The Examination of Change Management Using Qualitative Methods: A Case Industry Approach
Smith, Aaron C. T., Evans, Daniel M., Westerbeek, Hans M., The Qualitative Report
Despite the number of theories explaining the nature and antecedents of change, there is no consensus on a universally applicable model. Competing theories have been tested using deductive methods focusing on hypothesis testing. This study has utilized qualitative methods for collecting data within the sport industry to provide an initial understanding of change within that case industry Semi-structured interviews were conducted with 29 sport managers across Australian National and State Sporting Organizations and clubs participating in national league competitions. Interviews were transcribed and coded in a grounded interpretation culminating in a typology of change types. Results highlighted that Australian sport managers were inclined to be flexible in both their view of the origins of change, and its effective management. Key Words: Change Management, Sport Organizations, and Qualitative Methods
There is no paucity of theories describing or explaining change, its antecedents, content, and impact. Many of these theories have been cultivated after prolonged and rigorous qualitative field study, and paint seemingly accurate representations of the organizations that have been investigated. Subsequently, organizational change management studies seeking to test the veracity of such theories or metaphors for change within specific settings (Cooke & Szumal, 1993; Cousens, 1997; Fox-Wolfgramm, Boal, & Hunt, 1998) have tended to employ hypothesis testing, relying on the deductive verification of predetermined features or characteristics. If they are present to a degree of statistical significance, then the hypotheses may be confidently identified. This could demonstrate that a particular approach to change was employed within, or imposed upon, an organizational setting. The weakness of this approach to theory testing is that it can ignore the presence or avoid the collection of evidence reflecting other characteristics that expose the utility of other theoretical perspectives. This research seeks to use qualitative methods to investigate the importance of competing change management theories in the administration of sport in Australia. Importantly, the use of qualitative data means that conclusions about theoretical approaches are established inductively, and results can be compared to known characteristics associated with established theories, rather than beginning with these characteristics and searching for evidence of their presence.
The degree to which any of the well-established change management theories are applicable to the Australian sport industry is unclear. There are differences in both the study and the management of change across culture and industry. Indeed, what is meant by "change" can vary from study to study. Due to the inclusive, exploratory nature of this paper, George and Jones' (1995) generic definition of organizational change as the movement away from a present state toward a future state is employed here. This broad definition allows the greatest scope for recognizing change within an organizational setting, and therefore is ideal for this paper's purpose. In a turbulent global market the Australian sport industry has flourished in the past twenty years, navigating, among other things, significant shifts in global broadcasting stakeholders and methods, and local funding and participation issues (Westerbeek & Smith, 2003).
Notwithstanding a handful of pioneering studies in Australian sport (Skinner, Stewart, & Edwards, 1999), and the evidence accumulated overseas and in generic business studies, Australian sport change management researchers are largely beginning from "scratch." Thus, Australian sport management is faced with significant knowledge gaps: Generic change studies do not agree on the most fruitful theories for conceptualizing change, some providing detailed descriptions of change, while others emphasize specific prescriptive interventions (Ginsberg, 1988; Laughlin, 1991). In addition, the scarcity of local data means that few "leads" are forthcoming. This study has been undertaken in response to this lack of data.
The purpose of this exploratory study is to provide evidence to explain how sport managers conceptualize organizational change. It seeks to ascertain whether theoretical models accurately reflect the perceptions of practitioners, or whether sport managers' perceptions of change are suggestive of the need for further theory development. In other words, should studies on change in Australian sport focus chiefly on theory development, or should they focus on theory confirmation and refinement, and toward the progression and modification of current theories of change in the Australian sporting context? The purpose of this research can be conflated to the following question: How do Australian sport managers conceptualize the organizational change they experience? The extension and significance of this question can be stated in another: Is change in Australian sport so unique that present theories fail to adequately explain it? The results from this study are therefore examined in light of existing theories of change. This exploratory study may subsequently be used to further investigate change management models for sport. This information is essential if change management inquiry in Australian sport is to progress toward serious theory testing and development. Consequently, this study serves as a launching point for additional research focusing upon theories that are the most fruitful in explaining practitioners' perceptions of change.
The paper includes five sections, and proceeds as follows. First, established organizational theories associated with change types are identified. Importantly, this section concludes by recognizing that generic change studies do not agree on the most fruitful theories for conceptualizing change. Second, the limited research concerning change management in a sport context is examined. Having identified the current state of literature, section three of the paper explains the method employed in this study. Section four presents the results and a discussion, while finally in section five, the implications and conclusions from the work are considered.
A Brief Overview of Established Organizational Change Theory
Change theorists have developed a range of sophisticated conceptual models to explain the nature of change both within industries and organizations. One way of understanding the complexities of each approach is to consider them against two dimensions: mode and level, as suggested by Meyer et al. (1993). Mode refers to the size and speed of the change, while level describes whether the change is proceeding within a specific organization or as part of a broader sector-wide reform. Change may therefore be large or small and may occur inside an organization or within an industrial sector. Although the Meyer et al. typology does not capture every component of change, it is revisited in the discussion section of the paper as a vehicle for comparison for two reasons. Firstly, in our view the typology is the most comprehensive available. Secondly, the similarity between the results that emerged and the typology were conspicuous.
Incremental or "first-order" changes that occur within organizations assume that organizations adjust and respond to their changing environments constantly and deliberately. On the other hand, researchers such as Pettigrew (1985); Laughlin (1991); and Greenwood and Hinings (1996) propose sophisticated conceptual diagnoses for understanding and affecting change. While adaptation theorists are concerned with incremental intra-organizational change, metamorphosis theories (Meyer et al., 1993) concentrate on radical change within organizations. In other words, one considers measured change, the other sweeping. The fact that adaptation and metamorphosis theories compete suggests that organizations tend to be stable and inert, but occasionally undergo substantial transformation. A third category of theory focuses on incremental change within industries rather than individual organizations, and can be identified as evolutionary (Meyer et al.) in orientation. For example, population ecologists suggest that change comes about as a consequence of Darwinian-like selection where industries gradually evolve to match the constraints of their environmental context. The implication is that successful managers copy the behavior of already successful organizations because resource support from the institutional context is implicitly directed to homogenous organizations. The key to this theoretical standpoint is that organizations are coerced into change by pressures from within their institutional environment.
Revolutionary change theory (Meyer et al., 1993) emphasizes second-order or radical change within industries. Revolution theories such as the punctuated-equilibrium model (Anderson & Tushman, 1990; Kimberly & Miles, 1980) propose that industries are massively restructured periodically in between periods of stability. The distinctions highlighted by these theories helps to frame the nature of change attempts.
Change Management and Sport Research
Research examining the structural features and strategic approaches of Canadian National Sporting Organizations (NSO's) by Hinings and Slack (1987); Kikulis, Slack, Hinings, and Zimmerman (1989); and Thibault, Slack, and Hinings (1993) were the first studies to define and typologize sport organizations on the basis of complementary structure and "function." Slack (1988) determined that structural elements in organizations are closely aligned with their core values. Similarly, Kikulis, Slack, and Hinings (1992) and Cousens (1997) have used the archetype concept (Greenwood & Hinings, 1988) to differentiate organizations on the basis of institutional values and structural features, thus reinforcing the notion that organizational values and structural features are related. These studies tend to view change as an evolutionary process undertaken from within the firm. In the nomenclature of Meyer et al. (1993), they emphasize the importance of "adaptation" change theories.
Kikulis, Slack, and Hinings (1995b), investigated the response to institutionalized pressures for change in Canadian NSO's. They discovered that there is a tendency for organizations to maintain the status quo when existing organizational structural designs are maintained or reinforced. The same team went on in an additional study to reveal that the structures governing decision-making are strongly linked to core organizational values, which in turn have an impact on structural design (Kikulis, Slack, & Hinings, 1995a). They also observed that the degree of fit between structure and operational systems could influence whether an organization undergoes incremental or rapid change. In another study building on these results, Amis and Slack (1996) discovered that there is a relationship between size and structure: where the larger the organization gets, the more centralized, standardized, and specialized it becomes, although the strong presence of volunteers adds some complicating dimensions. These studies recognized the role of industrial pressure in the change process, and help explain the occurrence of evolutionary and revolutionary change.
In Australia, Skinner et al. (1999) investigated the change actions of the Queensland Rugby Union Association using the Laughlin (1991) concept of external jolts as a theoretical framework. They noted that crises or jolts are powerful drivers of change at both a strategic and structural level. Stewart and Xu (1998), using the Thibault, Slack, and Hinings (1990) model, classified a range of state sporting associations on the basis of structural features and strategic practices. They concluded that strategic practices affect structural design, as had Slack and Hinings (1989) for Canadian NSO's. Finally, Smith and Stewart (1995) observed that organizational values are the pivotal dimensions driving significant change in their analysis of a professional sporting club. These studies identified the presence of metamorphosis change in sport organizations.
Research providing a strong influence on this study was undertaken in Canada by Slack and Hinings (1992). Recognizing the potential in employing a range of theoretical perspectives to the study of change in Canadian NSO's, the authors examined the change process using resource dependency theory, institutional theory, organizational culture, and the role of transformational leadership. They determined that the application of different perspectives facilitated the construction of a more complete picture of the change process, where adaptation, metamorphosis, evolution and revolution all play a role. Slack and Hinings (1992) provided clear guidance concerning two issues relevant to this study. First, the use of more than one theory to examine change provides a more comprehensive understanding of the change process, and second, while integrating different theories of change can yield positive results, we should not be blinded by the need for a single, unified theory.
This paper employs an approach consistent with that undertaken by Slack and Hinings (1992) in Canada, in that it views establishing a range of theories that might be useful in unraveling the change process in sport as critical. Notwithstanding the methodological differences where Slack and Hinings (1992) used several theoretical perspectives as vehicles for quantitatively testing theories explaining the change process, and this study a qualitative approach, both studies implicitly return to the significance of identifying and employing more than one theoretical perspective.
In summary, research in sport organizations has provided some firm elements of change management content such as the relationships between structure, strategy, and values, but remains limited in its ability to describe and explain change. In addition, generic research in change management, while vast, has failed to incorporate the sport industry into its results, and potentially to benefit from its practitioners. This study attempts to contribute to filling this gap. The next section outlines the method employed.
The broad aim of this study is to explore sport management practitioners' perspectives of organizational change. Specifically, it seeks to:
1. Develop grounded theoretical categories of organizational change experienced by managers.
2. Consider the relationship between these grounded categories, and those previously established and common in literature.
In qualitative research, significant attention is paid to the research assumptions, and the subjective views of respondents. This interpretive approach holds that people's individual and collective thinking and action has a meaning which can be made intelligible (Minichiello, Aroni, Timewell, & Alexander, 1995). In other words, an interpretive approach seeks to explain the behaviors of people in terms of the meaning it holds for them. Interpretive research retains the positivist assumption that the goal of research is to describe and explain reality without a value bias. However, unlike the positivistic approach, the interpretive viewpoint rejects the possibility of creating generic laws (Bain, 1989). As a result, this research focuses on the perceptions, opinions, beliefs, and practices of individuals, and the assigning of these views with an underpinning meaning.
It was essential to select a research protocol that generated a rich variety of data about managers' perceptions of change, to maximize the possibility of revealing their conceptual views. In order to explain perceptions, it was essential to avoid bias by collecting data using a predetermined conceptual vehicle. It was determined that a grounded qualitative approach allows for the development of theory and category based on the nature of the data (Armenakis & Bedeian, 1999), thereby avoiding the effects of deductive pre-determination or classification of the data. The techniques of grounded data analysis require that concepts are not "imported" to aid in the analysis or interpretation of data (Eaves, 2001). In other words, we were seeking to avoid confirming or rejecting a single theoretical change management perspective, and therefore adopted a grounded research paradigm that allowed the data to be the sole expression of the theoretical possibilities.
Population and Sampling
Australian sport is characterized by a tiered system built from a club foundation. Club representatives form State Sporting Organizations (SSOs), which manage state development and infrastructure. In turn, representatives from each state form National Sporting Organizations (NSOs), which subsequently manage the sport from a national perspective. National league competitions are variously formed, some under the control of the respective NSO, and others as independent, club-based entities. Clubs competing in these competitions are largely membership based, but some are privately owned (Westerbeek, Shilbury, & Deane, 1995).
The population for the sample comprised senior managers from Australian NSOs (approximately 100), SSOs (approximately 80 per state or territory within six states and two territories in Australia), and professional clubs participating in national league competitions (approximately 150). This research focused on those sports managers who are involved in the day-to-day operation of an organization, are key decision-makers, and have roles of authority and responsibility. Two case-selection procedures were employed consistent with theoretical sampling principles (Minichiello et al., 1995). Sampling categories were initially selected on the basis of their prominence in literature, and the remaining cases were "discovered" as data collection proceeded; their relevance to the research question arising as the data was analyzed. Specifically, we sought to find research evidence suggesting variables that might mitigate the perceptions and practices of sport managers concerning change. When a variable was numerously highlighted by research evidence, it was included in the theoretical sampling frame. As the purpose of the sampling process was to introduce diversity in order to encourage the constant comparison of data, it was determined that the discovered variables would also be employed in the sampling process if the data illustrated the effect of the variable.
A review of literature established three factors that were viewed as influential upon sport managers' perceptions of change. These constituted purposeful sampling categories. The first purposeful sampling category was the reward system in place; sport managers may be paid or unpaid for their services (Auld, 1994; Cuskelly, 1994; Kahn, 1993; Love, 1993; Moore, 1993; Thibault, Slack, & Hinings, 1993; Watt, 1992). The second was the educational level the sports managers achieved, specifically whether they are tertiary (university) trained or not tertiary (university) trained (Hogg, 1989; Moore, 1993; Parkhouse, 1981; Watt, 1992). The final sampling category concerned financial resources. This category was differentiated into two components: organizations with a gross annual income of greater than one million dollars (represented as > 1 in Table 1) and organizations with a gross annual income of less than one million dollars (represented as < 1 in Table 1) (Ferguson, 1995; Frisby, 1988; Kikulis et al., 1989; Mills, 1994; Slack & Hinings, 1989). These factors were seen to be suggestive, rather than definitive, and therefore did not represent a conclusive explanation of the factors influencing managers' perceptions of change. On the other hand, these categories were demonstrably exposed by the literature.
A respondent fitting the first set of criteria as determined by the theoretical sampling frame (paid, tertiary trained, employed by an organization with revenues exceeding one million dollars) was contacted to participate in the study. This was undertaken on a random basis, constrained by convenience in that the list devised of eligible managers was limited to the three Australian states (Victoria, New South Wales, and the Australian Capital Territory), which house the majority of sport organizations. In addition, the pre-conceived list did not always accurately predict the training received by the potential respondent, in which case another was chosen until the criteria matched that desired. Consistent with the analytical induction method described in further detail shortly, the second respondent was selected from the other end of the sampling frame (unpaid, not tertiary trained, from an organization with revenues less than one million dollars). This process continued for the first eight interviews.
The data collected not only substantiated the use of the three sampling categories, but also identified another element that impacted upon managers' perceptions of change: whether funding was from public (government) or private (generated by the organization) sources. Based on data accumulated during the initial eight interviews, this category was assessed in terms of whether the organization received a significant amount of its funding from the government. Organizations which received more than forty percent of their income from the government were considered to be government dependent, while those receiving less than forty percent were seen to be "self funded." In Table 1, government dependent organizations are demarcated with an 'S', indicating a significant amount of funding. Equally, those with less than forty percent are demarcated as 'NS', indicating that the funding government provides is not significant. The forty percent threshold was derived from the first eight respondents, and represented an "average" of respondents' opinion. At the same time, this figure was seen as flexible, and although unnecessary, could have been altered at a later stage of the interview process in response to additional views and opinions. In total, twenty-nine interviews were undertaken, including at least one from each category. As 16 choices were available, the comparative process continued until saturation was reached, and it was believed that no new data was being uncovered. The sixteen possible categories are illustrated in Table 1.
It must be acknowledged that the sampling categories do not perfectly separate the sample into even and equally diverse components. However, the purpose of the model was to establish diversity, and to facilitate the constant comparative process. It should also be remembered that this sampling model does not pretend to be exhaustive or complete, and nor is it a perfectly representative sample of the population of sport managers.
The perceptions of a sample of sports managers were collected using semi-structured, in-depth interviews. Informants provided detailed descriptions of their administrative experiences in sport and revealed how those experiences affected their perceptions of change. Semi-structured interviews begin with broad interview schedules, but relied substantially on the interaction between interviewer and informant in order to gain information (Minichiello et al., 1995). They took the form of conversation between the interviewer and the respondent, and according to Burns (1994) they focused on the informant's perception of himself or herself, of his or her environment, and of his or her experiences. No questions were asked directly from an interview schedule, but rather the medium for information gathering was largely free-flowing conversation, relying on the interaction between the researchers and respondents. The onus was chiefly on the researchers to use their interpersonal skills to subtly direct the conversation toward the topic at hand (Fontana & Frey, 1994) without compromising the respondents' independence and sense of control.
The recursive model of interviewing was employed. With the recursive technique, open-ended questions are asked, encouraging a broad range of information to be brought forth. Specific questions are subsequently asked to narrow the field of inquiry. Questions become more specific with each successive interview (Minichiello et al. 1995).
Interviews were conducted based on the analytical induction method, an approach consistent with the constant comparison approach implicit in grounded theory. The analytical induction method incorporates the following processes: (1) conduct first interview; (2) record data; (3) analyze data via recognizing and coding dominant themes, issues, and concepts; (4) return to the original question and analyze it in light of the results from the first interview; (5) select a respondent who would likely have an alternative viewpoint; (6) conduct interview; (7) analyze interview and re-analyze first interview in light of second interview; (8) formulate, revise, and extend proposition until original question becomes more focused; (9) continue cycle, becoming more deductive; and (10) develop proposition to the stage where no new information is forthcoming, and saturation is achieved (Minichiello et al., 1995).
All interviews were audio-recorded, transcribed, and subsequently coded. Interviews were transcribed "word for word" into a computer database for use in subsequent analysis. Data obtained from interviews were systematically organized and classified into "chunks" of varying size; words, phrases, sentences, or whole paragraphs which were linked by common themes or issues (Denzin & Lincoln, 1998; Lee, 1998).
Coding requires the systematic examination of units of data. In this study the units of data were change incidents, which were subsequently de-constructed into categories that described the perceptions of sport managers. These categories were examined to reveal sub-categories and dimensions that emerged, and were systematically coded and compared. This process culminated in the specification of a number of core categories that encapsulated the change perceptions of sport managers, and represented a tight integration of the concepts in the data collected. These core categories form the basis for the results discussed later.
The transcripts were first broadly studied to gain a general familiarity of the contents. During this process, dominant concepts, themes, and issues were noted to form categories; these categories became the codes from which the transcript was interpreted and meanings were developed. Thus, coding by using keywords was employed in order to categorize and classify the text. These codes reduced the data and highlighted trends, themes, patterns, and causal processes.
Coding was undertaken in three forms, as recommended by Strauss and Corbin (1994). They were: open coding, axial coding, and selective coding. These three coding methods are not different, but sequential and interrelated; each builds upon the previous. In this way, the lines between each type of coding are artificial (Strauss & Corbin, 1990). Thus, in a session of coding, it was possible to move between one form of coding and another. The reason for this, as Strauss and Corbin (1990) appreciated, is that during the final or selective stage of coding, there are always some concepts that remain undeveloped and ambiguous. At the same time, during the initial or open stage of coding, some concepts will naturally progress to the more developed stages, achieved during selective coding.
Open coding is the process of breaking down, examining, comparing, conceptualizing, and categorizing data (Strauss & Corbin, 1990). Open coding was initially used, and the interview transcripts were studied and codes were assigned to every piece of information. Every sentence spoken by the respondents was classified into broad categories. According to Strauss and Corbin (1990), a category is a classification of concepts. Concepts are labels placed on discrete happenings, events, and other instances of phenomena. In this initial phase, the categories were broad and inclusive, rather than specific and selective.
The purpose of this initial coding stage was to reduce the data into a more condensed form, allowing definitive and apparent categories to emerge. Once particular phenomena were identified in the data, concepts were grouped around them. This was done in order to reduce the volume of data. Thus, the codes broke the data into more manageable "chunks" or preliminary concepts; ready for subsequent and more precise reduction when concepts were grouped into categories. Thus, the coding process moved in ascending order, combining and reducing data from the "raw" interview transcripts, which when grouped according to similar themes became concepts, which in turn when combined, became themes. Whenever concepts, categories, or themes were applied they were given an abstract and broad conceptual name.
Open-coding was undertaken on three levels. First, at the broadest level, an entire transcript was read and then the question asked: "What are the general themes coming from this interview?" Having answered this question, the document was then reviewed by paragraph. Again, the same question was asked: "What are the general themes coming from this paragraph?" Finally, at the most detailed level, interview transcripts were analyzed on a line-by-line basis. Here, the transcript was examined closely, looking at phrases and specific words. While this method was tedious, it ensured that no categories or concepts were overlooked, which was particularly important because some categories provided evidence for the introduction of additional sampling categories.
The interview transcripts were physically separated into the initial categories where all data from particular categories were relocated. By the conclusion of the open coding process, the twenty-nine interview transcripts were transformed into three open codes, namely, Types, Techniques, and Context. These open codes were re-located into separate computer directories and printed in order for the next stage of coding to be undertaken.
Axial coding was undertaken next. Using the initial codes, the axis of the key concepts was identified. In other words, deeper patterns, strategies, categories, and concepts were identified from the initial codes and sub-categories/codes were developed. Further computer files were established for each of the new codes within the existing codes, creating sub-directories resembling family trees. From the first three open codes, nine sub-codes were created, which were subsequently printed. These were named: Fast, Slow, Inside, Outside, Accidental, Action, Culture, Background, and Commercialism.
Finally, selective coding was used to refine the existing codes and sub-codes. This form of coding was the most specific, and was geared towards generating the most precise themes. Twenty-eight selective codes were developed, namely: Performance, Crises, Facilitators, Adaptation, Timing, Context, Discharge, Filtering, Continuous, Strategic, Ongoing, Policy, Organizational, Systems, Procedures, Catalyst, Sudden, Regulation, Chaos, Structure, Strategy, Quality, Tradition, Institutionalization, Tertiary, Competition, Stakeholders, and Power. The coding structure is reproduced in Table 2.
According to Minichiello et al. (1995), typologizing is a method that researchers commonly use to understand phenomena more completely by grouping ideas and then forming ideal types which conceptualize situations that have similar or different characteristics. In other words, it is a method of making sense of complex or abstract ideas. Ideal types are termed so because they do not exist in reality, but instead are intellectual constructs which represent reality for the purpose of understanding reality. Minichiello et al. noted that ideal types do not generate knowledge. In fact they are tools which help researchers ask particular questions and formulate useful propositions. Thus, the purpose of typology construction was to find patterns amongst the codes. In other words, while the codes provide substantial insight into the perceptions and practices of respondents, they do not fully explicate the relationships that exist between them. To this end, as the codes emerged, their relationships and the patterns created through the imagining of ideal types were considered. For example, several of the axial codes suggested a pattern when combined, forming a parsimonious division between the types of change experienced by respondents. These ideal types were pursued and typologized.
As a final word on the methodological process, it is noteworthy that the process was not linear, as any systematic explanation inevitably implies. The process of data analysis and typology building was concomitant, evolving, and emerging with new data as it was discovered.
Typically, qualitative researchers want to ensure that their findings are confirmable, dependable, credible, and transferable (Anfara, Brown, & Mangione, 2002; Creswell, 1994; Krefting, 1991). Several research processes are noteworthy concerning these issues. Dependability (or reliability) in qualitative measurement was addressed by noting the methods of recording data and the use of verbatim accounts of interviews. Two researchers were involved in coding interview data. "Check-coding," a technique where the researchers separately code the same data and subsequently come together to compare codes, was employed to enhance reliability (Miles & Huberman, 1994). A working reliability score was calculated by dividing the total number of coding agreements by the total number of agreements plus disagreements. This process was performed intermittently until reliability reached a satisfactory level. As recommended by Miles and Huberman, inter-coder agreement was considered reliable when it reached 90%.
Establishing credibility (internal validity) and transferability (external validity) of the data was attempted by seeking to find exceptions to the conventional interpretations determined by researchers. A limited amount of rudimentary triangulation was employed to facilitate this. For example, organizational documents were reviewed as a simple form of data triangulation. In addition, member checks, or the process of verifying the findings with the research participants themselves, were also employed in order to bolster the credibility of data. Special attention was paid to respondents' demographic characteristics in order to provide a platform for transferability and to guide future researchers in (potentially) quantitative efforts.
There are several important assumptions that this research has embraced. First, it takes a perspective consistent with Slack and Hinings (1992) in that a range of theories associated with change might be useful in unraveling the change process in sport, rather than one pivotal theory, or even no common theoretical ground. In this sense, it is important to acknowledge that the researchers anticipated more than one theoretical approach to emerge from the data, which could potentially have affected coding. Secondly, this research assumes a fundamentally systems-based view of change, in that it is viewed as a concept that cannot be isolated from other forms of organization; change is pervasive and does not affect organizational components independently (Greenwood & Hinings, 1996; Kotter, 1995). This assumption encouraged the researchers to seek comprehensive data about the occurrence of organizational change and also to look for relationships between organizational elements during coding.
In any research it is necessary to acknowledge the limitations intrinsic to the system of data collection and analysis used. As this research was qualitative in method, there are a number of limitations associated with qualitative research that must be highlighted. First, it is clear the great strength of qualitative strategies is their usefulness in uncovering depth of response. The corollary of this is that qualitative research is severely limited in its ability to cover breadth. In other words, only small segments of the population can be used in the sample. Subsequently, qualitative data does not lend itself to statistical manipulation. As a result, data obtained from qualitative sources are subject to lengthy analysis and discussion, only rarely being statistically representative in any way. In the case of this research, the data cannot be statistically interpreted, and cannot be summarized simply and efficiently in numerical form.
Another consequence of this style of research is the relative importance of sample selection. As there are comparatively few respondents, they must be chosen appropriately, providing as much information into the research question as possible. Much of this difficulty is controlled with the proper use of a suitable sampling technique. As theoretical sampling was used in this research, the limitations of that process must be recognized. Theoretical sampling hinges upon the original and discovered sampling categories selected. If at first, they are chosen incorrectly, they can be modified depending on the data obtained from the initial interviews. However, this process is demanding, requiring great care, and analysis.
Results and Discussion
In this section of the paper the respondents' perceptions of change are presented according to the typologization determined as a result of coding. In addition, these results are compared with the four dimensional typology proposed by Meyer et al. (1993): adaptation, metamorphosis, evolution, and revolution. This typology proved to dovetail strongly with the results, which can be explained by the respondents desire to describe and understand change in terms of its size and its impetus; the same two determinants used by Meyer et al. in their typology (mode and level). A fifth category of responses that emerged from this study is described as "other," for those results that did not fit neatly with the typology, and which provides some key evidence that the established typology requires some expansion.
Perceptions of Change
Change is a modification of anything occurring within an organization (Bridges & Roquemore, 1996). Consistent with this definition, all respondents agreed that change is continual and ongoing, and that all organizations are in the process of change. Notably, however, respondents made a clear distinction between fast change and slow change. Fast change typically occurs when sport organizations are overtly under-performing, often to the extent of endangering their long-term survival. To put it simply, in these instances changes are made in direct response to crises. In contrast, slow changes are "more lazy" and are of the kind that continues indefinitely. These "lazy" changes occur almost every day and arrive in the form of minor alterations to existing policies and practices. A third form of change described by respondents did not neatly fit into either the slow or rapid categories. The third type was identified as "accidental" change, where modifications are made to existing operations, not through design or strategic intent, not through evolutionary improvement, not even through external pressures, but as a result of inadvertent and unintended circumstance.
The concept that external pressures and crises may instigate change is not new, and has been demonstrated by Skinner, Stewart, and Edwards (1999) in Australian sport. Similarly, the notion of incremental change is a consistent inclusion in change theories (Pettigrew, 1985). However, the idea that change can occur as a consequence of accident has not commonly been recorded.
Respondents in this study specified that "accidental" change is a frequent occurrence. New practices and processes are periodically introduced, without the consent or knowledge of an organization's senior managers. While senior managers attempt to direct changes to fit with strategic goals, they admit to a degree of powerlessness over these random, unexplained, and sometimes hidden events.
One example involved a change in the way club merchandise was distributed. Although the "standard" policy was to sell club merchandise from the club only, as a result of an "accidental" change where a mail order was accepted and discharged promptly by an unknowing work placement student, sales of merchandise almost doubled as additional mail orders arrived. These "accidental" changes are characterized by the difficulty in tracing their source and consequences. They may result in either positive or negative outcomes and are not always overtly related to overzealous students. Such changes can also be easily concealed by an organization's institutionalized practices, particularly when no one can explain why a certain procedure is employed, or how it first began.
Respondents frequently lamented that while change had to be dealt with, they were uncertain about the best way of dealing with it. In the first instance, it was commonly admitted that a "tried and tested" approach to change is noticeably absent from the operations manuals of most sporting organizations. This lack of information concerning the best methodology for initiating and sustaining long-term changes is indicative of a general confusion about the two fundamental elements in any change program: what to change and how to change. Thus, not only are sport managers unclear about where to direct their energies in order to initiate change, they are also hesitant when it comes to nominating their preferred tools and techniques for managing, directing, and controlling change.
This confusion is exacerbated by a widespread fear that even mentioning change will be met by an outcry from constituents such as associated clubs, spectators, and members. Thus, proposals to modify traditional ways of operating are often considered with the clandestine sensitivity normally reserved for international espionage! Several respondents observed that while the mechanisms for change and the institutionalized obstacles preventing change may be similar in sport in contrast to commercial operations, the way change must be handled is entirely different. They pointed out, for example, that the stakeholders in commercial businesses demand constant change in order to optimize competitive advantage, while in sport organizations, the stakeholders tend to insist that change will compromise success or worse, destroy a long held tradition. Nevertheless, despite a reluctance to contemplate reformation, one-third of respondents observed that many sport organizations are vulnerable in the highly competitive sport marketplace without a substantial shift in their practices and policies.
Theories of Change
Respondents complained that while change was frequently necessary, they were uncertain about the best ways of approaching it. In the first instance, it was generally admitted that a "tried and tested" approach to change is missing from their theoretical knowledge of sport and its management. This lack of information concerning the best methodology for initiating and sustaining long-term changes is indicative of a general confusion about the nature of change. Respondents were capable of clearly discussing issues such as the nature and magnitude of change they have experienced and observed.
The most significant result was that respondents did not exclusively propose one theoretical model of change. All 29 respondents insisted that change could not be explained within one concept. However, there was a consensus that change can occur rapidly or slowly, within an organization or within the environment. These codes were employed to form a typology, which forms the basis of discussion here. Each typological category reflects significant differences held by respondents concerning the degree to which fast, slow, internal, and external change is embraced as a pivotal catalyst in the change process. Even a cursory glance at these "grounded" categories reveals their similarity with the Meyer et al. (1993) typology. Although this research was exploratory, the emergence of these categories might suggest that potential confirmation may be found, with further research in the sport industry, of the theoretical perspectives conflated by Meyer et al. However, a further category emerged from this study that provided a fifth typological category that is not encapsulated by the Meyer et al. model.
Data revealed a number of consistencies in the units of analysis respondents employed. A demographic profile also accompanies each section to highlight the shared characteristics of individuals proposing common ideas.
Slow / internal change
This approach to change was the most popular with managers working within larger, well-resourced organizations. Specifically, the organizations that employed respondents who mainly perceived change within adaptive terms were large, financially secure, and independent. These organizations predominantly comprised clubs participating in national league competitions as well as several national sport associations. All respondents who fit into this theoretical approach were paid, and tended to be tertiary trained.
Respondents who used these theories to explain change agreed that change is continual and ongoing, and that all organizations are somehow in the process of change. Ongoing changes occur almost every day and arrive in the form of minor alterations to existing policies and practices. Thus, these respondents took the view that the mode of change is generally slow. In the terminology offered by Kanter, Stein, and Jick (1992), their chief experience of change has been of the "long march" variety. Change is slow, incremental, and continuous. In addition to the continuous mode of change, the level of change was viewed as intra-organizational. Respondents' experiences of change were naturally continuous and occurred independently within an organization rather than being imposed by the environment or broader industry. In short, they perceive change as internally driven through processes. One respondent epitomized this view in the following comment concerning their change activities:
Change comes about because we want it to. This idea that it's someone else or something else is a bit of a cop-out. Change here is fundamentally systematic. Bit by bit, we make changes as things <