culture was assessed on the organizational level while
work engagement was assessed on the employee level.
After a first introductory email, HR-representatives and managers from several small
and large companies received an information letter explaining the background and purpose of
the studies, as well as information about confidentiality, anonymity and participant rights.
Following, an email with URL-links to online questionnaires was sent. The two
questionnaires were sent out to each participating company with appropriate instructions.
Each participating company was asked to recruit a minimum of 10 employees, and a
maximum of 50 employees to fill in the employee-questionnaire. Completion time for the
questionnaires was approximately 15-20 minutes. Two weeks after sending out
questionnaires, reminder emails were sent.
In total, 183 companies were contacted. 35 companies accepted the invitation to
participate, giving a response rate of 19 %. Of the 183 companies who were contacted, 84 did
not reply. Out of those who replied, 55 % accepted the invitation to participate. A total of 35
14
innovation, or risk taking”), market culture (e.g. The management style in the organization is
characterized by hard-driving competitiveness, high demands, and achievement”), and
hierarchy culture (e.g. “The glue that holds the organization together is formal rules and
policies. Maintaining a smooth-running organization is important”), each comprised of six
items. The OCAI is scored by instructing respondents to judge the degree of fit between their
organization and each of the in total 24 items. Respondents are asked to do this by dividing
100 points among the four items presented in each of the six dimensions, depending on the
extent to which the item is representative of their own organization. The item that is most
similar to the respondents’ organization is given a higher number of points, while the
alternative that is the least familiar to the respondents’ organization is given a lower number
of points. As such, the OCAI requires simple arithmetic calculations on behalf of the
respondents. Scores on the six items representing a specific culture were then added up across
the six dimensions, such that each organization received a summated score on each of the four
culture scales. Internal consistencies (Cronbach’s alpha) for the clan culture scale was .82, for
the adhocracy culture scale .75, for the market culture scale .87, and for the hierarchy culture
scale .84.
Control variables included in the analyses are the employee’s gender (male/female)
and tenure within the particular organization (years), measured at the employee level.
Additionally, organizational size (number of employees) as measured on the organizational
level was included as a control variable.
Statistical analyses
The data obtained from the questionnaires resulted in two separate datasets. All
datasets were exported directly from the electronic survey. One dataset consisted of data from
the questionnaire version filled in by an HR-representative, and the other dataset consisted of
data from the questionnaire version filled in by the employees. The datasets were merged into
one file, and sorted by a coded firm number in order to conduct the analysis on groups of
organizations. All items included in the surveys were mandatory, such that missing values
were not a problem. Even so, the survey did not control for invalid input. Both datasets were
therefore thoroughly screened.
The Level 1 dataset contained three cases with invalid input on the control variable
“tenure”, and eleven cases with invalid input on the control variable “age”. Invalid data input
was replaced using the Expectation-Maximization method (EM). First, Little’s Missing
Completely at Random (MCAR) test was performed in order to test the hypothesis of whether
the values were missing at random. As the test turned out to be non significant for both
16
variables, the missing values were replaced using the EM method. The Level 2 dataset was
also screened for errors. The Level 2 dataset contained some typing errors due to the
particular response format of the OCAI. In very few cases the total score on the culture
dimensions did not add up to 100. In one extreme case, the scores on the six culture
dimensions added up to values between 300 and 400. In this particular case, scores were
adjusted in a logical sequence that considered the ratio of the relationship between the total
score on the dimension (e.g. 300) and the expected total score on the dimension (i.e. 100).
The datasets were further processed with SPSS 20. Items were renamed and scales were
computed. Means and standard deviations were calculated for key variables.
Before conducting the analyses to test the hypothesized relationships, all predictors
were centered. Centering of the predictors increases the interpretability of the intercept in the
models (Hox, 2010). Grand mean centering was applied, and executed by subtracting the
grand mean of each independent variable from the values of each independent variable. As
such, the four independent variables all had a mean around zero.
In order to test the relationship between organizational culture and work engagement,
multilevel modeling was chosen as the appropriate method. Multilevel modeling is
conceptualized as a series of equations considering different levels of data, and allows for
analyzing nested data structure (Hox, 2010). In the present study the data had two levels:
employees (Level 1) nested in organizations (Level 2), where work engagement was
measured on Level 1 while organizational culture was measured on Level 2. As such, data in
this study included a multistage sample from a hierarchical system, where variables that
characterize the group and variables that characterize the individual were considered. Using
multilevel modeling decreases the chances of making a Type I error as the method is able to
take clustered data into consideration (Hox, 2010). Additionally, common methods bias is
avoided by using multilevel modeling.
The multilevel analysis was conducted in a sequence of four steps for each of the three
dependent variables, following the bottom-up procedure recommended by Hox (2010). In the
first step, a model without predictors was computed in order to retrieve the intercept-only
model. In the second step, employee tenure, employee gender and organizational size were
computed in order to examine the control variables’ effects on the dependent variable.
Considering the high intercorrelations between the independent variables (see Table 2), the
four culture types were computed in two separate steps. In the third step, the independent
variables of clan culture and adhocracy culture were entered. In the fourth step, the
independent variables of hierarchy culture and market culture were entered. In each step, the
17
estimates and standard error, significance of parameters, and residual error of distinct levels
were inspected. Computing the variables in the sequence described above is advantageous for
calculating the proportion of explained variance resulting from adding variables at the
different levels. The explained variance is calculated for both the employee level and for the
organizational level, as well as for the total.
In the present study, three measures of work engagement were used as dependent
variables, including the dimensions of vigor, dedication and absorption. Four independent
variables were used; clan culture, adhocracy culture, hierarchy culture and market culture.
Additionally, gender, tenure and organizational size were included as control variables.
Results
Descriptive statistics
Tables1 and 2 present descriptive statistics for employee level variables (Level 1) and
organizational level variables (Level 2). Descriptive statistics are presented separately for the
Level 1 and Level 2 variables in order to avoid aggregation and disaggregation of the
variables. Mean values (M) and standard deviations (SD) are included, as well as correlations
between variables, and estimates of internal consistency (Cronbach’s alpha).
Internal consistencies (Cronbach’s alpha) for the engagement scales were .91, .91 and
.90 for vigor, dedication and absorption respectively (Table 1). The culture scales had
reliabilities ranging from .75 to .87 (Table 2). According to DeVellis (2003), reliabilities
within the range of .70 and .80 are considered respectable, while reliabilities within the range
of .80 and .90 are considered very good. Accordingly, all engagement scales have reliabilities
that are considered very good. The clan culture scale, hierarchy culture scale, and market
culture scale with alphas of .82, .84, and .87 respectively, are also considered very good. Last,
the adhocracy culture scale with an alpha of .75 is considered respectable. All engagement
scales had high means, indicating that participants gave high scores on all variables of work
engagement (Table 1). The clan-, adhocracy-, hierarchy- and market culture scale all had a
mean close to 25 (Table 2). According to Cameron and Quinn (2006), an even score among
the four culture types indicates that organizations emphasize the four culture types nearly
equally. In some cases, intercorrelations between independent variables were high (Table 2).
Especially clan culture and hierarchy culture (r = -.72, p < .01), as well as market culture and
adhocracy culture were highly correlated (r = -.52, p < .01). As there is no multicollinearity
test in multilevel analysis, two separate models were computed when analyzing the effects of
the independent variables (as described above) to deal with potential multicollinearity.
18
Table 1
Descriptive Statistics, Internal Consistency Reliabilities, and Correlations for Level 1 Variables (Employees)
Variable M SD 1 2 3 4 5
Control variables
1. Gender (1 = female, 2 = male) - - -
2. Tenure 5.26 6.41 -.07 -
Employee level (Level 1)
3. Vigor 5.50 1.17 -.04 .06 (.91)
4. Dedication 5.59 1.26 -.06 .07 .82** (.91)
5. Absorption 5.31 1.25 -.10* .06 .84** .83** (.90)
Note: N = 463. Scale reliability (Cronbach's Alpha) in brackets on the diagonal.
* p < .05. ** p < .01.
Table 2
Descriptive Statistics, Internal Consistency Reliabilities, and Correlations for Level 2 Variables (Organization)
Variable M SD 1 2 3 4 5
Control variable
1. Organizational size 820.60 3375.23 -
Organizational level (Level 2)
2. Clan culture 33.78 12.37 -.13 (.82)
3. Adhocracy culture 19.25 8.23 -.19 .14 (.75)
4. Hierarchy Culture 25.38 13.84 .05 -.72** -.30 (.84)
5. Market culture 21.59 11.13 .22 -.32 -.52** -.23 (.87)
Note: N = 35. Scale reliability (Cronbach's Alpha) in brackets on the diagonal.
** p < .01.
19
Multilevel analysis
In tables 3, 4 and 5 the results of the multilevel analysis are shown. Table 3 presents
the results of models including vigor as the dependent variable. In Table 4 the results of the
different culture types’ effects on dedication are shown. Last, Table 5 presents the models
computed for the engagement dimension of absorption. Standardized estimates for the
independent variables are presented in the last column of the tables for the purpose of
facilitating proper interpretations when comparing effects of the independent variables in the
sample. In order to compare the fit of the models presented, Akaike’s Information Criterion
(AIC) was used as a general fit index. According to Hox (2010), a decrease in the AIC from
one model to the next is indicative of better fit.
Hypothesis 1 predicted that the clan culture type would have a positive relationship
with work engagement. In line with the hypothesis, the clan culture type was positively
related to all of the three dimensions of work engagement. Clan culture showed a significant
relationship with dedication (ß = .16, p < .10), but none of the estimates in the cases of vigor
(ß = .14, p > .10) and absorption (ß = .11, p > .10) were significant. Overall, the results partly
support Hypothesis 1.
Hypothesis 2 investigated whether the adhocracy culture type would have a positive
relationship with work engagement. Results show that the adhocracy culture type did not have
a positive relationship with vigor (ß = -.11, p > .10), dedication (ß = -.03, p > .10) or
absorption (ß = .00, p > .10), nor where they significant. As there were no significant effects
of adhocracy culture on any of the dimensions of work engagement, Hypothesis 2 was not
supported by the data.
Hypothesis 3 stated that the hierarchy culture type would have a negative relationship
with work engagement. In line with the hypothesis, hierarchy culture type was negatively
related to all of the three dimensions of work engagement, including vigor (ß = -.03, p > .10),
dedication (ß = -.10, p > .10), and absorption (ß = -.11, p > .10). However, estimates of
hierarchy culture’s effects on three of the work engagement dimensions were not significant.
Consequently, Hypothesis 3 was not supported by the data.
Hypothesis 4 proposed that the market culture type would have a negative relationship
with work engagement. The market culture type was negatively related to vigor (ß = -.04, p >
.10), dedication (ß = -.08, p > .10), and absorption (ß = -.05, p > .10). Although relationships
were negative as hypothesized, the estimates showed no significant relationships with any of
the three dimensions of work engagement. Thus, no support for Hypothesis 4 was found.
20
Table 3
Multilevel Analysis Predicting Vigor
Model 1 Model 2 Model 3 Model 4 Standardized estimates
Fixed Effects Est. (SE) Est. (SE) Est. (SE) Est. (SE) for predictors
Intercept 5.64 (.10)*** 5.57 (.20)*** 5.56 (.20)*** 5.57 (.20)***
Control variables
Tenure .01 (.01) .01 (.01) 01 (.01)
Gender -.01 (.11) -.00 (.11) -.01 (.11)
Organizational size .00 (.00) .00 (.00) -.00 (.11)†
Predictors
Clan culture .01 (.01) .14
Adhocracy culture -.02 (.01) -.11
Hierarchy culture -.00 (.01) -.03
Market culture -.00 (.01) -.04
Random Effects Est. (SE) Est. (SE) Est. (SE) Est. (SE)
Variation between employees 1.26 (.09)*** 1.26 (.08)*** 1.25 (.08)*** 1.25 (.08)***
Variation between organizations .13 (.07)* .11 (.06)* .11 (.06)* .12 (.07)*
AIC 1447 1450 1450 1453
†p<.10. *p<.05. ***p<.001.
21
Table 4
Multilevel Analysis Predicting Dedication
Model 1 Model 2 Model 3 Model 4 Standardized estimates
Fixed Effects Est. (SE) Est. (SE) Est. (SE) Est. (SE) for predictors
Intercept 5.70 (.10)*** 5.72 (.21)*** 5.70 (0.21)*** 5.71 (0.21)***
Control variables
Tenure .01 (.01) .01 (.01) .01 (.01)
Gender -.08 (.12) -.07 (.12) -.08 (.12)
Organizational size 00 (.00) .00 (.00) .00 (.00)
Predictors
Clan culture .02 (.01)† .16†
Adhocracy culture -.00 (.01) -.03
Hierarchy culture -.01 (01) -.10
Market culture -.01 (01) -.08
Random Effects Est. (SE) Est. (SE) Est. (SE) Est. (SE)
Variation between employees 1.49 (.10)*** 1.49 (.10)*** 1.47 (.10)*** 1.48 (.10)***
Variation between organizations .13 (.08)† .12 (.08)† .12 (.07)* .11 (.07)†
AIC 1524 1526 1527 1528
†p<.10. *p<.05. ***p<.001.
22
Table 5
Multilevel Analysis Predicting Absorption
Model 1 Model 2 Model 3 Model 4 Standardized estimates
Fixed Effects Est. (SE) Est. (SE) Est. (SE) Est. (SE) for predictors
Intercept 5.45 (.09)*** 5.62 (.20)*** 5.61 (.20)*** 5.61 (.20)***
Control variables
Tenure .01 (.01) 01 (.01) .01 (.01)
Gender -.20 (.12) -.19 (.12) -.20 (.12)†
Organizational size .00 (.00) .00 (.00)† .00 (.00)†
Predictors
Clan culture .01 (.01) .11
Adhocracy culture -.00 (.01) -.00
Hierarchy culture -.01 (.01) -.11
Market culture -.00 (.00) -.05
Random Effects Est. (SE) Est. (SE) Est. (SE) Est. (SE)
Variation between employees 1.49 (.10)*** 1.48 (.10)*** 1.47 (.10)*** 1.48 (.10)***
Variation between organizations .07 (.05)† .06 (.05)† .07 (.05)† .04 (.05)
AIC 1517 1517 1519 1519
†p<.10. ***p<.001.
23
For all of the three dependent variables, Model 1 showed moderately significant
unexplained variance between organizations, leaving little variance to be explained when
adding the organizational level variable. For vigor, Model 1 showed that the unexplained
variance between organizations was significant (p < .05). The intraclass correlation (ICC), for
vigor showed that 9% of the unexplained variation was at the organizational level. For
dedication, Model 1 showed that the unexplained variance between organizations was
minimally significant (p < .10), and the ICC showed that 8% of the unexplained variation was
at the organizational level. For absorption, Model 1 also showed that the unexplained variance
between organizations was minimally significant (p < .10). The ICC for absorption showed
that 4.5% of the unexplained variance was at the organizational level.
As the predictors were computed in two separate models, there is no basis for
comparing the fit of these two. Even so, the AIC did not decrease compared to the intercept-
only model when introducing the predictors in Model 3 and Model 4. This implies that neither
Model 3 nor Model 4 seem to have a better fit than the intercept-only model (Model 1). As all
predictors were located at Level 2, only proportions of variance explained at Level 2 are
relevant to inspect. For vigor, calculations showed that Model 3 explained 15% of the
variation at Level 2, while Model 4 explained 8% of the variation at Level 2. For dedication,
Model 3 explained 8% of the variation at Level 2, while Model 4 explained 15% of the
variation at Level 2. Last for absorption, calculation showed that Model 3 did not explain any
of the variation at Level 2, while Model 4 explained 43% of the variation at Level 2.
Summarized, Model 3 and Model 4 explained up to 43 % of the variation at Level 2 for all of
the dimensions of work engagement.
Discussion
This investigation focused on a higher-level organizational predictor of work
engagement, in an attempt to extend both work engagement and organizational culture
literature. The work engagement literature contains only a handful of investigations focusing
on higher-level drivers of work engagement. As such, this investigation adds to the line of
inquiry trying to explore how engagement is affected by broader organizational factors.
Similarly, the organizational culture literature contains investigations of only a limited
selection of employee level outcomes. This investigation adds to the line of inquiry trying to
explore in what ways employees are affected by the organizational culture present at their
workplace. Accordingly, by investigating the relationship between specifically organizational
24
culture and work engagement, the aim of the present study was to advance our understanding
of the broader organizational factors that impact on work engagement.
Two sources of information were used in the current investigation: employees and the
organizations they belong to. Considering the hypothesis in light of the results, one hypothesis
was partially supported by the data, while the other hypotheses were not supported by the
data. More specifically, clan culture showed a partially significant positive relationship with
work engagement. In contrast to the hypothesis, adhocracy culture did not show a significant
positive relationship with work engagement. Both hierarchy culture and market culture
showed a negative relationship with work engagement as hypothesized, but none of the
relationships were statistically significant. Overall, the current investigation finds support for
claiming that organizational culture has a limited direct effect on work engagement. Similarly,
the effects of culture on the work engagement of employees seem to minimally differ
depending on the type of organizational culture existent in a firm. The discussion will aim to
shed light over different interpretations and explanations of the results.
Although the results of the current investigation did not support the majority of the
hypothesized relationships, this paper contributes to both the work engagement and
organizational culture literature in several ways. On an overall level, an important
contribution of this investigation is the discussion of a relationship between two constructs
that have previously not been theoretically merged or empirically explored. As such, this
paper provides the first investigation of the relationship between organizational culture and
work engagement. Although a limited number of previous studies have linked higher-level
organizational variables to work engagement (see Dollard & Bakker, 2010; Tuckey et al.,
2012), to date no investigation has discussed the relationship between specifically
organizational culture and work engagement. Bakker et al. (2011) argue that “we need to pay
more attention to the broader contextual organizational factors that impact on engagement”
(p. 23). In line with this argument, the current paper represents one of the first leaps towards
connecting a broader contextual organizational factor with work engagement. While the
current investigation only provide partly significant results for one of the hypotheses, the
theoretical discussion of how culture and engagement potentially are linked forms the basis
for future attempts to connect these constructs.
The current investigation extends theory on facilitators of work engagement, and
contributes to work engagement theory by showing that the clan culture type moderately
predicts work engagement. In line with the partly significant results, previous investigations
also suggest that higher-level variables have the power to influence work engagement (see
25
Dollard & Bakker, 2010; Tuckey et al., 2012). As the majority of the hypotheses in the
current investigation were not supported, it seems relevant to compare the current paper to
previous investigations on work engagement. Accordingly, one of the clear differences lies in
how the hypothesized relationships are studied. Previous investigations on work engagement
have included a mediating variable when examining the relationship between a higher-level
organizational factor and work engagement. Thus, there is reason to believe that there is a
significant relationship between the different culture types and work engagement, but that the
relationships need to be explored through a more complex model that includes a mediator. If
empirically explored as a mediated relationship, a higher-level organizational variable seems
to have a powerful and significant relationship with work engagement.
The current investigation extends theory on cultures effects on individual level
constructs, and contributes to culture theory by showing that the clan culture type and work
engagement are moderately related. While the current investigation only found a partly
significant relationship between clan culture and work engagement, previous investigations
have found both direct and significant relationships between various culture types and
employee attitudes and behaviors (see Goodman et al., 2001; Hartnell et al., 2011; Odom et
al., 1990; Sheridan, 1992). On the one end is the possibility that the relationship between
organizational culture and specifically work engagement is manifested differently than the
relationship between organizational culture and other employee attitudes. On the other end is
the possibility that methodological limitations associated with the current investigation have
obscured results for the non-significant findings. If the first possibility is assumed, the
relationship between organizational culture and work engagement should not be investigated
assuming a direct relationship between the two constructs. If the second possibility is
assumed, limitations associated with the current investigation should guide how future
researchers approach the relationship between organizational culture and work engagement.
As the current investigation contrasts to both previous culture research and
engagement research regarding methodology and empirical assumptions, there are several
paths of explanations that can be discussed when considering why this study only obtained
partly significant results.
First, there is reason to believe that including job resources in the statistical
investigation would be advantageous for attaining a clearer picture of how different culture
types and engagement are related. Acting as a mediating variable, job resources could give
insight into the statistical variation of job resources between the different cultural contexts’.
Job resources have previously been added as a mediating variable when predicting work
26
engagement. For example, Tuckey et al. (2012) examined the role of empowering leadership
applying multilevel analysis on a model where job resources acted as a mediating variable.
Similarly, Dollard and Bakker (2010) applied multilevel thinking when investigating the role
of psychosocial safety climate in a model where job resources acted as a mediating variable.
Similar to the current investigation, the higher-level predictor was assumed to precede
working conditions by impacting to what extent job resources were available in the
organizational context. While Tuckey et al. (2012) obtained measures of the predictor and
outcome at different levels, Dollard and Bakker (2010) obtained measures of the predictor and
outcome at the same level, before aggregating the predictor to a higher level. As such, there
are clear differences in how the relationship between the higher-level variable and work
engagement is studied. Although, Tuckey et al. (2012) also found a direct relationship
between empowering leadership and work engagement, both investigations included job
resources as a mediating variable explaining the significant relationship between the higher-
level predictor and engagement.
Second, there is reason to believe that the relationship between an organization’s
culture and work engagement is more complex then assumed in this paper. According to
Schneider et al. (2013) most research on organizational culture addresses a direct relationship
between culture and some outcome variable, whereas theory usually forwards the assumptions
of more complex relationships between such variables. In line with Schneider et al.’s (2013)
observations, previous empirical investigations of organizational culture have forwarded a
direct link between culture and employee attitudes or behaviors. Thus far, culture has been
linked to constructs such as job involvement and empowerment (Goodman et al., 2001), job
satisfaction (Hartnell et al., 2011), commitment (Odom et al., 1990), and retention (Sheridan,
1992). While research methods in these studies vary, they all have investigated and found
support for a direct link between organizational culture and the employee level outcome.
Thus, several researchers argue for significant relationships between culture and employee
behaviors and attitudes. For example, Odom et al. (1990) argue that an important impact of
organizational culture is its effect on employee level outcomes, such as behavior. Similarly,
Richard et al. (2009) states that previous research indicates that organizational culture impacts
on employee attitudes. While the above-cited studies found significant and direct relationships
for all of the different culture types investigated, this paper only found a partly significant
relationship for one of the culture types. Differing methodologies applied can be one of the
reasons for this incongruence.
27
The main difference between above-cited investigations and the current study is the
application of a multilevel model. Previous studies have mainly applied correlation analyses,
analysis of variance (ANOVA) and multivariate analysis of variance (MANOVA), and
measured organizational culture as how it is experienced by the employee. This paper
forwards a different view; because culture represents something that is shared among
organizational members (Denison et al., 2012; Schneider et al., 2013), the operationalization
of culture must be correct, namely representing something that exists at a higher
organizational level (Chan, 1998; Hartnell et al., 2011). On the one end, by obtaining data for
predictors and outcomes from different sources, i.e. applying multilevel methods, this paper
avoids the fallacy of common method bias (Podsakoff, MacKenzie, & Podsakoff, 2011). On
the other end, this both conceptually and statistically puts the predictor and outcome further
apart, compared to when obtaining data from the same source (Podsakoff et al., 2011). With
reference to the mainly non-significant results, an interpretation could be that culture and
engagement operate contextually too far apart to be studied in a multilevel framework that
assumes a direct link. Following, the relationship between culture and engagement may be
better conceptualized as a rather complex relationship between the organization and the
employee.
Third, the CVF was chosen based on its influence and popularity in the literature, and
because of its usefulness in organizing and distinguishing between distinct culture types
(Cameron & Quinn, 2006; Hartnell et al., 2011). Although the CVF organizes culture
characteristics into meaningful types, results show the culture types are highly intercorrelated,
and exhibit weak relationships with the three dimensions of work engagement. The CVF is
originally an organization theory, stemming from the disciplines of sociology and
anthropology (Denison, 1996), whereas the engagement concept stems from the JD-R theory
located within occupational health psychology (Bakker et al., 2008). A different interpretation
of the results is that the concept of work engagement and the CVF are theoretically too far
apart so that an empirical investigation such as the current will have difficulties in
establishing significant findings. An advantageous alternative could be to consider climate as
a higher-level precursor of work engagement. While the distinction between culture and
climate has been debated for decades, Denison (1996) concluded that the difference between
culture and climate lies in how one interprets the constructs rather that their actual
phenomena. Similarly, Schneider et al. (2013) state that culture and climate are two
alternatives for conceptualizing the way people describe and perceive their work settings,
meaning they are overlapping perspectives. Thus, the constructs of culture and climate can be
28
said to be very much alike. What makes climate more appropriate is that the climate concept,
similar to engagement, stems from the discipline of psychology (Denison, 1996), and should
therefore be theoretically more closely linked. As such, an alternative framework for studying
how the broader organizational context impacts on work engagement could be found in the
climate literature.
Limitations and future research
The findings presented in this paper need to be considered in light of limitations
associated with the design and methodology of the investigation. Possible directions for future
research are suggested.
Concerning the sample, several aspects need to be pointed out. First, the sample size is
considered sufficient but not big (Maas & Hox, 2005). As a rule of thumb, Hox (2010)
suggests to strive for 30 groups with 30 individuals per group, if the main interest lies in fixed
effects. In the current investigation, the complete dataset consisted of 35 cases in the Level 2
sample, and 463 cases in the Level 1 sample. While the Level 2 sample is sufficient, the Level
1 sample includes an average of 13 (SD=37,58) individuals per organization, which is below
what is recommended. Additionally, as the standard deviation indicates, there was an unequal
distribution of employees in the organizations; a consequence of one organization being fairly
large in comparison to the other organizations included in the sample. As such, the sample
was skewed and a lack of heterogeneity was evident. Overall the sample used in this
investigation is neither perfect nor optimal, but is acceptable for the purpose of a master’s
thesis. Future research should strive for ensuring a lager sample size on the organizational
level and a more evenly distributed sample at the employee level. Second, the study was
conducted only on a Norwegian sample. Schneider et al. (2013) point out that national culture
has an impact on organizational culture. Therefore, future research should consider examining
the relationship between organizational culture and work engagement in a different country.
Another limitation is related to the OCAI and the measurement of organizational
culture. First, the particular response format of the OCAI requires that the responder
understands the rating instructions and is able to apply simple arithmetic calculations. In the
current investigation, great effort was put into giving clear instructions to respondents. Still,
in very few cases the total score on the culture dimensions did not add up to 100 points as was
expected. Future research should make efforts to provide the responder with clear and proper
instructions so that response errors are minimized. Second, considering that the responder has
to rate each statement presented in the OCAI, good sense of the organizational culture is
required. In the current investigation, an HR-representative was asked to fill in the OCAI,
29
under the assumption that the HR-department would have good knowledge of the culture
present in a firm. On this note, organizational seniority may be a better premise when
choosing a responder for the culture measures. This because, an individual with low tenure
within an organization may not be familiar with the existent culture, making it difficult to
weight the alternatives presented in the OCAI. The current research setting did not control for
tenure of the employee filling in the culture scales. Future research should consider letting a
long-tenured individual rate the OCAI to ensure that the response reflects the cultural profile
of the organization.
Analyses revealed that the culture scales were highly correlated, especially clan
culture and hierarchy culture, as well as adhocracy culture and market culture. In the case of
multicollinearity, the explained variance is arbitrary distributed between predictors. Thus, the
predictors were computed in two stages when analyzing the theorized model in order to avoid
potential multicollinearity. Computing the predictors in two separate stages makes
comparison of the models problematic. As such, interpretation of results must be considered
in light of the high correlation between the different culture scales. Future research might
avoid this by providing a different research setting and sample. Future research might also
consider using a different culture framework, or in line with the discussion, consider
introducing organizational climate as a higher-level precursor of work engagement.
In the current paper, organizational culture was addressed as a direct influencing factor
on work engagement under the assumption that specific culture types are a precursor of a
conducive work environment that facilitates engagement. However, only a direct link between
organizational culture and work engagement was investigated. A limitation associated with
the current investigation is that employees’ experiences of their work environment were not
included as a source of information. With reference to the discussion, there is reason to
believe that the relationship between organizational culture and work engagement is more
complex than assumed in this paper, where a mediated relationship between culture and
engagement could explain variance between organizations. The relationship between
organizational culture and work engagement needs to be further explored by using possible
mediating variables, such that efforts are made in capturing the complexity of the relationship
between the two constructs.
Last, for moving engagement research in a new direction, broader organizational
variables need to be considered as predictors when exploring how the wider organizational
context impacts on work engagement. Examples of such variables other than culture are
amongst others leadership style, climate, and HRM-practices. When investigating phenomena
30
that include variables operating at different levels of the organization, future research should
ensure to adopt multilevel approaches.
Implications for practice
Although the present study only provided partly significant results for one of the
hypothesized relationships, there are some implications for HR and management practice to
be inferred from the results.
To foster and facilitate employee engagement, organizations depend on knowledge of
what drives work engagement. This is especially relevant for HR and managers, as they often
are responsible for initiatives directed towards creating work environments where employees
are satisfied, dedicated, committed and contributors of organizational success (Goodman et
al., 2001). Yet, in order for such initiatives to be successful in facilitating positive attitudes
and behaviors, knowledge of what predicts these positive attitudes and behaviors is key. The
current investigation suggests that initiatives directed towards increasing work engagement of
employees do not guarantee success when focusing only on broader contextual variables such
as culture. For example, initiatives directed towards adopting or developing a culture that has
more clan-like characteristics, may only contribute minimally to facilitating and fostering
work engagement. Ultimately, there is limited support for the assumption that specific culture
types optimize working conditions for employee engagement. Rather, research to date
indicates that job resources are the most important predictors of work engagement (Bakker et
al., 2014). Thus, initiatives directed towards increasing the availability of resources will be
more effective if HR or managers want to foster and facilitate employee engagement in their
organization.
Conclusion
This study is one of a very few studies to consider a higher-level variable as a
predictor of work engagement; also it is the first to consider the relationship between
specifically organizational culture types and work engagement. The present study was
conducted on 35 Norwegian organizations within a variety of different sectors and industries.
A multilevel approach was applied to test four hypothesized relationships. Results of the
analyses provided partial support for one hypothesis, while three of the hypothesized
relationships were not supported. More specifically, the analyses revealed that the clan culture
type has a moderate relationship with work engagement, while adhocracy culture does not
have a significant positive relationship with work engagement. Similarly, hierarchy culture
and market culture did not have a significant negative relationship with work engagement. A
31
discussion of the results explored work engagement and organizational culture research, in
order review different explanations for the results. Differing assumptions and methodological
differences were especially relevant to point out. Additionally, the results of the current
investigation need to be considered in light of study limitations. The current paper contributes
to both work engagement theory and organizational culture theory, by showing that there is a
moderate link between these two concepts. Future research is needed in order to shed light
over the potential complex relationship between culture and engagement within an
organization.
32
References
Bakker, A. B., Albrecht, S. L., & Leiter, M. P. (2011). Key questions regarding work engagement. European
Bakker, A. B., & Bal, M. P. (2010). Weekly work engagement and performance: A study among starting
10.1348/096317909X402596
Bakker, A. B., & Demerouti, E. (2007). The job demands-resources model: State of the art. Journal of
Bakker, A. B., & Demerouti, E. (2008). Towards a model of work engagement. Career Development
Bakker, A. B., Demerouti, E., Hakanen, J. J., & Xanthopoulou, D. (2007). Job Resources boost work
engagement, particularly when job demands are high. Journal of Educational Psychology, 99, 274-284.
doi: 10.1037/0022-0663.99.2.274
Bakker, A. B., Demerouti, E., & Sanz-Vergel, A. I. (2014). Burnout and work engagement: The JD-R approach.
10.1146/annurev-orgpsych-031413-091235
Bakker, A. B., Demerouti, E., & Verbeke, W. (2004). Using the job demands-resources model to predict brunout
Bakker, A. B., & Schaufeli, W. B. (2008). Positive organizational behavior: Engaged employees in flourishing
Bakker, A. B., Schaufeli, W. B., Leiter, M. P., & Taris, T. W. (2008). Work engagement: An emerging concept
in occupational health psychology. Work & Stress: An International Journal of Work, Health &
Brown, S. P., & Leigh, T. W. (1996). A new look at psychological climate and its relationship to job
involvement, effort, and performance. Journal of Applied Psychology, 81(4), 358-368. doi:
10.1037/0021-9010.81.4.358
Cameron, K. S., & Quinn, R. E. (2006). Diagnosing and Changing Organizational Culture: Based on the
Chan, D. (1998). Functional relations among constructs in the same contant domans at different levels of
analysis: A typology of composition models. Journal of Applied Psychology, 83, 234-246. doi:
10.1037/0021-9010.83.2.234
Crawford, E. R., LePine, J. A., & Rich, B. L. (2010). Linking job demands and resources to employee
engagement and burnout: A theoretical extension and meta-analytic test. Journal of Applied
Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001). The job demands-resources model of
Denison, D. R. (1996). What is the difference between organizational culture and organizational climate? A
native's point of view on a decade of paradigm wars. Academy of Management Review, 21, 619-654.
doi: 10.5465/AMR.1996.9702100310
Denison, D. R., & Mishra, A. K. (1995). Toward a theory of organizational culture and effectiveness.
Denison, D. R., Nieminen, L., & Kotrba, L. (2012). Diagnosing organizational cultures: A conceptual and
empirical review of culture effectiveness surveys. European Journal of Work and Organizational
Denison, D. R., & Spreitzer, G. M. (1991). Organizational culture and organizational development: A competing
values approach. In R. W. Woodman & W. A. Pasmore (Eds.), Research in Organizational Change and
Deshpandé, R., & Webster, F. E. J. (1989). Organizational culture and marketing: Defining the research agenda.
DeVellis, R. F. (2003). Scale Development. Theory and Applications. (2nd ed. Vol. 26). Thousand Oaks, CA:
Dollard, M. F., & Bakker, A. B. (2010). Psychosocial safety climate as a precursor to conducive work
environments, psychological health problems, and employee engagement. Journal of Occupational and
Glisson, C., & James, L. R. (2002). The cross-level effects of culture and climante in human service teams.
Goodman, E. A., Zammuto, R. F., & Gifford, B. D. (2001). The competing value framework: Understanding the
impact of organizational culture on the quality of work life. Organizational Development Journal,
19(3), 58-68.
Hackman, J. R., & Oldham, G. R. (1976). Motivation through the design of work: Test of a theory.
resources, and consequences. In A. B. Bakker & M. P. Leiter (Eds.), Work Engagement: A Handbook of
Hallberg, U. E., & Schaufeli, W. B. (2006). "Same same" but different? Can work engagement be discriminated
from job involvement and organizational commitment? European Psychologist, 11, 119-127. doi:
10.1027/1016-9040.11.2.119
Hartnell, C. A., Ou, A. Y., & Angelo, K. (2011). Organizational culture and organizational effectiveness: A
Hobfoll, S. E. (2001). The influence of culture, community, and the nested-self in the stress process: Advancing
conservation of resources theory. Applied Psychology: An International Review, 50, 337-421. doi:
10.1111/1464-0597.00062
Hobfoll, S. E., & Freedy, J. (1993). Conservation of resources: A general stress theory applied to burnout. In W.
B. Schaufeli, C. Maslach & T. Marek (Eds.), Professional Burnout: Recent Developments in Theory
Hox, J. J. (2010). Multilevel Analysis: Techniques and Applications (2nd ed.). New York: Routledge.
Lok, P., & Crawford, J. (2001). Antecedents of organizational commitment and the mediating role of job
Maas, C. J. M., & Hox, J. J. (2005). Sufficient sample sizes for multilevel modeling. Methodology, 1(3), 86-92.
doi: 10.1027/1614-1881.1.3.86
Odom, R. Y., Boxx, W. R., & Dunn, M. G. (1990). Organizational cultures, commitment, satisfaction, and
Podsakoff, P. M., MacKenzie, S. B., & Podsakoff, N. P. (2011). Sources of method bias in social science
research and recommondations on how to control it. Annual Review of Psychology, 63, 539-569. doi:
10.1146/annurev-psych-120710-100452
35
Quinn, R. E., & Kimberly, J. R. (1984). Paradox, planning, and perseverance: Guidelines for managerial
practice. In J. R. Kimberly & R. E. Quinn (Eds.), Managing Organizational Transition (pp. 295-313).
Quinn, R. E., & Rohrbaugh, J. (1983). A spatial model of effectiveness criteria: Towards a competing values
Richard, O. C., McMillan-Capehart, A., Bhuian, S. N., & Taylor, E. C. (2009). Antecedents and consequences of
psychological contrats: Does organizational culture really matter? Journal of Business Behavior, 62,
Salanova, M., Agut, S., & Peiró, J. M. (2005). Linking organizational resources and work engagement to
employee performance and customer loyalty: The mediation of service climate. Journal of Applied
Schaufeli, W. B. (2004). The future of occupational health psychology. Applied Psychology: An International
Schaufeli, W. B., & Bakker, A. B. (2003). Utrecht Work Engagement Scale (UWES). Test manual. Utrecht
http://www.beanmanaged.com/doc/pdf/arnoldbakker/articles/articles_arnold_bakker_87.pdf
Schaufeli, W. B., & Bakker, A. B. (2004). Job demands, job resources, and their relationship with burnout and
10.1002/job.248
Schaufeli, W. B., Salanova, M., González-Romá, V., & Bakker, A. B. (2002). The measurement of engagement
and burnout: A two sample confirmatory factor analytic approach. Journal of Happiness Studies, 3, 71-
Schneider, B., Ehrhart, M. G., & Macey, W. H. (2013). Organizational climate and culture. Annual Review of
Sheridan, J. E. (1992). Organizational culture and employee retention. Academy of Management Journal, 35,
Tuckey, M. R., Bakker, A. B., & Dollard, M. F. (2012). Empowering leaders optimize working conditions for
engagement: A multilevel study. Journal of Occupational Health Psychology, 17, 15-27. doi:
10.1037/a0025942
Xanthopoulou, D., Bakker, A. B., Demerouti, E., & Schaufeli, W. B. (2009). Work engagement and financial
returns: A diary study on the role of job and personal resources. Journal of Occupational and