The adopted role of change agents as an explaining
mechanism for the relation between change agents’
self-efficacy and change effectiveness: A multiple case study.
Master thesis, MSc Business Administration, specialization Change Management. University of Groningen, Faculty of Economics and Business
22nd of June, 2015
Word count: 14.874 excluding references and appendices
Maureen Offenberg Student number: S2017652 Castorweg 44 7556 ME Hengelo Tel.: +31 (0)622124465 E-mail: maureenoffenberg@hotmail.com Supervisor/university: Dr. A.G. Regts – Walters
Second supervisor/university: Dr. J.F.J. Vos
2
THE ADOPTED ROLE OF CHANGE AGENTS AS AN EXPLAINING MECHANISM FOR THE RELATIONSHIP BETWEEN CHANGE AGENTS’ SELF-EFFICACY AND
CHANGE EFFECTIVENESS: A MULTIPLE CASE STUDY
MAUREEN OFFENBERG University of Groningen
Abstract
This qualitative research aims at contributing to the three adopted roles of the change agent – change initiator, change implementer and change facilitator - as an explaining mechanism for the relation between change agents’ self-efficacy and change effectiveness. In order to do so, the three adopted roles of the change agents were categorized as change agents’ behavior. For investigating the different relationships, a multiple case study was conducted within a hospital setting. In total, four physicians and seven nurses participated in semi-structured interviews. These physicians and nurses originated from four distinctive cases in which the change was implemented effectively or ineffectively. Data analysis was performed at the level of each particular case individually (within-case analysis), followed by comparing the four distinctive cases with each other in order to find differences and/or similarities (cross-case analysis). The findings suggest that a link doc with a high level of self-efficacy should adopt all three different roles as change initiator, change implementer and change facilitator, whereby he, in the role as change initiator, needs to be convinced that he can improve business processes. This vision together with the adoption of the three different roles will, eventually, lead to change effectiveness. A main practical contribution concerns the conviction of change agents that they are able to improve business processes. Theoretically, this study improves the insights in the adopted role of change agents and the consequences for change effectiveness.
3
TABLE OF CONTENTS
1. Introduction………...4
2. Theoretical framework……….9
2.1 Self-efficacy………..9
2.1.1 High self-percept of efficacy and an individual’s behavior………...9
2.1.2 Low self-percept of efficacy and an individual’s behavior………9
2.1.3 Change agents’ self-efficacy and change effectiveness………...………10
2.2 Adopted role change agent………..11
2.2.1 Roles of a change agent………11
2.2.2 Self-efficacy and the adopted roles of the change agent…..………12
2.2.3 The adopted roles of the change agent and change effectiveness……..………..12
2.3 Differences between physicians and nurses………13
3. Methodology………....15
3.1 Multiple case study………...………..15
3.2 Quality criteria………15
3.3 Case selection……….…………...16
3.4 Data collection………....18
3.4.1 Sample & Procedures……….…..18
3.4.2 Measures……….……..……19
3.4.2.1 Self-efficacy……….…………..19
3.4.2.2 Adopted roles……….………...……..20
3.4.2.3 Change effectiveness……..………20
3.5 Data coding & analysis………...………20
4. Results………..23 4.1 Within-case analysis………...…………23 4.1.1 Effective cases………..23 4.1.1.1 Case A………23 4.1.1.2 Case B………26 4.1.2 Ineffective cases………...…………28 4.1.2.1 Case C………...28 4.1.2.2 Case D………...30 4.2 Cross case-analysis……….35
4.2.1 Comparison effective cases………..35
4.2.1.1 Self-efficacy...35
4.2.1.2 Adopted roles………...…..35
4.2.1.3 Change effectiveness………...………..36
4.2.2 Comparison ineffective cases………..36
4.2.2.1 Self-efficacy………...…………36
4.2.2.2 Adopted roles……….36
4
4.2.3 Comparison effective and ineffective cases……….37
4.2.3.1 Self-efficacy………...37
4.2.3.2 Adopted roles……….37
4.2.3.2 Change effectiveness……….38
5. Discussion………...…………..39
5.1 Change agents’ self-efficacy and change effectiveness………..………39
5.2 Change agents’ self-efficacy and the adopted roles of the change agent.………..39
5.2.1 Change initiator………40
5.2.2 Change implementer………40
5.2.3 Change facilitator……….40
5.2.4 Additional elements role of the change agents……….41
5.3 The adopted roles of the change agents and change effectiveness……….41
5.4 Differences in functions………..43
5.5 Practical implications………..44
5.6 Theoretical implications………..45
5.7 Strengths, limitations and directions for future research………46
5.7.1 Strengths………..46
5.7.2 Limitations………...46
5.7.3 Directions for future research………..46
5
1. INTRODUCTION
The fact that a change implementation’s degree of effectiveness heavily depends on the change agent who leads the change is widely acknowledged in academic literature (e.g., Battilana, Gilmartin, Sengul, Pache, & Alexander, 2010; Eisenbach, Watson & Pillai, 1999; Fiol, Harris & House, 1999; Higgs & Rowland, 2011; Kavanagh & Ashkanasy, 2006; Oreg & Berson, 2011; Trice & Beyer, 1993). Many researchers have examined several characteristics of change agents which are likely to influence the effectiveness of a change project. Amongst these factors are change leaders’ behavior, such as the influence of task-oriented and person-oriented behavior, facilitating and engaging behavior and leader-centric behavior (Battilana et al., 2010; Higgs & Rowland, 2011). Moreover, the influence of change agents’ leadership style and competencies on change effectiveness has been found in research as well (Battilana et al., 2010; Kavanagh & Ashkanasy, 2006; Eisenbach, 1999; Fiol et al., 1999; Oreg & Berson, 2011; Trice & Beyer, 1993). In addition, there is evidence for the influence of change agents’ personal attributes, such as dispositional resistance, conservation values and openness to change, on change effectiveness (Oreg & Berson, 2011).
Another personal attribute is self-efficacy, and this personal attribute plays an important role in times of organizational change (Jimmieson, Terry & Callan, 2004). Self-efficacy concerns the perception, belief, judgment or expectation that a person is motivated and can execute the behavior that is required in order to produce certain outcomes (Bandura 1977, 1982; Gist, 1987; Gist & Mitchell, 1992; Paglis & Green, 2002; Robertson & Sadri, 1993; Stajkovic & Luthans, 1998). Consequently, self-efficacy influences an individual’s actions and behavior (Bandura, 1977, 1982; Gist, 1987).
6
that change agents who have high self-percepts – thus, a high level of self-efficacy - are more ready for change and, therefore, change effectiveness is very likely. This is underlined by Pearlmutter (1998), who found that only people who possess high perceptions of self-efficacy have the required abilities, such as confidence and commitment, in order to move their organizations from a present state to the desired end state. From this literature, it can be derived that there are several explaining mechanisms, such as change agents’ readiness for change, confidence and commitment, which explain the relationship between change agents’ self-efficacy and change effectiveness. However, to my knowledge, no research has been conducted regarding change agents’ behavior as an explaining mechanism for the relationship between change agents’ self-efficacy and change effectiveness. As self-efficacy deals with conviction of a person in order to produce the behavior that is required, it is important to see how this required behavior is translated into the adopted role of change agents. The level of self-percepts determines which actions people take (Armenakis, Mossholder & Harris, 1993), since perceived self-efficacy affects the choice of behavioral settings, implying that it influences the way in which people behave (Bandura, 1982). Pearlmutter (1998), in addition, found that only people who perceive themselves as highly efficacious are likely to initiate change, because they have the confidence they can master the change. People who possess low self-percepts are, however, less likely to initiate the change because they lack the confidence that is necessary in order to initiate the change (Bandura, 1977; Pearlmutter, 1998). People make an estimation whether they can master the situation, and based on this estimation they will decide which actions to take (Bandura, 1982).
Since change agents’ behavior is too broad to investigate, this study will focus on the adopted role of change agents, as a categorization of change agents’ behavior. Cawsey et al. (2012) state that each role that can be adopted by a change agent represents different types of actions and this implies that self-efficacy determines the role change agents adopt during change implementation.
7
So, while previous research shows that change agents’ self-efficacy positively influences change effectiveness, less is known about the adopted role of change agents, or change agents’ behavior, as an explaining mechanism for this relationship. Because self-efficacy influences behavior (Bandura, 1977; 1982), and change agents’ behavior influences change effectiveness (Battilana et al., 2012; Falbe & Yukl, 1992; Higgs & Rowland, 2011), in this research I will examine whether the adopted role of change agents, or change agents’ behavior, serves as an explaining mechanism of the relationship between change agents’ self-efficacy and change effectiveness. While previous research has focused on change recipients’ self-efficacy (Eby, Adams, Russell & Gaby, 2000; Oreg, Vakola & Armenakis, 2001), this research, however, focuses on change agents’ self-efficacy. The central research question in this research is therefore:
What is the relationship between change agents’ self-efficacy and change effectiveness and to what extent can this relationship be explained by the adopted role of change agents?
A visual representation of the central research question is depicted in Figure 1.
This research focuses on a specific context, that is, a hospital setting in which both physicians and nurses act as change agents. Research has indicated that people who have a better function in terms of level of salary and status, have higher self-percepts than those people who have lower salary levels and status (Abele & Spurk, 2009; Ballout, 2009; Day & Allen, 2004). When this is translated to hospital settings, it appears that physicians have a higher status and higher salary levels compared to nurses (Devine, 1978; Keddy et al., 1986). Therefore, I will focus on the difference in hospital functions between physicians and nurses as well. Hence, the following sub question is formulated:
Is there a difference between physicians’ and nurses’ level of self-efficacy, do they adopt different change agents’ roles, consequently influencing change effectiveness?
8
outcomes of the change implementation. Last, it contributes to academic literature concerning the adopted roles of the change agent and their influence on the effectiveness of a change implementation. More specifically, it draws on the roles which are necessary in order to implement change effectively. While many scholars have focused on self-efficacy as a concept (Bandura, 1977; 1982) and self-efficacy and leadership (Paglis & Green, 2002), this research specifically focuses on the influence of self-efficacy on the adopted role of change agents during change implementation, in an attempt to explain the positive relationship between change agents’ self-efficacy and change effectiveness.
Figure I – conceptual model
9
2. THEORETICAL FRAMEWORK
2.1 Self-efficacy
People possessing high self-percepts are more likely to choose environments that are unstable and need to be mastered, whereas those perceiving a low level of self-efficacy are likely to choose environments that are more stable and, thus, do not bring about many challenges. This is underlined by Robertson & Sadri (1993) who argue that efficacy expectations determine behavior and by Stajkovic and Luthans (1998), who state that self-efficacy concerns “how people’s beliefs in their capabilities to affect the environment control their actions in ways that produce desired outcomes” (p. 63). It is, thus, plausible that self-percepts influence behavior. The subsequent section elaborates upon types of behavior that are frequently encountered by people who have either a high percept or a low self-percept.
2.1.1 High self-percept of efficacy and an individual’s behavior. Bandura (1977) argues that people who find themselves capable of handling a situation will behave assuredly. This is underlined with the argument that people “undertake and perform assuredly those activities that they judge themselves capable of managing” (Bandura, 1982, p. 123). Furthermore, the level of self-efficacy determines the level of effort (Gist, 1987; Bandura, 1977; Bandura, 1982). Gist (1987) describes that people having high self-percepts tend to engage more frequently in activities that are task related and those people will not give up easily. This behavior will in turn enhance their self-efficacy. Bandura (1982) supports this assumption by arguing that self-percepts of efficacy determine the amount of effort people will extend and how long they will persist when they face obstacles. In addition, he states that people with a strong sense of efficacy will exert more effort in order to master a challenge.
10
give up easily. Hence, it can be inferred that judgments of self-efficacy determine both the intention to engage in tasks and the level of effort.
2.1.3 Change agents’ efficacy and change effectiveness. Change agents’ efficacy has its impact on change effectiveness in different ways. Firstly, the level of self-percepts of efficacy determines a person’s readiness for change (Cunningham et al., 2002). As readiness for change is one of the key determinants for the success of the change intervention (Armenakis et al., 1993), it is important to identify how self-efficacy plays a role in this process. Following Cunningham et al. (2002), people possessing high levels of self-percepts score higher on readiness for change and this implies that change effectiveness is more likely, whereas people with lower levels of self-efficacy are less ready for the change and change effectiveness is, therefore, less likely. Moreover, Bandura (1977; 1982) argues that, in general, individuals who perceive themselves as highly efficacious will contribute successful effort that produces successful and effective outcomes, whereas people perceiving low self-efficacy are more likely to fail because they cease their efforts. Wood and Bandura (1989) justify this line of argumentation by stating that a person’s conception of ability has a substantial and direct influence on performance attainments, whereby high levels of self-efficacy will lead to higher levels of performance and effectiveness and vice versa. This is all underlined by Pearlmutter (1998) who found that only change agents who possess high perceptions of self-efficacy have the required abilities in order to move their organizations from a present state to the desired end state. Only these agents have the commitment and confidence that is needed to realize change. This can be ascribed to the fact that people who have a high level of self-percepts focus on analyzing and figuring out solutions and are willing to manage more complex dilemmas. Because of these levels of self-percepts, they recognize the need for change and are able to introduce the change and are willing to undertake the tasks inherent in this change process (Pearlmutter, 1998). In short, change agents who are highly efficacious are more likely to implement change effectively than low efficacious change agents. Hence, it appears that self-efficacy influences the level of effectiveness of a certain change outcome.
11
2.2 Adopted role change agent
The definition of change agents that is previously provided clearly distinguishes between certain types of roles a change agent can adopt. It, thus, appears that change agents adopt many different roles when they lead change (e.g., Burnes, 2014; Caldwell, 2003; Cawsey et al., 2012). A commonly used categorization is the classification into the three different roles as change initiator, change implementer or change facilitator (Cawsey et al., 2012; Saka, 2003). Caldwell (2003) makes another distinction, that is, between change leaders and change managers, but these concepts show several similarities with respectively change initiators and change implementers. In this paper, therefore, the three different roles change initiator, change implementer and change facilitator are utilized.
2.2.1 Roles of a change agent. Each role has its own characteristics and important aspects that need further explanation. Therefore, the different roles will be discussed now.
According to Caldwell (2003a), a change initiator, or a change leader, can be identified as the person who gets things moving, takes action and stimulates the system. A change initiator is the one who is constantly seeking change in order to improve certain business aspects. They do so by means of identifying the need for change, visioning a better future and consequently initiating the change (Caldwell, 2003; Cawsey et al., 2012).
A change implementer or change manager following Caldwell (2003), however, is the middle-manager who gets orders from top management to make the change happen, chart the path forward, foster support and reduce resistance (Caldwell, 2003; Caswey et al., 2012). A careful note should, nevertheless, be made regarding the position of change implementers since they are very likely to experience powerlessness (Oshry, 1993), whereby they face the challenge to deal with both senior management and subordinates (Oshry, 1993). They may feel trapped between tops and bottoms and become ineffective as a result (Cawsey et al., 2012).
12
encouraging empowerment and responsibility (Cawsey et al., 2012; Higgs & Rowland, 2000; Saka, 2002).
These roles differ significantly from each other. For instance, a change facilitator remains impartial whereas both change initiators and change implementers aim at fostering support, less taking into account the concerns of the change recipients (Caldwell, 2003; Higgs & Rowland, 2000; Saka, 2002). Additionally, the change initiator focuses specifically on creating a vision, whereas change implementers and change facilitators put this vision into practice (Caldwell, 2003; Cawsey et al., 2012; Saka, 2002). However, although these roles are quite different, they appear to be complementary (Caldwell, 2003). Moreover, Caldwell (2003) argues that switching to another role might be helpful in case the current approach is not working appropriately. This is acknowledged by Grandia (2015) who argues that the role of change agents should not be considered as fixed, but as evolving throughout the change process. Moreover, sometimes you have to engage in a different role in order to manage change successfully. Hence, the different roles are complementary, implying that a change agent should adopt different roles during the change implementation in order to implement the change successfully (Caldwell, 2003; Higgs & Rowland, 2000).
2.2.2 Self-efficacy and the adopted roles of the change agent. Since Cawsey et al. (2012) argue that each role represents different types of actions, the level of self-efficacy determines which role a change agent adopts. Pearlmutter (1998) and Paglis and Green (2002) found that those change agents who perceived themselves as highly efficacious are more likely to initiate the change, because they believe that they can master the change. Nevertheless, change agents who have low-self-percepts of efficacy are less likely to initiate change because they lack the confidence that is necessary in order to initiate change. Based on this literature, I expect that especially the role as change initiator will only be adopted by change agents who have a high level of self-efficacy, whereas the roles change implementer and change facilitator will be adopted by change agents who have both a high and low level of self-efficacy, as these two roles are less focused on taking the lead.
13
successful than those change agents who focus less on their colleagues. The more directive approaches, such as leader-centric or task-oriented behavior (Battilana et al., 2012; Higgs & Rowland, 2011), are negatively related to change effectiveness (Falbe & Yukl, 1992). This means that those change agents who behave like a real leader and solely focus on tasks instead of persons, are less likely to implement change successfully.
When this is translated to the adopted roles of change agents, a distinction can be drawn between these three different roles. Following Cawsey et al. (2012), change initiators are the persons who take up the change and see the action, whereby perseverance is enormously important. Since change initiators, thus, focus on the tasks to fulfill and less on the person, it can be assumed that change initiators are more likely to engage in task-oriented and leader-centric behavior. Change facilitators, however, focuses on fostering support and alleviating resistance. Moreover, they have strong relationships with other parties involved in the change (Cawsey et al., 2012). Since this role focuses more on the human aspect of a change project and takes into account a person’s concerns and tries to take this away to make the change a success, it is plausible that a change facilitator engages more in person-oriented behavior. Last, a change implementer is more difficult to classify as it shows characteristics from both behavior classifications, namely task-oriented and person-oriented behavior. First, change implementers decide how to chart the path forward, whereby they focus on which tasks are necessary in order to make the change successful. On the other side, they work closely with people who have a stake in the change project. Thus, since they focus on both the tasks and the persons who take part in this change project, it is plausible to argue that a change implementer engages in both task-oriented and person-oriented behavior.
2.3 Differences between physicians and nurses
14
status differences between doctors and nurses, whereby “physicians are in a dominant position within the hospital and nurses are the subordinates in the organization” (Devine, 1978, p. 278). This is underlined by Keddy, Jones Gillis, Jacobs, Burton and Rogers (1986) who argue that the nurses looked up to the doctors, which caused these status differences. These differences in status find its roots in the setting for learning and the amount of formal education that each profession requires. In addition, wide salary differences among doctors and nurses exist (Devine, 1978; Keddy et al., 1986). Devine (1978) points out that the level of education and the amount of responsibilities given to the physicians are key determinants for the level of salary. These levels of salary are in favor of the physicians (Keddy et al., 1986) and they, thus have a higher level of income. Hence, since people with higher salary levels and degree of status have higher self-percepts (Abele & Spurk, 2009; Ballout, 2009; Day & Allen, 2004) and doctors have a higher status and higher salary levels, it can be inferred that doctors should have a higher sense of self-efficacy. Therefore, it seems plausible that physicians are more likely to adopt the role as change initiator, while nurses are more likely to adopt the role as change implementer and change facilitator.
15
3. METHODOLOGY
3.1 Multiple Case Study
Given the fact that the relationship between self-efficacy and the adopted roles of change agents have not been investigated yet, an explorative study is deemed necessary as a first step in this research. A case study is most suitable for answering how or why questions (Yin, 2009) and is therefore the most appropriate knowledge generating process for this objective, because this research design focuses on providing an answer to the question how the adopted role of change agents serve as an explaining mechanism for the relationship between change agents’ self-efficacy and change effectiveness. Moreover, to my knowledge, no research has been conducted regarding change agents’ behavior as an explaining mechanism for the relationship between change agents’ self-efficacy and change effectiveness.
Thus, the adopted role of change agents is insufficiently explained in academic literature. To get a more comprehensive understanding, an embedded multiple case study is conducted. The logic behind this type of case study calls for a careful selection of the cases, so that it either (a) predicts the similar results – which is also known as literal replication – or that is (b) produces contrasting results but for valid reasons – which is also known as theoretical replication (Yin, 2009). Given this information, this research design is most suitable for this study as it will focus on the differences and similarities among both effective and ineffective cases. Based on the selection of two effective and two ineffective cases, I will examine whether the level of change agents’ self-efficacy and the adopted role(s) of the change agents lead to a level of change effectiveness, whereby literal and theoretical replication is considered.
3.2 Quality Criteria
inter-16
subjective agreement, controllability, reliability and validity need to be assured (van Aken, Berends & van der Bij, 2012).
Following the criteria of Swanborn (1996), controllability was achieved by publishing the paper and describing in detail how the research has been conducted. Furthermore, both deductive and inductive codes were published in a so-called code book in Appendix I, II and III of this paper. Last, inter-rater agreement is an important aspect to guarantee controllability and, therefore, I strived for homogeneity and consensus by asking fellow master students whether they agreed with several interpretations and propositions made in the paper. In this way, researchers are able to replicate the study.
Reliability and validity are research-oriented criteria that need to be considered as well (Yin, 1994). Reliability means that propositions concerning the empirical world should be at least independent at three different aspects; the need to be researcher independent, instrument independent, respondent independent and circumstances independent (van Aken, Berends & van der Bij, 2013; Swanborn, 1996). To fulfill these requirements, a few of the coded interviews were controlled by other fellow master students, so researcher bias is likely to be reduced. To control for the respondents bias, both link docs and nurses were interviewed and, in addition, the coordinator of the change project has been interviewed during an introductory interview in order to grasp an overall view of the link docs and nurses and how they acted during the dresscode project following her point of view. Last, to increase reliability with regard to circumstances, the data collection included a time span of one month to avoid that all people will respond because of certain triggers that only occur at specific moments in time.
Last, it is of vital importance that the interview will measure what is intended to measure in order to guarantee construct validity. Therefore, before data collection started, all interview questions were assessed and criticized by two independent supervisors.
3.3 Case Selection
17
specifically, these requirements concerned the haircut, the jacket and jewelry. To let the dresscode succeed, one or more so-called link docs and link nurses for each department were appointed as change agents. The hospital has called them link docs and nurses as they are perceived to be the ‘link’ between the infection prevention department and the employees who work at the departments, for instance the intensive care. These link docs and nurses were supposed to encourage their colleagues to fulfill the dresscode requirements. Throughout this research, the label link doc and link nurse will be utilized. For this project, many data existed regarding its effectiveness because of measurements that were organized twice a year. These measurements focused on whether a department fulfilled the dresscode requirements. This was the reason for choosing this change project. Moreover, this project was the ideal situation to pick two effective and two ineffective cases in order to consider both literal and theoretical replication, as data regarding the change effectiveness for each case already existed. Hence, documents regarding the outcomes of the dresscode in December 2014 for each department were analyzed in order to select the two most effective and ineffective cases. A clearer overview of the four selected cases and their change effectiveness is depicted in Table I. Since each department consisted of a few smaller departments, the average score for the overarching department is reported
Table I – Case characteristics and change effectiveness per case
Case Case description Scores December 2014 (10 random persons
observed per case)
E ff ec tiv e A Intensive care
Total of 2 link docs and 5 link nurses
10 persons fulfilled the 6 dresscode requirements.
B
General internal medical science
Total of 2 link docs and 2 link nurses
9 persons fulfilled all 6 dresscode requirements, 1 person complied with 5 dresscode requirements. In eff ec tive C
Child medical science
Total of 1 link doc and 2 link nurses
5 persons fulfilled the 6 requirements, 3 people complied to 5 requirements 2 people to 4 requirements.
D
Thorax surgery
Total of 2 link docs and 3 link
18
nurses 4 persons complied to 4 requirements 2 persons to 3 requirements.
3.4 Data Collection
3.4.1 Sample & Procedure.
Data collection started by sending an email to nineteen link docs and nurses of the four departments of the sample. There were eleven persons who responded and after a week, another reminder was sent to the link docs and nurses who did not respond yet. What became apparent was that the eight out of eleven link docs and nurses for the effective cases were willing to participate and sent an email within one week, whereas only two out of seven link docs and nurses for the ineffective cases responded to the emails within one week with the message that they were willing to participate. Because many physicians were not willing to participate, I decided to interview only one link doc for each case. Moreover, as several departments only had two link nurses, I have chosen to interview two link nurses to avoid a skew distribution of respondents. When the reminding email did not improve the response rate, people were contacted by phone and suddenly all link docs and nurses who were contacted by phone were willing to participate. After their acknowledgement of their willingness to participate, an appointment was made with all the link docs and nurses for conducting a semi-structured interview.
In total, twelve semi-structured interviews were conducted. Two different categories of actors were interviewed, namely eleven link docs and nurses, and the overall coordinator of the infection prevention trajectory was interviewed during an introductory interview. Per case approximately one link doc and two link nurses were interviewed. All respondents participated in the infection prevention trajectory at the hospital. An overview of these respondents for each case can be found in Table II.
Table II – Overview interviewees
Function Function as change agent Code Gender
Case A
Head of department Link Doc D01 Female
19
Head nurse Link Nurse N04 Female
Case B
Medical specialist Link Doc D04 Male
Senior nurse Link Nurse N02 Male
Case
C
Child infectious Link Doc D03 Female
Child nurse Link Nurse N05 Female
Nursing consultant Link Nurse N07 Female
Case
D
Nurse practitioner Link Doc D02 Female
Senior nurse Link Nurse N03 Female
Executive nurse Link Nurse AQ Female
Expert/Manager infection prevention
Coordinator of dresscode project
C01 Female
From this table, it can be derived that for case B, only two interviewees participated in this study instead of three, whom participated in the other cases. This can be ascribed to the fact that the link nurse for this department was not willing to participate and this department had no other link nurses.
All interviews were between fifteen and thirty minutes in length, were recorded and transcribed. Link docs and link nurses answered the same questions concerning their role during the dresscode trajectory, their self-efficacy and their satisfaction regarding the effectiveness of the dresscode trajectory. I also conducted an introductory interview with the coordinator of the change project whereby she told me about the project in general and the link docs and nurses from whom she had the intention to approach for an interview with me. For this interview, no interview scheme was utilized. Therefore, only the questions of the interview with the link docs and nurses can be found in Appendix IV.
20
When analyzing these documents, I decided to utilize the most recent measurement moment; December 2014.
3.4.2 Measures. In this section, for each variable a description is given about the measurement method.
3.4.2.1 Self-efficacy. Self-efficacy was measured with the generalized self-efficacy
scale from Schwarzer and Jerusalem (1995). A sample item is “Thanks to my resourcefulness, I know how to handle unforeseen situations.” Responses were given on a four-point scale ranging from 1 (Not at all true) to 4 (Exactly true). The scores for all ten statements needed to be added in order to reach an overall score. The maximum score for this scale is 40. Scores below 28 are categorized as a low level of self-efficacy, scores of 28 and 29 are considered as a medium level of self-efficacy and scores of 30 and higher are considered as a high level of self-efficacy (Schwarzer & Jerusalem, 1995).
3.4.2.2 Adopted roles. The role of change agents was measured by asking questions
such as: “How do you describe your role as link doc/nurse during the dresscode trajectory?” or “In which type of activities did you engage as link doc?” The answers for these questions were transcribed and coded according to the procedure that is described in the following section.
3.4.2.3 Change effectiveness. Change effectiveness was measured by analyzing
documents reporting about the outcomes of the dresscode trajectory. I received several documents with outcomes of the measurements from each half year, however I decided to pick the measurement moment in December 2014 as this was the most recent measurement. This document clearly illustrated how well each department scored on the dresscode requirements. Ten randomly chosen employees per case were observed by employees from the infection prevention team and for all ten persons an overview was made whether they complied with the six requirements.
3.5 Data coding and analysis
21
concerns the interplay between both deduction and induction due to the fact that theory building is likely to occur in an ongoing dialogue between pre-existing theory and new insights generated as a result of empirical observation. First, deductive codes were established. These codes were derived from the existing literature. From there on, selective coding, described as “The process of selecting the core category, systematically relating it to other categories, validating those relationship, and filling in categories that need further refinement and development” (Strauss & Corbin, 1990, p. 116) was utilized in order to divide the statements belonging to one code in several clearly defined sub-categories. Subsequently, inductive codes were derived from reading the transcripts of the conducted interviews. Both deductive and inductive codes and examples are depicted in the code book, which can be found in Appendix I, II and III. Moreover, an overview of solely the code names is depicted in Table III. What becomes evident in this Table III is that although each role has its own code and belonging subcategories, there exist a general code for the role of change agents as well. This can be ascribed to the fact that during the inductive coding, several subcategories were identified which belong to a role of the change agent. However, it was not possible to assign these subcategories to one specific role and, thus, a general code for all roles was established. After conducting within-case analysis, the cross-case analysis started by comparing first the effective cases with each other and the ineffective cases with each other. Subsequently, a comparison between the effective and ineffective cases took place. For each comparison, an analysis was conducted in order to grasp an in-depth understanding of the similarities and differences among these cases. More specifically, I examined whether differences in the level of change agents’ self-efficacy and the adopted role of the change agents lead to different or the same outcomes, whereby literal and theoretical replication is considered.
Table III – Overview codes
OVERVIEW CODES
Code Subcategory Type
Self-efficacy Expectation Deductive
Level of effort Deductive
22
Change implementer Charting path forward Deductive Receiving orders from the top Deductive Change facilitator Identifying emerging issues Deductive
Fostering support Deductive
Encouraging responsibility Deductive
Communicating Deductive
Role of change agent Creation of culture Inductive Lack of responsibility Inductive
Lack of priority Inductive
Refusing role Inductive
Being an example Inductive
Perceived change effectiveness
23
4. RESULTS
In this section, the results of the analysis of the implementation of the dresscode at the university medical centre will be discussed. In Appendix I, II and III you can find the coding schemes including the quotes from all respondents. The results showed that the change agents in the effective cases adopted the three different roles as change initiator, change implementer, and change facilitator. The role as change initiator came forward at the very beginning of the dresscode trajectory, where physicians and nurses could voluntarily sign up for a role as link doc or link nurse. The other two roles came forward during the dresscode trajectory, after the link docs and nurses had signed up and the dresscode project had started. Moreover, the so-called link docs and nurses who adopted these roles, had a high level of self-efficacy. However, in the ineffective cases two aspects came forward. First, in one ineffective case a link nurse adopted all three different roles. Nevertheless, her level of self-efficacy was medium. Second, all other link docs and nurses in the ineffective cases did not adopt all of the three different roles, especially the role of change initiator was lacking.
The next section will elaborate on the main findings for each case, providing more details about the variables change agents’ self-efficacy, adopted roles of the change agent and change effectiveness. After this within-case analysis, Table IV is depicted in order to summarize all the findings and provide a clear schematic overview. Subsequently, an overview of the differences and similarities among the four different cases is provided.
4.1 Within case-analysis
4.1.1 Effective Cases. The first section of this within case-analysis discusses the results for both effective cases, case A and B.
4.1.1.1 Case A. This case is, with a score of 10 out of 10, the best scoring participant
for the dresscode. For this case, one link doc and two link nurses have been interviewed and certain things became apparent during these interviews.
First of all, the link doc was head of the department as well and was given a very high profile. She was one of the initiators of the dresscode project and was highly motivated to make it work for her department. Furthermore, she had the highest measured self-efficacy level for all respondents, namely 40 out of 40. This, additionally, became evident during the interview, where sentences such as “I am 55 years old, when you have such a function like
24
(D01)” confirmed her extremely high level of self-efficacy. She argues, in addition, that she
keeps on going, because it is important for her and she is not willing to stop until goals are attained. Moreover, she adopted all three roles during the change implementation, whereby her role as change initiator was apparent. As previously mentioned, she was one of the initiators of the project because she identified several aspects that needed further improvement. For example, she identified the need for change: “I am one of the initiators of
the project (D01)” and she was willing to improve business processes: “The connection between the infection prevention and the intensive cares needed to be improved (D01).”
However, the stake of the other two roles – change implementer and change facilitator – during her role as link doc may not be underestimated. For instance, she “followed all the
introduction meetings in which they told us (the link docs and nurses) what to do and how to do it (D01).” This quote illustrates her role as change implementer, where she receives orders
from the top. Last, by following her own instinct, she charted the path forward: “I followed
my own instinct and did not take into account the orders from the top (D01).” That this link
doc adopted the role as change facilitator as well became evident with the following sentences whereby she identified emerging issues: “We are looking at the results, are they good or not?
If not, how can we improve that? (D01)” and encouraged that her colleagues took
responsibility as well: “All employees at the intensive care take initiative if people do not
comply with the dresscode. That is something I appreciate (D01).” In addition to these roles,
this link doc “strives for a culture in which everyone makes each other aware of the dresscode
by constantly pointing out and making people aware of the dresscode (D01).” An overview of
all the roles and the belonging facets for this case is depicted in Figure II. Last, when mapping the change effectiveness for this case, it became apparent that although she was convinced that the goals for her department were met, she was not completely satisfied because there are always people who will reduce their attention for the dresscode after a certain period of time. Therefore, the attention for the dresscode should never weaken, because that would eventually lead to a reduced compliance to the dresscode.
Moreover, two link nurses were interviewed. Both link nurses were in an executive position for this department. The level of self-efficacy for both nurses was high, namely 32 out of 40 (N01) and 30 out of 40 (N04). Sentences such as: “I have an executive position, so it
– taking the lead and pointing out to people - is not new for me. I know how to deal with these situations (N01)” confirm this level of self-efficacy. One of the two link nurses (N04) was
25
nurse (N01) solely adopted the roles as change implementer and change facilitator. One of these link nurses identified that she voluntarily signed up because she wanted to improve business processes: “I want to balance the matters from the infection prevention with our
department (N04)”, indicating her role as change initiator. Both link nurses, additionally,
engaged in the roles as change implementer and change facilitator. For instance, they charted the path forward: “We, the team of link docs and nurses, are making a plan about how to
implement the plan of the infection prevention (N01)” and encouraged responsibility from
other colleagues: “It is also up to my colleagues, they should have the feeling that they should
comply with the requirements (N04)”. Again, when mapping the change effectiveness for
their department, it became evident that both nurses were not convinced that when goals are attained, it is enough. They keep on emphasizing that there will attention concerning the dresscode will decline, and therefore they should keep on making people aware of the dresscode. This is illustrated with the following quote: “You always have to pay attention to
the dresscode; otherwise people will forget (N01).”
All in all, two out of three link docs and nurses adopt the three different roles as change initiator, change implementer and change facilitator. Furthermore, the level of self-efficacy for all participating link docs and nurses in this case is high. In addition, although the change effectiveness for their department is outstanding, they all emphasize that once the goals are accomplished, it is a matter of staying focused. Otherwise people will weaken their attention for the dresscode.
Figure II – overview roles and belonging facets for case A
Case A
Change facilitator - Identifying emerging issues - Foster support - Encourage responsibility - Communicating Change implementer- Charting path forward - Receiving orders from top Change initiator:
- Identifying need for change
26
4.1.1.2 Case B. This case scores relatively good on the dresscode compliance (9 out of
10). One link doc and one link nurse have been interviewed, which has led to many findings.
The link doc already had a background in infection prevention, implying that he already had enough knowledge before the infection prevention trajectory started. Because of this knowledge, he was willing to take his responsibility. One of the reasons for becoming a link doc is illustrated with the following sentence: “Before I was a link doc, I saw things at my
department which could have been improved (D04).” Moreover, he made people aware in
case they were risking an infection, this was already before the infection prevention trajectory had started. Thus, this link doc already identified the need for change before he was appointed as link doc, illustrating the role as change initiator. Moreover, he fulfilled the roles as change implementer and change facilitator by communicating to his colleagues and trying to give them responsibility as well in order to make the dresscode a success for his department. This is illustrated with the following sentences: “I point out to people when they do not comply
with the dresscode requirements (D04)” and “People take their own responsibility when it comes to the dresscode. And that is exactly what is necessary (D04).” In addition to these
roles, this link doc pointed out that you should “create a sphere in which colleagues perceive
it as normal to show when others do not comply (D04).” Here, the creation of a culture is
important for this link doc. An overview of all the roles and the belonging facets for this case is depicted in Figure III. Moreover, the generalized self-efficacy scale has proven that his level of self-efficacy was high. This link doc had a score of 33 out of 40.When this link doc spoke about the effectiveness, it became evident that the goals for this department were attained. However, they are not finished with the dresscode project, because in his opinion, they have to keep people conscious otherwise a decline in goal attainment is likely to occur.
The link nurse who has been interviewed was very motivated and by giving responsibility to his colleagues, he tried to make the dresscode a success for his department, showing similarities with the role as change facilitator. This is illustrated with the following sentence: “The policy is taken up by the whole department (N02).” He adopted the role as change implementer as well. For instance, by charting the path forward by means of discussing with colleagues: “How can we do something about it? That is one of the things we
constantly discuss with our team of link docs and nurses (N02).” However, the role as change
27
the feeling that I could deal with the difficulties I encountered (N02)” and that he should
never stop with convincing people to comply with the dresscode. Last, the goals for this department are accomplished; everybody complies with the dresscode requirements according to this link nurse. However, he is aware of the fact that the goals are never fully accomplished: “One of the goals is to keep it like this and that calls for continuously paying
attention (N02).”
To conclude, one out of two link docs and nurses adopted the three roles as change initiator, change implementer and change facilitator. Furthermore, the level of self-efficacy for all two participants was high. Although the dresscode scores for their department were relatively good, they want to ensure this level of dresscode compliance. Therefore, attention should never weaken.
Figure III – overview roles and belonging facets for case B
Case B
Change facilitator - Fostering support - Encouraging responsibility - Communicating Change implementer- Charting path forward Change initiator:
- Identifying need for change
28
4.1.2 Ineffective cases. The second section of this within case-analysis discusses the results for both ineffective cases, case C and D.
4.1.2.1 Case C. This case scores quite bad concerning the dresscode requirements,
namely 5 out of 10 persons complied with all the requirements. For this case, one link doc and two link nurses had been interviewed, leading to several remarkable findings.
First of all, the link doc was initially not willing to participate, because she did not have any intention to convince her colleagues of the benefits for the dresscode. After kindly requesting her again to participate, she was willing to participate but only under the condition that she could explain why she did not convince her colleagues about the interests of the dresscode. What came forward during this interview was that the link doc for this department refused her role and responsibility as link doc, even though she had a background in infection prevention. She never had the intention to change certain aspects with regard to the dresscode for her department. The reason for this is illustrated with the following sentence: “I have
consciously chosen not to take into account the requirements of the infection prevention, because I consider it as not important (D03).” Therefore, she does not adopt any of the three
roles as change initiator, change implementer or change facilitator, since she somehow refuses her role as link doc. Although she receives the orders from the top, she does not take them into account. Furthermore, the general self-efficacy scale has indicated that her self-efficacy level is high, namely a score of 31 out of 40. As this link doc was not willing to follow the complete interview protocol, no interview data exist in order to affirm this level of self-efficacy. Last, although she does not have the intention to change for her department, she is also not satisfied with the outcomes of the dresscode for her department. When this link doc was asked about her level of satisfaction regarding the dresscode for her department, she answered: “I am absolutely not satisfied concerning the dresscode for our department
(D03).”
In contrast, the two link nurses who had been interviewed had a total different point of view than the link doc. Both link nurses were convinced about the importance of the dresscode and tried to make their colleagues aware of that. One of the link nurses (N05) took several actions that conform with all three roles. The following quote stresses why she became a link nurse: “In the past I always saw things, and I thought: I should do something
about this. This can way better (N05)”, whereby she, thus, identified the need for change
29
hoped to improve those things, since she was quite conscious about several aspects for her department that needed improvement. Moreover, this link nurse engaged in the roles as change implementer and change facilitator as well. For instance, she charted the path forward:
“I have a lot of plans about how to improve the dresscode, for instance a presentation (N05)”
and she communicated: “I constantly give arguments, especially about why they have to wear
a white jacket (N05).” However, interview data proved that she refused her role many times,
because she “has not enough time to fulfill my role as link nurse. So, it is out of my system
(N05).” Although the link nurse adopted all three different roles, her level of self-efficacy was
medium. The generalized self-efficacy scale identified a score of 29 out of 40. The differences in status among physicians and nurses was one of the major problems for her and this has led to a lower level of self-efficacy, as she was convinced that “physicians do not take into
account my concerns, it takes someone else to make the dresscode work for physicians (N05)”. As a result, she did not make physicians aware of the dresscode in case they did not
comply with the requirements. When mapping the change effectiveness, the link nurse is quite satisfied with the outcomes for her department: “I am quite satisfied, we are on track. Only
the jackets are a problem (N05).”
The other link nurse specifically focused on trying to find ways to implement the dresscode for her department and when she found ways, she tried to make it work for her department by constantly communicating. This implies that she adopted two different roles during the implementation of the dresscode, namely change implementer and change facilitator whereby the facets ‘charting path forward’ and ‘communicating of both the roles change facilitator and change implementer came forward. This can be justified with the following sentences: “You should implement the dresscode in such a way that is more
appropriate for your team (N07)” and “You listen to them about how they experience the dresscode, and then you try to understand them and find a solution (N07).” Moreover, the
generalized self-efficacy scale has proven that her level of self-efficacy was high with a score of 31 out of 40. This was affirmed during the interview: “Yes, I could deal with these
difficulties (N07).” Last, when the change effectiveness for her department was mapped, it
became evident that she was quite satisfied with the outcomes for her department. However, she admitted that it was a continuous process that will never stop: “You always continue with
the dresscode, you will never end because attention among people will lower (N07).”
30
to change this. Moreover, one of the link docs and nurses adopted all three different roles, but her efficacy was medium, which is in contrast to the other link doc and nurse whose self-efficacy level was high and who adopted solely two roles or no role at all. Last, an overview of the adopted roles and belonging facets for this case is depicted in Figure IV.
Figure IV - overview roles and belonging facets for case C
4.1.2.2 Case D. This case scores worst on the dresscode requirements, namely a score
of 4 out of 10 people who complied with all the dresscode requirements. For this case, one link doc and two link nurses have been interviewed, which had led to several findings.
During the interview with the link doc, she did not comply with the dresscode requirements and she tried to explain why not. It came forward that she somehow understands why people did not comply to the dresscode, sentences such as “I understand it when they do
not comply, because they only talk to the patient (D02)” and “It is all about priority, in my opinion the hand hygiene is more important (D02)” were quite common during the interview.
Such sentences prove of a lack of priority for the dresscode and of refusing her role as link
Case C
Change facilitator
- Communicating
Change implementer
- Charting path forward - Receiving orders from top Change initiator:
31
doc. This, additionally, indicates that she did not adopt all three roles during the implementation of the dresscode. Both roles as change initiator and change implementer were lacking. The role as change facilitator, nevertheless, came forward during the interview, whereby she communicated: “I make people aware of the dresscode by just speaking to them
in person (D02)” and identified emerging issues: “It is all about checking whether your colleagues fulfill the dresscode requirements (D02).” Although the generalized self-efficacy
scale identified a high level of self-efficacy, namely a score of 30 out of 40, interview data concerning the level of self-efficacy proved the contrary. Especially her level of effort proved the opposite, since this link doc argued that: “Once I have drawn attention for the dresscode
when someone does not comply, I will not say it again three days later (D02).” This proves of
a low level of self-efficacy, since the aspect ‘level of effort’ should have been high for people who possess high self-percepts. Last, when I asked her about her opinion concerning the change effectiveness for her department, it became apparent that she was quite satisfied and that the process worked out very well: “The process worked well for my department (D02).”
In addition, two link nurses have been interviewed. It became apparent that one of the two link nurses (N03) did not consider the dresscode as important. During the interview, it became clear that she did not give any priority to the dresscode: “I paid more attention to the
hand hygiene (N03).” This implies that the dresscode was less important to her. As a result,
she did not adopt all three different roles as change initiator, change implementer and change facilitator. She mainly communicated with her colleagues, which can be classified under the role change facilitator and this is illustrated with the following sentence: “I speak to the
persons when they do not comply with the dresscode (N03).” Moreover, as a change
facilitator she identified the emerging issues: “That you signal when things are going wrong
(N03).” The other two roles, change initiator and change implementer, are lacking. The
generalized self-efficacy scale identified a high level of self-efficacy for this link nurse, namely a score of 32 out of 40. This was affirmed during the interview, whereby she stated that beforehand “she had the feeling that she could deal with all the difficulties which she
could possible encounter (N03).” Last, although she admitted that the scores for this
department were far below the standard, she was very satisfied with the implementation trajectory: “From the beginning of the dresscode, the trajectory went well (N03).”
32
change implementer: “We discussed how to make the dresscode a success for our department
(N08).” Moreover, by constantly pointing out to people and communicating to them that they
should comply with the dresscode, she tried to foster more support. These actions reveal a role as change facilitator that focuses on communicating, since she: “points out when people
forget their ring (N08).” Thus, two types of roles can be identified, namely the role as change
implementer and change facilitator. However, she told that she is not the only person who is responsible, aiming at the fact that all her colleagues are old enough to realize that they should comply and, therefore, have their own responsibility: “I am not the only person who is
responsible (N08).” This proves of a lack of responsibility. The generalized self-efficacy
scale identified a high level of self-efficacy with a score of 39 out of 40, however interview data proved a lower level of self-efficacy because she argued that: “Especially in the
beginning I was hesitant, because I was wondering how I should deal with people who did not comply (N08).” When mapping change effectiveness, she argued that all goals were obtained
and she could be extremely satisfied with the results: “I am very satisfied for the process as a
whole, because people know that they should comply (N08).”
33
Figure V - overview roles and belonging facets for case D
Table IV – overview findings for each case
Respondent Score self-efficacy (40 is maximum) Indication level self-efficacy
Adopted roles Adopted facets of roles Outcomes
dresscode
C
ase
A
D01 40 High
Change initiator Identifying need for change Improving business processes
Effective (10 out of
10) Change implementer Charting path forward
Receiving orders from the top Change facilitator Identifying emerging issues
Encouraging responsibility Communicating
- Creation of culture Being an example N01 32 High
Change implementer Charting path forward
Receiving orders from the top Change facilitator Fostering support
Encouraging responsibility Communicating
N04 30 High
Change initiator Improving business processes Change implementer Charting path forward Change facilitator Identifying emerging issues
Case D
Change facilitator - Identifying emerging issues - Communicating Change implementer- Charting path forward Change initiator:
34 Encouraging responsibility Communicating C ase B D04 33 High
Change initiator Improving business processes Identifying need for change
Effective (9 out of
10) Change implementer Charting path forward
Change facilitator Fostering support
Encouraging responsibility Communicating
- Creation of culture N02 36 High
Change implementer Charting path forward Change facilitator Encouraging responsibility
Communicating - Creation of culture C ase C D03 31 High
Change implementer Receiving orders from the top
Ineffective (5 out of 10) - Lack of priority Refusing role N05 29 Medium
Change initiator Identifying need for change Change implementer Charting path forward Change facilitator Communicating - Refusing role N07 31 High
Change implementer Charting path forward Receiving order from the top Change facilitator Communicating
- Being an example
C
ase
D
D02 30 High
Change facilitator Identifying emerging issues Communicating Ineffective (4 out of 10) - Lack of responsibility Lack of priority Refusing role N03 32 High
Change facilitator Identifying emerging issues Communicating
- Lack of responsibility Lack of priority N08 39 High
35
4.2 Cross case-analysis
Now the findings for each case are discussed and schematically represented, this section elaborates upon the differences and similarities among the different cases. First, a comparison between the effective cases will be made. Subsequently, the ineffective cases will be compared and, last, the effective and ineffective cases will be compared to each other.
4.2.1 Comparison effective cases. From the previous sections, many similarities and differences among the effective cases have been encountered. For each variable, an overview is provided with the similarities and differences of the effective cases.
4.2.1.1 Self-efficacy. When it comes to self-efficacy, all interviewed link docs and
nurses had a high level of self-efficacy, whereby for both cases one link doc or nurse had an extremely high level of self-efficacy. This high level of self-efficacy has led to the fact that the level of effort each link doc or nurse puts in his/her role was very high. Most of the link docs and nurses for these cases noticed that they never stop making people aware in case one of their colleagues did not comply with their dresscode.
4.2.1.2 Adopted roles. What came forward in the previous section is that for both
36
this might be an additional dimension which may possibly lead to a higher level of change effectiveness.
4.2.1.3 Change effectiveness. Last, although for both cases the effectiveness of the
change is very high, all link docs and nurses admitted that goals are never attained and that they should never stop paying attention to the dresscode. In addition, all link docs and nurses were aware of the fact that once they weaken their attention, a decline in dresscode compliance is very likely to occur.
4.2.2 Comparison ineffective cases. From the previous sections, many similarities and differences among the ineffective cases have been encountered. For each variable, an overview is provided with the similarities and differences of the ineffective cases.
4.2.2.1 Self-efficacy. In addition, the level of self-efficacy for both cases is high. Only
one link nurse has a medium level of self-efficacy, all the other link docs and nurses have a high level of self-efficacy. However, one link doc and link nurse showed contrary results for her level of self-efficacy. Although her high level of self-efficacy, they admitted that they will not continuously point out to people when they do not comply with the dresscode requirements or that they were hesitant in the beginning. Therefore, although their level of self-efficacy is high according to the generalized self-efficacy scale, the interview data revealed a lower level of self-efficacy.
4.2.2.2 Adopted roles. The link nurse with the medium level of self-efficacy of one
