Top Management’s Influence on Middle
Management’s Performance During Change
Master thesis, Msc.BA, specialization Change Management
University of Groningen, Faculty of Management and Organization
Pauline Smulders
Student number: s1613480
Jacob van Lennepstraat 17hs
1053 HA Amsterdam
Phone: +31 (0)652665890
E-mail: P.M.J.Smulders@student.rug.nl
Supervisor university: Dr. J. Rupert
Top Management’s Influence on Middle
Management’s Performance During Change
ABSTRACT
In this paper certain top management’s behaviors are proposed to moderate the relationship between uncertainty and role conflict on middle management’s performance during times of change. Because middle managers are both targets as
well as agents of change, their contributions are supposed to be pivotal for the success
of the initiated change. Eighty middle managers participated in this research by
answering an online questionnaire. Contrary to what is hypothesized, no moderating
relationships are found.
Keywords: middle management, change, top management, support, performance, recognition
Acknowledgements:
First of all I would like to thank my parents, who have stood by me and supported me
from afar for these past five years that I have tried to make the best out of my study.
Another special acknowledgement goes to Mr. Cloosterman, CEO of the Academy of
Middle Management, who has put great effort in helping me to get in contact with as
many middle managers as possible. He has been a great and enthusiastic supporter as
of the first moment. I would also like to thank Roderick Kasteel for his help and
support in times of need. And last, but certainly not least, I would like to thank my
supervisors, dr. Rupert and dr. de Poel, as well as my second evaluator dr. Prins for
TABLE OF CONTENTS
1. INTRODUCTION 5
2. THEORETICAL BACKGROUND 8
2.1. Middle management’s performance 8
2.2. Role Conflict 9
2.3. Uncertainty 13
2.4. Top Management’s Support 15
2.5. Role Recognition 18
3. CONCEPTUAL FRAMEWORK 20
4. RESEARCH DESIGN AND METHOD 21
4.1. Sample 21
4.2. Change Events 22
4.3. Measures 23
4.4. Control Variables 24
4.5. Middle Management’s Uncertainty 25
4.6. Role Conflict 25
4.7. Role Recognition 26
4.8. Informational and Instrumental Support 26
4.9. Performance 28 4.10. Data Analyses 28 4.11. Correlations table 31 5. RESULTS 31 5.1. Correlations 32 5.2. Regressions 32
5.4. Multicollinearity 37
6. CONCLUSION AND DISCUSSION 38
7. LIMITATIONS AND FURTHER RESEARCH 45
8. REFERENCES 48
1. INTRODUCTION
To be successful, organizations nowadays are under the constant pressure to improve
their performance. In order to satisfy investors and consumers, and stay ahead of
competition faith is often put into charismatic, and sometimes even heroic senior
managers that lead and push organizations through necessary changes. However, what
is often overlooked is the role that middle managers play in these changes (Mayer & Smith, 2007). And although top management’s leadership is very important, the role of the middle managers is often wrongly understood. Middle managers have been
subject of much criticism as they often seem to have a negative impact on change
(Balogun, 2003). Middle managers can be defined as “those managers holding
positions between first level supervisors and the level of executives, below those who have company-wide responsibilities” (Frohman & Johnson, 1992). And as Likert (1961) once stated, middle managers have traditionally been seen as the linking pins
that supply information upwards and consume those decisions which are being passed
down between the strategic apex and the operating core (Thompson, 1967). Hence, it
would seem that middle managers play an important role in managing and
implementing change processes. However, according to Balogun (2003), middle managers are often portrayed as the ones that are resistant, ‘drag their feet’, or sabotage change processes. On the whole, middle management has received a lot of
Another point of view is that they are also important players in successful and
effective change because they connect an organization’s strategic and operational
levels (Floyd & Wooldridge, 1997). As Floyd & Woolridge (1994) have found,
middle managers synthesize information for top management on issues like
opportunities and threats by using their position and external contacts. They also
facilitate adaptability within the organization by supporting fledgling endeavors
within their own departments, as well as acting as champions for innovative ideas by
using resources. Balogun (2003) furthermore, states that middle managers are crucial
in the creation of new organizational structures and processes, and hence they are the
ones that eventually define the real and realized outcomes of change.
Thus, a sole focus on middle management’s ‘foot dragging’ potential, may obstruct a broader view on their strategic importance. And so, reasons to focus on middle
managers are twofold. Firstly, this group of managers has generally been regarded as
pivotal in the process of change implementation, and because of this, they play a key
role in the ultimate success or failure of major change initiatives (Balogun, 2003;
Giangreco & Peccei, 2005; Floyd & Wooldridge, 1994). A second reason for focusing
on middle managers is that their role in organizational change initiatives comprises
both that of a target, as well as an agent of change (Fenton-O’creevy, 2001; Balogun
& Johnson, 2005). This way, negative feelings towards change on the side of middle
management, can have seriously damaging effects on the performance of middle
managers and consequently the outcomes of the change (Balogun, 2003). Hence, it
seems necessary for top management to acknowledge what these middle managers
can offer (Frohman & Johnson, 1992), because if they do hold an important role in
change, then, given the frequency to which organizations nowadays have to change in
middle managers can add to the process (Floyd & Lane, 2000). Floyd & Lane (2000)
furthermore argue that before accepting a change, middle managers first have to be
able to appreciate the pressures in the organization that create the need to change.
Because if they do not understand that need for change, the middle manager cannot act as an agent for change. This will obstruct middle management’s performance with the consequence that the change may never end up down the line (Balogun, 2003).
And therefore, top management should be receptive to middle management’s negative
feelings towards change, and exert such behavior that positively influences middle management’s performance.
Therefore, this paper’s research question is:
How can certain top management’s behaviors, such as top management’s role recognition and top management’s support, influence middle management’s
performance during change processes?
This research question will be backed up with the following sub-questions:
- How is middle management’s performance influenced by middle management’s uncertainty and middle management’s role conflict during change?
- How do top management’s role recognition and informational support influence the relationship between role conflict and middle management’s performance during
change?
2. THEORETICAL BACKGROUND
The theoretical background will shed light on the different variables which will be
elaborated upon in this paper. The first part will start by elaborating on middle management’s performance, followed by the effects of role conflicts and uncertainty within a change context. In the second part, top management’s behaviors of support
and role recognition will be further explained as well as how the use of these will
affect the negative relationship between uncertainty and role conflict, and
performance.
2.1. Middle management’s performance
Successful change initiatives should comprise good middle management’s performance, because, according to Balogun (2003), unproductive middle management’s performance can have detrimental effects on the outcome of change. In this paper, middle management’s performance is the ability of the middle manager to be both a target, as well as an agent of change (Fenton-O’creevy, 2001; Balogun &
Johnson, 2005). This requires them to be adaptable and flexible (McKenna, 1999) in
both understanding the change, and being able to pass it along (Balogun, 2003). Therefore, middle management’s performance is related to the middle manager’s ability to swiftly understand the reasons for change, and to act accordingly. Meaning
that they should be able to perform both their existing tasks, as well as any new tasks that may accompany the change. Increased middle management’s performance is therefore, also found in the middle manager’s ability to help in, for example, strategy-making (Wai-Kwong et al., 2001) and their capability to facilitate in the emotional
However, these differences in tasks that middle managers are assumed to perform
during change, may influence the role a middle manager has, and may therefore lead
to role conflicts (Floyd & Lane, 2000).
2.2. Role conflict
According to Dopson & Neumann (1998), during major change initiatives, middle
managers are expected to move from more specific roles to more generalist or
managerial roles due to their increasing responsibilities that come with the change.
Simultaneously though, middle management is also responsible for the concrete
implementation of the change (Balogun, 2003; Floyd & Wooldridge, 1994). Hence,
they often experience high pressure levels because of their changing roles as well as
their task to be the agent of change (Floyd & Lane, 2000). And so, because of their
changing experiences and their transitions between roles, role conflict is an often
experienced phenomenon at middle management levels (Currie & Procter, 2005). Role conflicts can be defined as “concurrent appearance of two or more incompatible expectations for the behavior of a person” (Biddle, 1986), and creates distress for the individual experiencing it (Floyd & Lane, 2000). A role is a set of
behaviors that others expect of people in a specific situation (Floyd & Lane, 2000;
Friedman & Podolny, 1992; Katz & Kahn, 1978). Employees may hold a variety of
roles within an organization, that involve a wide spectrum of behaviors, due to
differences in contexts and expectations of the people with whom they interact. If
these differences in expectations are large, role conflicts may occur. Because most
organizations do not have enough slack resources to support parallel activities,
employees are facing role conflicts. The problem herewith lies in the conflict between
encouraging necessary behavior in order to develop new competencies (Floyd &
Lane, 2000).
Balogun (2003) in her case study on the roles of middle managers found that, during a
change implementation, the middle manager performs the complex role of a being a ‘change intermediary’, instead of being only a subject of change or an agent of change. Middle managers not only need to implement the change in their respective
departments, they also need to understand and interpret the change and be able to deal
with it in order to pass it on. In fulfilling this position, it is important that the middle
manager engages in a variety of actions in order to interpret the reasons to change.
These actions include their attempts to change personally, to aid others in going
through the change, in keeping the business going during the change process, as well
as implementing changes within their own departments.
FIGURE 1
Middle managers as change intermediaries
Personal change- Organizational change requires far reaching personal change for middle managers in both their roles and responsibilities, as well as the way in which
they interpret their roles. The activity of interpretation of this role is important. The
managers need to interpret what kinds of implications the intended change will have
for themselves in how they think about and perform their job (Balogun, 2003). Hence,
for middle managers to perform well they should be able to communicate with and
understand top management, their ideas, and why the changes are important, as well
as communicating clearly and convincingly throughout the rest of the organization to
spread the word and get people on board (Huy, 2001).
Aiding others in the change process- The middle manager may also act as a role
model who is responsible for explaining staff about the nature of change in both
formal and informal ways. And so, managers must not only have close relationships
with top management, but also with line staff (Maurer, 2005). Therefore, line staff
should be able to trust the middle manager in order to follow them in the change
process. According to Balogun (2003), the difficulty in performance lies in the fact
that the middle manager should not only be a coach to others, but also a co-worker
and friend, while doing that at the same time as going through their own personal
transitions, and combining different roles.
Keeping the business going- Another role for middle management is keeping the business going while all the changes are taking place. Often staff is expected to
perform the same amount of work, whilst simultaneously implementing and dealing
with the change, as well as giving information on a range of aspects of their work. It is
not uncommon during change to have a change in staff, by means of lay-offs, new
staff members or the integration of different teams. Therefore, keeping the business
challenge for middle management. What is difficult here, is to pay constant attention
to what is important. It is about continuously juggling priorities on where to put
attention and effort into (Balogun, 2003). In order for this to happen, top management
needs to recognize and support the middle manager if they, in times of change, want
to hold on to core values and capabilities while also changing how work gets done and
shifting the organization in new strategic directions (Huy, 2001). In this light, middle management’s performance is dependent upon their ability to keep the focus on the daily work at hand, while simultaneously understanding the new vision an being a
translator between the two (Maurer, 2005).
Implementing changes to departments- Implementing the needed changes often brings along a wide range of activities spread over the whole organization. In which, often
times, responsibility comes down to the middle manager, who has to be innovative
and encouraging in order for the change to take place successfully. Huy (2001) argues
that the middle manager has an exceptional position; they are close to everyday
operations, customers, and frontline employees, and hence know better than anyone
where potential problems may lie. However, they are also faraway enough in order to
see the big picture, that permits them to see new possibilities for solving problems, as
well as for encouraging growth.
During change, these four different roles are expected to conflict when different
requests are made from top managers and employees (Balogun, 2003; Hales, 2005), and when behaviors between the different roles are not in consensus with people’s expectations (Floyd & Lane, 2000) . This is in line with Wooldridge (et al., 2008) who
state that role conflict is often caused by different interpretations of environmental
cues and contradictory expectations among managers at different hierarchical levels.
the differences in the roles middle managers perform. Subsequently, the first
hypothesis is the following:
Hypothesis 1 : In the context of change, role conflicts are negatively related to middle
management’s performance.
Role conflicts however, are not the only negative consequences of change on middle management’s performance. Uncertainty, for example, is a result of role conflicts, where coping mechanisms have to be adopted for individuals to deal with
inconsistencies (Herzig & Jimmieson, 2006). Hence, the next section will further explain middle management’s uncertainty.
2.3. Uncertainty
In today’s ever changing business world, organizations need to change their strategic direction, structure, and staffing levels continuously in order to stay ahead of
competition ( Bordia et al., 2004; Armenakis & Bedeian, 1999; Cascio, 1995). These
changes may lead to a large amount of uncertainty among employees (Bordia et al.,
2004; Callan, 1993; Terry & Jimmieson, 2003). Uncertainty is one of the most
commonly reported psychological states within a context of organizational change
(Bastien, 1987; Buono & Bowditch, 1989; DiFonzo & Bordia, 1998; Terry et al.,
1996; Bordia et al., 2004). This uncertainty may lead to negative consequences for the
performance of middle management, because with uncertainty come feelings of
doubt about the future or about cause and effect relationships in the environment
situation is unpredictable and when information is inconsistent or unavailable”, and it has been recognized as being experienced by employees during change (Bordia et al., 2004). Middle management’s performance is dependent upon the middle manager’s ability to firstly grasp the importance of change (being able to undergo the change
themselves), as well as passing the change down the line (be the implementer of a
change process) (Balogun, 2003). Because middle managers must understand the
change in order to pass it on, any uncertainty on their side must be managed properly
so as not to jeopardize the outcome of the change initiative (Bordia et al., 2004).
Traditionally, decisions relating to change initiatives are made by top management,
while implementation is often carried out by middle management. As a result, those
managers experience cognitive disorder (McKinley & Scherer, 2000), where they are
left with feelings of uncertainty about how to best implement the change (Herzig &
Jimmieson, 2006). Literature (Greco & Roger, 2001) states that uncertainty is a
stressful psychological state that individuals want to reduce as much as possible,
however other researchers believe that not all uncertainty is negative, and that in some
cases it is even preferable because it gives hope of potential positive outcomes instead
of the certainty of negative outcomes (Brashers, 2001). An example could be that
employees might favor uncertainty during restructuring over the certainty of job loss
(Herzig & Jimmieson, 2006). Before implementation of the change initiative, most middle management’s uncertainty stems from their inability to fully comprehend the change. This means that they do not fully understand the goals and the need for
change, as well as the expected profits that can be gained from the change, that causes
the uncertainty. During change, when the change is in progress, most uncertainty
seems to come from the process of change. During this implementation phase, most of
this phase, comes from not knowing the appropriate procedures and processes of how
implementation works; how to implement, what to implement and how to assist
employees in their transition (Herzig & Jimmieson, 2006). Hence, the following
hypothesis:
Hypothesis 2: In the context of change, middle management’s uncertainty about the
change is negatively related to middle management’s performance.
In order to positively influence the effects that role conflicts and uncertainty have on
performance, it has been argued that the provision of top management support is an
important factor (Rhoades & Eisenberger, 2002). Hence, two types of top management’s support are elaborated upon in the next section.
2.4. Top management’s support
During times of organizational change, when employees may have feelings of
uncertainty, are de-motivated and/or scared, it has been argued that top management
should exhibit behaviors that motivate people, that positively influences their behavior
(Judge & Piccolo, 2004; Tyler & De Cremer, 2005), and performance, as well as their
feelings of self-esteem (Kark et al., 2003) and self-confidence (Conger & Kanungo, 1998). Therefore, top management’s supportive behavior is considered important, because employees value the assurance that organizational aid will be available in times of need to perform one’s job effectively and to deal with stressful situations (Rhoades & Eisenberger, 2002). In times of change top management’s supportive behavior provides direction in uncertain situations that is necessary in order to
Jimmieson, 2006). Because of this, this second part will elaborate upon certain top management’s behaviors, namely support and role recognition, which top management can turn to in order to buffer the negative effects that uncertainty and role conflicts may have on middle management’s performance.
To commence with supportive top management’s behavior, Huy (2002), in his study on middle managers, found that, during change initiatives, the greater the support
demonstrated by top management, the greater the commitment to the initiative by
middle management, which led middle management to increase their promotion of the
project. Top management’s support has also proven to be important in reducing middle management’s uncertainty because it facilitates the middle manager to understand the need for change (Bordia et al., 2004; Herzig & Jimmieson, 2006).
Within the social support literature, different types of support have been
distinguished, e.g., instrumental support, informational support, emotional support
and appraisal support (House, 1981). However, in this present paper focus will be put
on informational and instrumental support. Reasons for this are because (a) these two
types of support are easier and more suitable to manage than, for example, appraisal
support (Schat & Kelloway, 2003), and (b), emotional support is, for a great deal,
incorporated in both instrumental and informational support (Barling et al., 1988).
House (1981) associated instrumental support to the provision of needed resources such as time and/or materials. It involves “instrumental behaviors that directly help the person in need” (pp. 24-25). This also incorporates emotional support, since instrumental acts have psychological consequences (House, 1981), and therefore it
may decrease feelings of uncertainty (Herzig & Jimmieson, 2006). Furthermore,
research (Herzig & Jimmieson, 2006) found that during change instrumental support
form of support, because it involves giving people direct help or assistance, which
may lower feelings of uncertainty (Herzig & Jimmieson, 2006). Hence, the negative
relationship between uncertainty and performance suggested earlier, is now proposed
to decrease when top management provides middle management with instrumental
support. Therefore this paper proposes the following hypothesis:
Hypothesis 3: In the context of change, the negative effects of uncertainty on
performance are buffered by top management’s instrumental support.
Informational support, on the other hand, is about providing necessary information or
advice that an individual can use to cope with personal and environmental problems
(House, 1981; Swanson & Power, 2001). This information can be provided both
informally, as well as formally (e.g., training) (House, 1981). It is a more indirect
source of support because it is about providing individuals with resources that they
can use to help themselves, in order for them to perform their tasks and multiple roles
more effectively. Because informational support focuses more on the provision of
indirect resources, and less on psychological consequences (such as uncertainty) than
instrumental support (House, 1981), this paper proposes that these two different kinds
of support moderate different independent variables. Therefore, the earlier suggested
negative relationship between role conflicts and performance is now assumed to be
weakened when top management provides middle management with informational
support. Hence, this paper proposes the following hypothesis:
Hypothesis 4: In the context of change, the negative effects of role conflicts on
In addition to the provision of support, this study also focuses on role recognition, which is argued to positively influence middle management’s performance. Research (Balogun, 2003; Herzig & Jimmieson, 2006) shows that top management’s
recognition of the different roles that middle managers fulfil can have a large
contribution during change. Therefore, role recognition will be further clarified in this
next section.
2.5. Role recognition
In dealing with uncertainty and, in particular, role conflicts, an important top
management behavior is that of recognition of the different roles middle management
has. Middle managers have difficult and challenging tasks that exist of different roles
(Balogun, 2003). However, the role of personal change and aiding others seems to
often go unrecognized by top management. However, it is these two roles that
comprise crucial aspects of the tasks of the middle manager. Here they do not only
make sense of the change intent into concrete actions for themselves, but also for their
teams (Balogun, 2003). The way the middle manager interprets the change affects
how they adopt the change themselves. It will influence the way they will try to
impose the change on their teams, as well as how they will keep the business running
during the change process. Furthermore, the way the change has been interpreted will
also affect what changes the middle manager will implement in their business units
and how they will implement it. Hence, the different middle management roles
Thus, according to literature (Aladwani, 2001; Herzig & Jimmieson, 2006), top management should recognize the importance of middle management’s sense-making of change, because how middle managers interpret the requirements of change, as
well as their actions and initiatives during the change, is a result of their
interpretations and affect the ultimate successfulness of change.
According to Balogun (2003), why middle managers are often not appreciated by
higher management is because the sense-making activity of interpreting the change
intent is often not being recognized as important by top management. Hence, this
activity does not receive the recognition it apparently needs. Therefore, a lack of top management’s recognition for the middle manager’s roles during change may create a work overload. This stems from the middle manager’s constant role juggling, which could potentially expose the middle manager to experiencing role conflicts (Balogun, 2003). This lack of top management’s recognition may lead to role conflicts between their roles of keeping the business going and implementing the changes (which are the
more visible aspects of their work), and the less concrete and intangible interpretation
aspects of their different roles. Moreover, the issue of inconsistent messages from top management also makes the middle manager’s role more difficult, because it increases uncertainty. Hence, Balogun (2003) proposes that middle management’s ‘foot dragging’ may be a reaction to organizational constraints such as lack of top management’s recognition for their different roles, which obstructs them from performing their job. Consequently, the above mentioned negative relationship
between role conflicts and performance is suggested to be weakened by top management’s role recognition. Therefore, the following hypothesis is:
Hypothesis 5: In the context of change, the negative effects of role conflicts on
3. CONCEPTUAL FRAMEWORK
Based on the literature review, two conceptual frameworks are created. In these
models a negative correlation is put forward between uncertainty and role conflict,
and performance. The independent variables propose moderating effects of certain top
management’s behaviors on the negative correlation between the direct effects and the
dependent variable.
FIGURE 2
Conceptual framework of the hypothesized relationships between role conflict and performance, and the moderating effects of informational support and
recognition
FIGURE 3
Conceptual framework of the hypothesized relationships between uncertainty and performance, and the moderating effect of instrumental support
4. RESEARCH DESIGN AND METHOD
This study focuses on whether certain top management’s behaviors have a moderating effect on how well middle management performs during change. From the theoretical
background it becomes clear that support and recognition are likely to affect the
negative relationship between role conflicts and uncertainty, and performance. Based
on what is put forward in the literature, a conceptual model is created with middle
management performance as the dependent variable. As the independent variables the top management’s behavior of role recognition is used, as well as top management’s instrumental and informational support to middle managers. Hence, it is proposed that
if top management supports middle management and recognizes the different roles
middle management can have within a change process, then this will probably affect
the way middle management will interpret the change and behave during the change
process.
4.1. Sample
In this research, a survey is conducted among carefully selected participants who fitted Frohman and Johnson’s (1992) definition of middle managers (described previously). With the help of the Academy of Middle Management over 70 middle managers were personally approached by email, by the Academy’s CEO Mr. Cloosterman, with the request to take part in the current study. Of these 70 managers,
around 30 responded to this email by filling out the online survey, which amounts to
approximately 42,8%. In addition, the Academy of Middle Management also posted
the survey on various (LinkedIn) Middle Management Forums. A letter was enclosed
to the emails sent by the Academy of Middle Management stating the reason for the
the results. However, it is unclear how many middle managers filled out the survey by
way of this method. Next to this, middle managers in the Amsterdam business district
were approached to participate in the survey. This amounted in about useful 30 surveys. Furthermore, an email was sent to the researcher’s personal network with the request to forward the letter to middle managers in their network. Again, it is unclear
how many middle managers were reached by this method. In total 80 out of the 92
surveys collected, within approximately six weeks, were considered useful. This
means that 13% of the surveys have been removed from the sample. Reasons were
either that the respondent was not a middle manager or that the respondent did not fill
out the whole questionnaire. The respondents differ in age, sex and tenure. The
average age of the respondents is 42 years (with a standard deviation of 9,906). 57,5%
of the sample are male, whereas 42,5% are female. Tenure ranges from 3 months to
33 years (with a standard deviation of .497). Furthermore, the respondents come from
eight different nationalities; Dutch ( 82,5%), USA (6,25%), UK (3,75%), Australian
(2,5%), Swiss (1,25%), Brazilian (1,25%), Moroccan (1,25%), Portuguese (1,25%).
4.2. Change events
Survey outcomes are only relevant if (a) the participant is a middle manager and (b) if
the participant has ever undergone an organizational change event, whether it was a
transformational or a transactional change (Kleiner & Corrigan, 1989).
Transformational change is a large-scale change event that affects not only an organization’s working methods or vision, but also the values or even the culture of a company (Kleiner & Corrigan, 1989; Herzig & Jimmieson, 2006). Transactional
change, alternatively, is regarded as being less imposing as it entails more
the expansion of products and services or developmental changes to already existing
systems and processes (Herzig & Jimmieson, 2006). In this study the specific types of
changes experienced by the middle managers are not important, becauseany changes
experienced by the middle managers in this research provide a useful context for studying the predictors and outcomes of middle management’s performance during change (Judge et al., 1999). The use of more than one type of change that affects
different levels of an organization is in line with past change research (Judge et al.,
1999; Caldwell et al, 2004; Herzig & Jimmieson, 2006).
4.3. Measures
Surveys were conducted among middle managers. The survey consists of 36 items
divided among six scales (uncertainty, role conflict, recognition, informational
support, instrumental support and performance). In the survey the middle manager
was asked to decide on a change event they had been involved in, and to answer
accordingly. When deciding on a change event, they were also asked to choose one
that happened rather recently; this in order to avoid a potential reduction in reframing
or remembering the event (Caldwell et al., 2004; Herzig & Jimmieson, 2006).
The first part of the survey contains general questions about the respondents regarding
amongst others age, sex and tenure. The scales used in this survey to measure the
different constructs, have been selected from validated scales from previous studies.
The selection of items from the original studies are based on their factor loadings
within their respective studies as well as on their face validity for the current study.
Every construct consists of six carefully selected items on a 5-point Likert scale, this
is done in order to give respondents a more easily defined choice. All items are
uncertain at all, (1) disagree to (5) agree, and from (1) not at all to (5) to an extreme
extent. All surveys were entered into an online database (www.thesistools.com). The
questionnaire can be found in the appendix.
Furthermore, internal consistency of all items is checked via Cronbach’s alpha, with an acceptable alpha of .60 and above. All items are also checked using ‘alpha if item deleted’, and all alphas, except for informational support remained highest when not removing an item from the scale. The alpha for informational support however,
increased from .64 to .69 when item 3 was deleted from the scale.
4.4. Control variables
Control variables are specific factors that are used by researchers to exclude
alternative explanations for their findings (Schmitt & Klimoski, 1991; Becker, 2005)
or to reduce error terms and increase statistical power (Schwab, 1999; Becker, 2005).
These variables help interpret the relationship between the other variables (Cooper &
Schindler, 2006). Two demographic variables are used as control variables in the
present study so that in the regression analysis the conclusions on performance will
not be confounded by demographics (Teo et al., 1999). The variables in this study are
age and gender. Age is chosen because aging may play a role in portraying how a
person changes over time, and consequently, how this may have an effect on how job
performance changes over time (Waldman & Avelio, 1993). Age has furthermore
been shown to be related to decreased performance on tests of learning, memory and
reasoning (Lindenberger & Baltes, 1994; Verhaeghen & Salthouse, 1997; Sturman,
2003). Also, according to Rogers & Wright (1998) age is one of the most commonly
used control variables when researching performance. Regarding gender, Tsui &
evaluation among middle managers. Some studies (Dipboye et al., 1977; Schein,
1973) state that male middle managers perform better, while other studies (Abramson
et al., 1977; Peters et al., 1983) prove that female middle managers perform better.
Furthermore, a number of studies have used gender as the differentiating variable in
performance in learning patterns as well as the outcomes produced by individuals in
both the academic and the corporate world (Meyers-Levy, 1989; Chung & Monroe,
2001; Graham et al., 2002). For these reasons this study incorporates age and gender
as control variables.
4.5. Middle management’s uncertainty
As is stated previously, uncertainty is considered to be one of the most commonly
reported psychological states within a context of organizational change (Bastien,
1987; Buono & Bowditch, 1989; DiFonzo & Bordia, 1998; Terry et al., 1996; Bordia et al., 2004). In this study, focus is put on middle management’s uncertainty during change, and the items are derived from a 21 item scale in the study of Schweiger and DeNisi (1991). The scale contains the following main question: “as you have thought about your future in the company, to what extent are you uncertain about the following?”. Statements regarding uncertainty include “whether you will have control over your job”, and “whether you will have influence over changes in your job”. The Cronbach’s alpha for this scale is .74.
4.6. Role conflict
The current study proposes that, due to the many different roles middle managers
have, role conflicts may occur due to different expectations in working behavior from
regarding role conflict come from Rizzo, House & Lirtzman’s study on role conflict and ambiguity in complex organizations (1970), in which they used a 30 item scale to
measure role conflicts and role ambiguity. Out of these 30 items, 13 items are related
to role conflict, and of these 13 items 6 items are chosen for this study. Examples of statements are “I have to do things that should be done differently” and “there is a lack of policies and guidelines to help me”. The Cronbach’s alpha for this scale is .85.
4.7. Role recognition
This study proposes that senior management’s recognition for the different middle manager’s roles will influence middle management performance. To test this construct, 6 survey questions are used from Hornsby, Kuratko & Zahra’s (2002) “Middle managers’ perception of the internal environment for corporate entrepreneurship: assessing a measurement scale”. In their paper, the researchers
divided the items over 5 scales. The items which are used in the present paper come from the ‘management support for corporate entrepreneurship’-scale (4 items), and the remaining 2 items come from the 6 item scale of rewards/reinforcements. Items
from those two scales are chosen because they relate most to role recognition as
described in this present paper. Statements include “my organization is quick to use
improved work methods that have been developed by me” and “upper management is
aware and very receptive to my ideas and suggestions”. This scale’s Cronbach’s alpha
is .82.
4.8. Informational and instrumental support
Informational support has been linked with the provision of necessary information or
(House, 1981; Swanson & Power, 2001), while instrumental support focuses on
behaviors that directly help the person in need (House, 1981). The reason to focus on
these two types of support has been elaborated upon in the theoretical background. In
researching this variable, the paper of Huntington (et al., 1989) on perceived
organizational support is used. Of the original 36 item scale, 12 items are chosen in
this current study. Six items for informational support and six items for instrumental support. Statements for these constructs include “help is available from the organization when I have a problem” and “the organization fails to appreciate any extra effort from me”. The Cronbach’s alphas for instrumental and informational support are .86 and .64 respectively. As is mentioned previously, the informational
support alpha increased with .05 when deleting item 3 from the scale. However, when
looking at both the factor analyses with item 3 as well as without item 3, the overall
analysis is less significant for the version without item 3. First of all, without item 3,
there are more overall double loadings (24 in total). Second of all, the informational
support item does not have its own component but loads on the instrumental support
component. Third of all, role recognition loads substantially higher on instrumental
support. And fourth of all, while the scale for informational support is better without
item 3, generally seen, the other items load lower on their specific components. Therefore, although removing item 3 from the scale will increase the Cronbach’s alpha for informational support, it is chosen not to do so because this will lower the
overall factor analysis. Moreover, even with item 3, the scale has an acceptable alpha
4.9. Performance
According to research (Balogun, 2003) unproductive middle management’s performance can have detrimental effects on the outcome of change. Therefore, a successful change initiative should comprise good middle management’s performance. Items are used from Tsui’s (1984) study on managerial role activities. The six most relevant items are picked for this study. In her study, Tsui describes six
managerial roles, which have been tested on 42 items. The roles, as described by Tsui,
that fit middle managers the most are those of a leader and entrepreneur, and
therefore, 4 items have been derived from the 11 item scale on leader roles, and the
remaining 2 items have been derived from the 3 item scale on entrepreneur roles. The question ‘do you perform these role activities’ is asked to answer the managerial role statements. Examples of statements are “forward important information to subordinates” and “allocate human resources to tasks”. These six items have a Cronbach’s α of .79.
4.10. Data analysis
The statistical approach to developing the scales in this research employs a factor
analysis. Factor analysis is a common psychometric technique which is used to
discover a set of factors that represent underlying latent constructs from a larger
number of observed variables (such as items on a survey) (Bundick, 2011).
A factor analysis using a Varimax Rotation with a Kaiser Normalization is performed
to identify underlying factors. The results can be found in table 1. Initially, the
analysis revealed 12 different components, however in a re-run they were forced back
TABLE 1
Factor analysis for all variables Components 1 2 3 4 5 6 Uncertainty Item 1 .424 -.159 -.098 -.348 .338 .258 Item 2 .328 -.115 -.086 .030 .650 .182 Item 3 -.163 .301 -.358 -.151 .499 .129 Item 4 .033 .035 .111 .032 .744 -.025 Item 5 .087 -.013 -.299 -.017 .755 -.124 Item 6 -.149 -.084 -.169 .060 .677 -.071 Role Item 1 -.095 -.380 .639 .011 -.082 -.136 Conflict Item 2 -.248 -.096 .688 -.024 .065 -.169 Item 3 -.454 -.134 .479 -.091 -.296 .034 Item 4 -.187 -.081 .805 -.138 -.119 .054 Item 5 -.188 .052 .768 -.090 -.249 .102 Item 6 -.017 -.021 .788 -.031 -.115 -.144 Informational Item 1 .501 .269 -.263 .225 .069 .151 Support Item 2 .289 .352 -.221 .025 -.185 .313 Item 3 .475 .076 -.142 .091 .081 -.309 Item 4 .428 .373 -.042 .226 -.059 .399 Item 5 .286 .174 -.124 -.126 -.075 .579 Item 6 .184 .612 .030 -.243 -.089 -.158 Instrumental Item 1 .229 .539 -.321 .132 -.085 -.055 Support Item 2 .585 .106 -.231 .167 .061 .440 Item 3 .559 .546 -.118 -.046 -.055 .180 Item 4 .783 .173 -.037 .204 .125 .124 Item 5 .724 .148 -.192 .072 -.184 .313 Item 6 .656 .429 -.182 .126 .139 .279 Role Item 1 -.052 .784 -.086 .081 -.098 .079 Recognition Item 2 .351 .701 -.108 .078 .067 .242 Item 3 .274 .598 -.086 .190 .143 .309 Item 4 .146 .566 -.014 .052 .059 .096 Item 5 .429 .108 -.313 .110 .000 .039 Item 6 .744 .363 -.059 .012 -.038 .116 Performance Item 1 .299 .387 -.053 .607 -.086 -.093 Item 2 .141 -.032 .046 .788 -.002 .076 Item 3 .218 -.001 -.114 .752 .056 .014 Item 4 -.096 .031 -.174 .675 .034 .228 Item 5 .237 .049 -.101 .526 .044 .584 Item 6 .097 .148 .005 .225 .048 .716
Notes: Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.
The rotated component matrix shows that most variable items load high on a specific
and 2 load high on the instrumental support component. An explanation could be that
uncertainty is taken away by the provision of help or assistance. Furthermore,
informational and instrumental support have a significant overlap, which could
indicate that it might be difficult to differentiate between these two kinds of support.
As can be seen informational support items 1,3 and 4 load high on the instrumental
support component, while item 2 and 5 of instrumental support load on the
informational support component. This could point out that informational support is
more often regarded as instrumental support, than instrumental support is regarded as
informational support. Moreover, role recognition items 2, 5 and 6, and item 3 also
load high on instrumental support and informational support respectively. An
explanation for this occurrence could be that both types of support could also be
perceived as role recognition from upper management. The same holds the other way
around, where informational support item 2,4, and 6, and instrumental support items
1,2 and 6 load high on the role recognition component.
Another observation is that although almost all performance items load high on
component 4, items 5 and 6 load high on the informational support component. This
can indicate that sometimes middle management performance is linked to the extent
to which they receive indirect resources to help themselves.
In addition, a re-run of the factor analysis was performed in which the items were
forced into 7 components. This was done to check whether some items loaded high on
an extra component, this was not the case. It is also worth mentioning that all
negatively stated items, which are informational support item 3 and item 6, as well as
instrumental support items 1 and 3, all have high loadings with other components
(although they have been computed), which could indicate that these items are
4.11. Correlations table
The correlations, means, standard deviations and alphas of all variables under study
are presented in Table 2. The internal consistency of the items is checked via Cronbach’s Alpha, with an acceptable alpha of .60 and above.
TABLE 2
Means, Standard Deviations, Alphas and Correlations for all Variables
M SD α 1. 2. 3. 4. 5. 6. 1. Uncertainty 3.63 0.77 .74 _ 2. Role Conflict 2.78 0.98 .85 -.37** _ Informational Support 3.44 0.70 .64 .12 -.40** _ 4. Instrumental Support 3.55 0.84 .86 .20† -.49** .76** _ 5. Role Recognition 3.28 0.86 .82 .08 -.35** .63** .69** _ 6. Performance 3.78 0.71 .79 .05 -.27** .37** .46** .36** _
5. RESULTS
This results section will explain the results found in the correlations and regressions.
To test the hypotheses, simple (H1 and H2) and multiple (H3, H4 and H5) linear
regression analyses are performed. Notes: **p<.01
* p<.05
† p<.10
5.1. Correlations
To start with the correlations, Table 2 presents many significant (p < 0.01)
correlations, indicating that the relationships between the variables are statistically
significant. Also, most of these correlations are, overall, in line with existing theory.
Interesting to see are those relationships that do not conform to existing theory, such
as the marginally positive correlation (r= .20, p < .10) between instrumental support
and uncertainty, which is contrary to literature (Herzig & Jimmieson, 2006). This
finding indicates that higher levels of instrumental support are correlated with higher
levels of uncertainty. Another observation that stands in contrast to literature is the
negative relationship between role conflict and uncertainty (r= -.37, p < 0.01), which
indicates that a higher level of role conflict should lead to a lower level of uncertainty,
and vice versa.
5.2. Regressions
To test the hypotheses, regression analyses are performed. Simple regression analyses
are performed in order to test whether the direct effects of organizational change
(uncertainty and role conflict) have an effect on performance. Multi regression
analyses are performed to check if instrumental support has a moderating effect on
uncertainty and performance. Furthermore the moderating effects of role recognition
and informational support are tested to verify their effect on performance.
Hypothesis 1 states that in the context of change, uncertainty is negatively
related to middle management’s performance. The regression analysis shows (B = .03,
p = .75, R² = .04, F = 1.01) that this hypothesis cannot be supported, and therefore has
to be rejected. Regarding the control variables it turns out that, when uncertainty is
performance (B = .01, p = .97). Age both has a positive, as well as a significant effect
on performance (B = .01, p < .1).
Hypothesis 2 states that in the context of change, role conflicts are negatively
related to middle management’s performance. The effect of role conflicts on
performance is significant (B = -.19, p = .015, R² = .11, F = 3.12) at a significance
level of p<.05, showing a negative relationship between role conflict and
performance. Hence, these results support the hypothesis that there is a negative
relationship between role conflicts and performance; a decrease in role conflicts
increases performance, and vice versa. With regard to the control variables it can be
said that, when role conflict is added to the regression, gender has a negative, not
significant effect on performance (B = -.004, p = .99). Age, again, has a positive and
significant effect on performance (B = .01, p < .1).
5.3. Influence of the moderator
A moderator is a variable that affects the strength and/or direction of the relationship
between the independent and the dependent variables.
(Source: Baron & Kenny, 1986)
A basic moderator effect can be seen as the interaction between an independent
variable and a factor that sets the appropriate conditions for its operation (Baron &
Kenny, 1986). For example, if we look at hypothesis 3, it states that uncertainty is
moderated by instrumental support. As can be seen in Figure 4, three causal paths
come together in the outcome variable (performance), where the predictor is
uncertainty (path a), the moderator is instrumental support (path b), and the
interaction of the two (path c). The moderator is considered to be supported if the
interaction (path c) is significant. Besides, there could also be significant main effects
for the predictor and the moderator (paths a and b), however these are not directly
relevant when testing for the moderator (Baron & Kenny, 1986).
Hypothesis 3 states that in the context of change, the negative effects of uncertainty on performance are buffered by top management’s instrumental support. The multi regression analysis in Table 3 shows that this hypothesis cannot be
supported, and therefore has to be rejected. Although a negative relationship (B =
-.02) is found, it is not significant (p = .77). Another noticeable relationship, however
not hypothesized, is that instrumental support, while not having a moderating effect,
does have a direct positive relationship with performance (B = .33, p = .00).
Regarding the control variables it can be stated that gender has an overall negative,
although not significant, effect on performance (p = .99; .46; .45 for models 1, 2 and 3
respectively). Gender is coded as 1 = male; 2 = female, therefore lower values for
gender are linked to higher performance, and vice versa. Conversely, age is positively
related to performance. However, this is only significant for model 1 (B = .01, p
TABLE 3
Results for Regression Analysis for Hypothesis 3 for Performance
Model 1 Model 2 Model 3
Variables B B B Gender -.001 -.11 -.11 Age .01† .01 .01 Uncertainty -.04 -.04 Instrumental Support .33** Uncertainty X Instrumental Support -.02
Hypothesis 4 & hypothesis 5 state that in the context of change, the negative effects of role conflicts on performance are buffered by top management’s informational support, as well as top management’s role recognition. The multi
relationship is found between informational support and role conflicts (B = .04),
indicating that the provision of informational support has a positive influence on the
negative relationship between role conflict and performance.
TABLE 4
Results for Regression Analysis for Hypotheses 4 & 5 for Performance
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Another positive but not significant relationship (B = .06, p = .51) is found between
role recognition and role conflicts, meaning that the provision of role recognition has
a positive effect on the negatively hypothesized relationship between role conflicts
and performance. What can also be seen is that informational support, although not
having a moderating effect, does have a significantly positive direct effect on
performance (B = .21, p = .01). Moreover, looking at the control variables it can be
said that gender has a negative effect in all models (B = -.001). However, these
findings are not significant (p = .99). Age on the other hand has a positive effect on all
models (B = .01), indicating that with increased age also comes increased
performance. For model 1 and 4 these findings are significant (p = .09).
As previously described, no significant relationships are found for the independent
variables uncertainty, instrumental and informational support, and role recognition
with the dependent variable performance. Both direct and moderating effects are
hypothesized, and, except for one, no initial relationships are found, and therefore
hypotheses 1,3, 4 & 5 can be rejected.
However, a relationship between role conflicts and performance is found, and
therefore hypothesis 2 cannot be rejected. Moreover, direct relationships are found
between performance and both types of support.
As a final note to this section, it should be mentioned that this research has a relatively
low sample size, which can have an effect on the results.
Multicollinearity occurs when two variables are very highly correlated, and it is
usually regarded as a problem because this means that the regression coefficients may be unstable. Therefore, the Pearson’s r between each pair of independent variables should not be higher than .80 (Bryman & Cramer, 2009, p. 300). As can be seen in
Table 2, no correlations exceed the .80 cut-off point. As an extra check the tolerance
statistics is also performed, and again, there are no direct signs for multicollinearity.
When tolerance is low, the multiple correlation is high, and the possibility of
multicollinearity exists (Bryman & Cramer, 2009, p. 306). As can be seen in Table 7
no tolerance figures are close to zero, and therefore multicollinearity is unlikely.
TABLE 7
Multicollinearity diagnostics
Hypotheses Tolerance statistics
H1 .97
H2 1.00
H3 .92
H4 .94
H5 .89
6. CONCLUSION AND DISCUSSION
This conclusion and discussion section will start by providing answers to the research
questions, which will then be followed by a discussion of the researched variables.
The objective of this research is aimed at gaining insights into the important dynamics
middle managers are both targets as well as agents of change (Fenton-O’creevy, 2001;
Balogun & Johnson, 2005), but also because organizational change often brings along
feelings of uncertainty as well as role conflicts, which can have negative effects on
middle management performance. Therefore, the focus of this study has been on
certain behaviors top management can exert in order to help their middle managers
overcome uncertainty and role conflicts during change without jeopardizing their
performance. Starting with the first sub-question: how is middle management’s
performance influenced by middle management’s uncertainty and middle management’s role conflict during change? This study does not find any relationships regarding the influence of uncertainty on middle management performance. Role
conflicts however, are found to be negatively related to performance, meaning that
higher degrees of role conflict lead to lower degrees of performance, and the other
way around. Regarding the second sub-question: how do top management’s role
recognition and informational support influence the relationship between role conflict and middle management’s performance during change? This study finds that both
moderator variables of role recognition and informational support do not have a
significant moderating influence on the hypothesized negative relationship between
role conflicts and performance. However, a significant positive direct relationship is
found between informational support and middle management performance. Meaning
that informational support has a positive effect on middle management performance.
Then looking at the third sub-question: how does top management’s instrumental
management performance. This indicates that the provision of instrumental support has a positive effect on middle management’s performance.
In conclusion, the answer to this paper’s main research question of ‘how can certain top management’s behaviors, such as top management’s role recognition and top management’s support, influence middle management’s performance during change
processes?’ is that that certain top management’s behaviors can have a direct positive influence on middle management’s performance. This study finds that the provision of both types of top management support have a direct positive effect on middle
management performance. Which means that the direct provision of both instrumental and informational support positively influences middle management’s performance.
In this research it is hypothesized that uncertainty and role conflicts have a direct
negative effect on performance, which can be moderated by respectively instrumental
support, and informational support and role recognition. The results provide some
inconsistent images with regard to existing literature, since four out of five hypotheses
are rejected. A possible explanation for these inconsistencies may be due to the
relatively low sample size in this research, which can be a sound explanation for the
overlaps found in the factor analysis, as well as the fact that no moderating effects are
found. With a larger sample it can be possible that there are in fact significant
relationships for the rejected hypotheses. However, with a larger sample it can also
turn out that there are no significant relationships at all.
firstly grasp the importance of change (personal change), as well as being able to pass
the change down the line (implement changes) (Balogun, 2003). Therefore, they are
regarded as both the targets as well as the agents of change (Fenton-O’creevy, 2001;
Balogun & Johnson, 2005). This requires them to be adaptable and flexible
(McKenna, 1999), while keeping the business going, and taking on new challenges
such as assisting in strategy-making (Wai-Kwong et al., 2001). Also, they should be
able to help others through the change process (Balogun, 2003) by facilitating in the
emotional management of their employees (Huy, 2002).
Looking at the hypothesized relationships, the most consistent effect is found for the
negative relationship between role conflicts and performance, which turns out to be an
immediate direct effect. This is in line with existing literature on role conflict during
change (Floyd & Lane, 2000; Balogun, 2003; Floyd & Wooldridge, 1994). Moreover,
Tubre & Collins (2000) state that role conflict is negatively related to job performance
because it either indicates an information overload or because it tends to weaken
performance-to-reward expectancies (Jackson & Schuler, 1985). Hence, it is
important that top management recognizes that middle managers need time and
understanding to process the change, and that sending inconsistent messages only
increases role conflicts. Therefore, it is recommended that in order to reduce role
conflicts as much as possible, top management should refrain from sending
conflicting messages. Also it should be kept in mind that, due to middle management’s changing roles, it takes time for middle managers to deploy existing competencies and adopt new ones (Floyd & Lane, 2000).
that middle managers have to cooperate and comply with their directives, and that
they (senior managers) do not want to hear about difficulties. They expect middle
managers to just adjust to the organizational changes. Hence, the researchers
emphasize that senior management plays a large role in how middle managers
perform. The provision of both instrumental and informational support in sense of
education and training, or the provision of extra resources has been mentioned by the
authors (Dopson & Neumann, 1998) as important for middle managers to cope with
change. In addition, their findings also state that middle managers gain comfort from
education and training because they suppose that greater specialist knowledge will
shield them from the worse aspects of change. Although research has found
moderating effects between support and performance (Erdogan & Enders, 2007), in
this study the provision of neither informational nor instrumental support by top
management show a moderating influence on middle management performance.
Reasons for this can be that the extra provision of resources or information may increase an individual employee’s perceived personal coping with the change or the employee’s knowledge about the change and hence increase the individual’s perceived self-competence and control within the changing environment (Shaw et al,
1993). Therefore these researchers state that support in sense of training or direct help
may have a direct effect on performance, instead of a buffering effect (Shaw et al.,
1993). This is in line with findings of the present research, where, although not
hypothesized, direct relationships are found between both instrumental and
informational support and performance. Consequently, it is recommended that top
management should provide both informational and instrumental support to middle
management because this has proven to directly influence performance in a positive
