xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003419
Extending situational crisis communication theory :
b attitude and reputation following the 2004-05 nhl lockout
h [electronic resource] /
by Melanie Formentin.
[Tampa, Fla] :
University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains X pages.
Thesis (M.A.)--University of South Florida, 2010.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
ABSTRACT: This exploratory case study positions the 2004-05 National Hockey League (NHL) lockout as an organizational crisis, studying it within a Situational Crisis Communication Theory (SCCT) frame. A mixed methodology approach used qualitative and quantitative content analyses and a survey of NHL fans to gauge the NHL's reputation five years after the lockout. For the content analyses, 282 newspaper articles from 7 newspapers in the U.S. and Canada were coded for SCCT variables and presentation of the lockout by news writers. NHL fans (n = 140) were surveyed with the goal of assessing SCCT variables as predictors of attitude. Results confirmed previous SCCT findings and showed links between SCCT variables and fan attitudes toward the NHL and the sport of hockey.
Advisor: Kelly Page Werder, Ph.D.
National hockey league
x Mass Communications
t USF Electronic Theses and Dissertations.
Extending Situational Crisis Communication Theory: Attitude and Reputation Following the 2004-05 NHL Lockout by Melanie Formentin A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts School of Mass Communications College of Arts and Sciences University of South Florida Major Professor: Kelly Page Werder, Ph.D. Kenneth C. Killebrew Jr., Ph.D. Randy Miller, Ph.D. Date of Approval: April 13, 2010 Keywords: reputation management, nationa l hockey league, communication managers, fans, professional sports Copyright 2010, Melanie J. Formentin
Dedication This thesis is dedicated to my mom and best friend, who so patiently and lovingly supported me throughout.
Acknowledgements Many thanks to Dr. Kelly Page Werder for her significant help w ith this study. Also, many thanks to Drs. Ken Killebrew and Randy Miller for their continued support of this ongoing project. Finally, thanks to the USF School of Mass Communications for providing the engaging environment that makes theses like this possible.
i Table of Contents List of Tables iii Abstract iv Chapter One: Introduction 1 The 2004-05 NHL lockout 3 Research Questions 6 Hypotheses 6 Chapter Two: Literature Review 8 Crisis Communication 8 Situational Crisis Communication Theory 11 Using Situational Crisis Communication Theory 13 Chapter 3: Method 20 Content Analysis 23 Categories and Coding 28 Survey Research 29 Materials 29 Measures and Design 31 Additional Surveys 36 Chapter 4: Results 39
ii Content Analysis 40 Research Questions 42 Fan Survey 49 Research Questions 50 Test of Hypotheses 52 Chapter 5: Discussion and Conclusion 58 Discussion 59 Conclusion 68 Limitations 73 Future Research 75 Recommendations 77 References 79 Appendices 88 Appendix A: Content Analysis Codebook 89 Appendix B: Fan Survey 98
iii List of Tables Table 1 Intercoder Reliability Analysis 28 Table 2 Newspaper Frequencies, by Date 41 Table 3 Newspaper Frequencies, by Source 41 Table 4 Newspaper Frequencie s, by Word Length 42 Table 5 Newspaper Frequencies, by Story Placement 42 Table 6 Newspaper Frequencies, by Article Type 42 Table 7 Overall portrayal of the NHL lockout 43 Table 8 Newspaper Analysis, Re putation Variables 44 Table 9 Discussion of labor history 45 Table 10 Respondent Demographics, Age 50 Table 11 Reliability Analysis for Variable of interest 53 Table 12 Regression Model for Fan Perception of SCCT Reputational Threat Variables Predicting Attitude Toward NHL 54
iv Extending Situational Crisis Communication Theory: Attitude and reputation following the 2004-05 NHL lockout Melanie Formentin ABSTRACT This exploratory case study positions th e 2004-05 National Hockey League (NHL) lockout as an organizational cris is, studying it within a Situati onal Crisis Communication Theory (SCCT) frame. A mixed methodology approach used qualitative and quantitative content analyses and a survey of NHL fans to gauge th e NHLs reputation five years after the lockout. For the content analyses, 282 newspaper articles from 7 newspapers in the U.S. and Canada were coded for SCCT variables and presentation of the lockout by news writers. NHL fans ( n = 140) were surveyed with the goal of assessing SCCT variables as predictors of attitude. Results confirmed previous SCCT findings and showed li nks between SCCT variables and fan attitudes toward the NHL and the sport of hockey.
1 Chapter 1 Introduction Like other unionized organizations, the profession al sports industry has experienced its share of strikes and lockouts in the previous two decades. Major League Baseball (MLB) lost an entire post-season to a labor stri ke in 1994, while the National Hockey League (NHL) stumbled through its first lockout in 199495, losing half of a hockey season to an ongoing labor dispute. However, it was the 2004-05 NHL lockout that resulted in the longest stoppage of play experienced by a major North American sports league. A strike is considered the primary bargaining power of organized labor, (Schmidt & Berri, 2004, p. 344). During a stri ke, workers withdraw their labor or services from an organization. The objective is to impose costs upon the organization and gain an upper hand during contra ctual negotiations. A strike is different from a lockout because, during a lockout, laborers are prevented from working at the choosing of the organization. Sports provide an interesting contex t through which to study labor relations. Unlike traditional unions, the players associ ations found in professional sports are not utilized to negotiate indi vidual salaries (Rosen & Sa nderson, 2001). Instead, collective
2 bargaining agreements are put in place to help standardize working conditions, pension benefits and insurance, grie vance procedures and league -wide arrangements such as a minimum salary, any direct restrictions on to tal payrolls or indivi dual salary caps, or owners incentives to compensate players, (p. F63). This results in disagreements between owners groups and players associat ions in regards to revenue distribution. Historically, players associations oppose at tempts to cap payrolls while owners often fight to restrict excessive player movement. The basic labor agreemen ts typically last for three to seven years prior to being reopened for negotiation. This becomes especially important in the big business of sports. Revenue can hinge on player salaries, franchise values, and stadium costs (Rosen & Sanderson, 2001). The collective agreements between ownership and player groups provide outlines that control revenue sharing and free agency, strikes and lockouts, player agents, endorsements, product licensing, and even me dia partnerships such as television broadcasting deals. All of these factors and moreaffect the bottom line of sports leagues, individual teams, a nd even individual players. For the NHL, the 2004-05 lockout represente d a labor disagreement that arguably spun out of control. When the NHL officia lly announced the beginni ng of the lockout, it left a somber impression of the state of the ongoing collective ba rgaining agreement negotiations. Although it has been suggested that strike cost s are limited to the strike period and that consumer demand returns in force immediately after the strike ends (Schmidt & Berri, 2004, p. 345), one could argue that even the NHL realized it wasnt facing a standard labor disagreement.
3 The 2004-05 NHL Lockout The NHL faced numerous popularity issues in the United States at the time of the lockout (Batchelor & Formentin, 2008). Atte ndance numbers were falling, a television contract with ESPN was in limbo, and teams were claiming large revenue losses due to rising player salaries. Further damaging th e image of the NHL was the notion that contract bargaining negotiations were stalled because of salary cap issues (Podnieks, 2005). While the league was calling for a salary cap tied to a percenta ge of overall league revenue, the National Hockey League Player s Association (NHLPA) was fighting the idea of limiting the amount of money a team c ould spend on its players. As the two sides continued to battle each other well into th e canceled season, public perception was that greed was at the root of the CBA-related problems (Toronto Star, 2004). With the NHLPA vehemently opposing a salary cap, the NHL argued that imposing such a cap was necessary to create cost certainty and align player salaries more closely with team and league reve nues (Podnieks, 2005). At one point, NHL commissioner Gary Bettman argued that imposing a salary cap would drop the average player salary from $1.8 million to $1.3 million. Additionally, average player salaries had risen from $733,000 in 1994-95 to $1.8 million in 2003-04 (Woods, 2004a). With numbers like these, it seemed understandable that the average sports fan might be unimpressed with the NHLPAs position on the issue. In an interview with Peter Mansbridge on the Canadian Broadcasti ng Corporations (CBC) Newshour show, The National NHLPA executive director Bob Goodenow was challenged by one fan who
4 asked, Do the players know the average salary of a National Hockey League fan? (cited in Podnieks, 2005, p. 42). Further exacerbating the tension between the two sides was the fact that Goodenow was publicly steadfast in his opi nion opposition to a cap (Podnieks, 2005). Despite early reports that some players started calling for changes in Goodenows bargaining tactics, it took nearly five months for the NHLPA to consider a salary cap. At the same time in February 2005, the NHL con ceded its position on linkage, or tying salaries to overall league revenues. The effort was too late. Even with increasing negotiations between the sides, Bettman proclaimed the entire season would be canceled if a deal wasnt made, and no team would compete for the Stanley Cup for the first time since 1919 when an influenza epidemic wiped out the final series. By then, players and fans were furious that the s eason seemingly was lost for no r eason as each side moved off its entrenched position in the 11th hour. At that point, the main concern was simply to create a contract bargaining agreement that could save the NHL from further embarrassment and the potential cancellation of a portion or all of the following 2005-06 season. With the lockout affecting everyone from fans to league employees and cab drivers to television broadcast networks, the league was in a vulnerable state. It became important for the two sidesthe NHL and the NHLPAto communicate their positions. The league was facing a crisis situation, and public perception that greed was the underlying issue presented a challenge to both the NHL and NHLPA During the course
5 of the ten-month lockout, the NHL emerged as better prepared to deal with the public scrutiny it endured. This case study argues that reputation mana gement objectives should have been at the forefront of the NHLs communicati on strategy during the 2004-05 NHL lockout. Specifically, this study employs explorator y case study methodology founded in Situation Crisis Communication Theory (SCCT), including a newspaper content analysis and a survey of NHL fans, to identify how the league s reputation has change d in the five years since the 2004-05 NHL lockout. According to Coombs (2007), SCCT posits that crisis responsibility is direct ly correlated with reputation, thus crisis communica tion strategies should be based upon perceptions of crisis re sponsibility. The SCCT is used as the guiding framework for a content analysis of how the lockout was presented in major newspapersthus, identifying some of the in formation fans received about the labor dispute. Through use of a fan survey, SCCT is used to identify a link between fan attitudes toward the NHL and the reputat ion management variables of crisis responsibility, crisis history, a nd prior reputation/relationships history. It is suggested that SCCT variables can be used to assess the l eagues reputation in the years following the crisis. The following research questions and hypothe ses are addressed in this study, and are separated per the research method that informs them:
6 Research Questions Content Analysis RQ1: Was the 2004-05 NHL lockout presented by major newspapers as an organizational crisis? RQ2: How was the 2004-05 NHL lockout presented by major newspapers in terms of SCCT variables? RQ3: What strategic communication strate gies did the NHL use during the 2004-05 lockout, as presented by major newspapers? Survey RQ1: Do fans have a positive per ception of the NHLs reputation? RQ2: Has the NHLs reputation improved in the five years since the lockout? Hypotheses H1: NHL crisis responsibility, crisis hist ory, and prior reputation influence the NHLs current reputation. H2: Hockey fan perception of reputat ional threat variables (crisis responsibility, crisis history, and prior reputation) related to the 2004-05 NHL lockout are related to fa n attitude toward the NHL. H3: Hockey fan perception of reputat ional threat variables (crisis responsibility, crisis history, and prior reputation) related to the 2004-05 NHL lockout are not related to fan attitude toward hockey.
7 H4: Among hockey fans, perception of high player responsibility for the 200405 NHL lockout is inversely rela ted to perception of NHL/owner responsibility for the 2004-05 NHL lockout. H5: Among hockey fans, perception of high player responsibility for the 200405 NHL lockout negatively influenc es perception of NHL reputation. H6: Hockey fan attitude toward the NHL is less positive than fan attitude toward hockey.
8 Chapter 2 Literature Review Crisis Communications A variety of definitions have been app lied to the concept of a crisis, and many scholars have developed work ing definitions for their own purposes. In early crisis management literature, definitions focused on a crisis as a disruption that could threaten a system, or organization, both on the physical an d existential level (Pauchant & Mitroff, 1992). Crises have been labeled as forewarni ng situations that can escalate in severity and may impede operations and affect re venues (Fink, 1986). Media or government scrutiny can add to the stress of the crisis. Less measurablebut just as importantis the concept that organizational image and reputa tion may also be affected during a crisis. Noted crisis communication expert Fearn-Ba nks (2007) echoes these sentiments when suggesting that the key concepts in defining a crisis are the assumpti ons that business is interrupted and an organization s existence can be threatened. In the previous two decades, crisis management has emerged as a way for organizations to strategically prepare for cr ises they may face. Crisis management has been defined as the preparation and applicati on of strategies and ta ctics that can prevent
9 or modify the impact of major events on th e company or organization (Stocker, 1997, p. 189). In this sense, crisis management can be a predictive function in which an organization tries to plan ahead for potential crisis situations. Based on a review of public relations-related crisis communication literature, Hagan (2007) posited that organizations that scan the environmen t for potential problem areas and maintain positive relationships with the organization s various publics handle and survive crises better than do others (p. 437). One of the criticisms that may be directed at crisis management literature is that suggested steps for communication during a cr isis can vary from author to author. Stocker (1997) suggests that crisis manageme nt is organization-centered and starts with an analysis of the company and/or industry and moves towards asking whether a plan or preparation is needed. In contrast, audience-centered approaches suggest that stakeholders should be a top pr iority, a level of responsibility should be accepted, central information centers should be developed, news coverage should be monitored and key stakeholders should be communicated w ith (Wilcox, Cameron, Ault, & Agee, 2005). Despite the importance of crisis mana gement offered by both approaches, many companies are not prepared for a crisis situation even though many crises can be predicted (Mitroff, Pauchant, & Shrivastava, 1989). Many managers deny that a crisis may befall their organization, while others sugg est that a major catastrophe would have to occur before efforts are made to be proact ive crisis managers. This happens despite managers recognizing that th ey are significantly likely to encounter a crisis.
10 Even so, this has not stopped scholars from devising methods aimed at planning for and approaching crisis situations. One particularly relevant crisis management approach that leads naturally into post-crisis management objectives comes from Stocker (1997), who suggests that crisis manageme nt can be built from three sequential objectives. The first objective is that an organization try to prevent crisis. If the crisis occurs, the second objective is for an organi zation to turn its effo rts towards countering the negative effects the crisis may have. Fina lly, the third objective suggests that, through behavior, an organization must provide a platform for th e companys future (p. 191). In recent years, public relations scho lars have started looking at crisis management from a public relations and re putation management perspective (Hagan, 2007). Within that time frame, communication schol ars have also started to look at crisis management as a symbolic approach (Coombs, 1998). This perspective places an emphasis on how communication can be used as a symbolic resource in attempts to protect the organizations image (Coombs, 1998, p. 177), which is consistent with Stockers (1997) second crisis manageme nt objectivemanaging or countering the negative effects of a crisis. As it has been developed by researchers, the symbolic approach rests on two assumptions: 1) crises present a threat to or ganizational image, and 2) the characteristics of a crisis situation influence the communi cative choices of an organization (Coombs, 1998). Crisis communication research has focu sed on strategies for concepts such as image protection and restoration (Benoit, 1997), contingency theories (Cancel, Cameron, Sallot, & Mitrook, 1997), and s ituational response strategies (Coombs & Holladay,
11 2002). A common criticism of the symbolic approach is that the literature is rife with case studies that offer more post-crisis situati on prescriptive recommendations than predictive theories (Coombs, 2007; Hagan, 2007). Although these rhetorical case studies have allowed communicators to focus on what to say or do based on given situations, it has also means that overall theory has remain ed relatively underdeveloped (Coombs, 2007). Despite these criticisms, one theory that has received support in the field of postcrisis communication is the Situational Cr isis Communication Theory (SCCT). In the mid-1990s, W. Timothy Coombs pioneered this symbolic approach, arguing that crises can be assessed on a situational basis along a continuum of response types that range from defensive to accommodative messages (Wilcox, et. al, 2005). SCCT addresses factors such as crisis respons ibility, prior history and prior reputation/relationship history. The case study approach dominantly found in SC CT literature will be maintained for this study, utilizing the NHL lockout as the specific example. However, as will be discussed, efforts will be made to expand the type of research and data collection methods traditionally used to analyze similar cases. Situational Crisis Communication Theory In 1995, Coombs synthesized existing crisis communication literature in an effort to make a list of existing cr isis response strategies. Co ombs (1995) defines crisisresponse strategies as public statements ma de after a crisis (p. 447) and worked to develop a set of guidelines to appropriately use the identified stra tegies. Positing that crisis-response strategies are an important symbolic resource for crisis managers
12 (Coombs, 1995, p. 447), Coombs also suggests that the crisis situation should be a major influence in strategy selection (p. 448). Coombs grounded his approach in attr ibution theory, arguing that the key characteristics of the theory complement the needs and uses of crisis communication (Coombs, 2007). Attribution theo ry suggests that people have a need to search for the reasons an event has occurredespecially unexpected and negative eventsand judgments about the event are made based on locus, stability, and controllability (Coombs, 1995; 2007). Locus is concerned with locus of control or whether the cause was internal or external to an actor (p. 448). Stability focuses on whether the event has happened more than once over time. Controllabi lity refers to whether the event was in or out of control of the actor. Thes e three attributes may vary acr oss crisis situations, leading to different response needs across cases. Str onger attributions of responsibility, frequency of event occurrences and increased harm to out side stakeholders invariably lead to more negative attitudes towards an organization. Development of SCCT focused on applying related concepts from attribution theory to a greater variety of cr ises (Coombs, 2007). As studies continue to be performed, SCCT has been developed as a way to give crisis managers a guiding framework for navigating crisis situations (Coombs & Holladay, 2002). Crises pose reputational threat s to organizations, a nd the notion is that protecting reputations can be a central focus of crisis communications. An organizations reputation is defined as the way stakeholde rs perceive an organization, and research suggests that reputation can be considered a v alued resource in and of itself. As such, SCCT acts as a mechanism for predicting how st akeholders will react to a crisis by using
13 an evidence-based framework for understa nding how to maximi ze the reputational protection afforded by post-crisis communication (Coombs, 2007a). In developing SCCT, a model is provided for developing a response system based on factors of the event or situation at hand. Through his research, Coombs (2007) suggests that proper assessment of the crisis situation and pr oper selection of a crisis response strategy can lead to better strategic co mmunication approaches that will protect an organizations reputation. Nume rous studies have been used to test and define SCCT, with the main assumption being that a relatio nship exists between crisis responsibility and organizational reputati on (Coombs & Holladay, 2002). Cr isis responsibility is defined as the degree to which stakeholders attribute responsibility for a crisis to an organization (Coombs, 2004, p. 268). A correla tional effect exists between the two concepts, with lower attributions of crisis responsibility resulti ng in less reputational damage in crisis situations (Coombs & Holladay, 1996; 2002; 2006). Likewise, if attributions of crisis responsibility increas e, the crisis poses increasing threats to organizational reputation (Coombs, 2004). Using Situational Crisis Communication Theory Applying SCCT to a crisis situation begins with two steps, the first of which involves the crisis manager examining the situ ation using three fact ors that can affect reputational threats (Coombs, 2007) The three factorsinitial cr isis responsibility, crisis history, and prior reputation/ relationshi p historyare compared along a continuum.
14 In the first step, perceived responsibility for the crisisor locus of controlis determined (Coombs, 2007; Coombs, 1995). In other words, did the organization create the crisis? Prior research indica tes that the more an organization is attributed with crisis responsibility, the lower reputa tional scores are received a nd higher reputational threats exist (Coombs, 2007; Coombs & Holladay, 2002; 2004). Crisis types are used to make this assessment, with the crisis type acting as a frame to indicate how stakeholders will interpret the crisis (Coombs, 2007; Coombs & Holla day, 2002). Coombs and Holladay (2002) synthesized existing crisis typology lite rature into a list of 13 crisis types, attempting to differentiate the assortment of crises an organization may face. The goal of SCCT, then, is to create lists of both cr isis types and response types that may be integrated to become practical for a crisis communicator. By refining existing typologies, variations were identified to expand the definitions of the crisis types an organization may face. In addition to acknowledging a variation between human and technical breakdowns, variations in the definitions of organizational misdeeds were also identified. After the crisis type is identified, adjustments are taken into consideration based on severity of the situation a nd organizational performance hi story. A negative or positive assessment of either of these two factors can change the perceived severity of the crisis situation, signaling a need for different re sponse strategies. As such, the suggested adjustments in typologies give organizations a more precise method of identifying crisis type, and subsequently more appropriate re sponse strategies. A suggested criticism of these typologies, however, is that they are not mutually exclusive or exhaustive. Regardless, moving along the c ontinuum, three crisis cluste rs have been identified to
15 create specific and predicta ble levels of cris is responsibility (C oombs, 2004; 2007; 2007a; Coombs & Holladay, 2002). The victim cluste r results in low attributions of crisis responsibility and low reputat ional threat. The accident al cluster shows minimal attributions of crisis responsibility in whic h the organization is he ld accountable but the event is considered unintentional or uncontrollable. The third cluster is the intentional cluster, in which the organization faces strong attributions of purposeful crisis responsibility. By determining which of the three clusters applies to a crisis situation, an organization can begin to gauge attribution perc eptions of stakeholders and initial threats to reputation. In the case of the 2004-05 NHL lockout, one might argue that the crisis falls into the intentional cluster. The league purposefully locked out its players to achieve a more viable economic model and took it upon itself to cancel the entire 2004-05 season. In the second step, cris is history and prior reputat ion/relationship history are assessed. Also known as consistency, crisis hi story examines whether an organization has experienced a similar crisis in the pa st (Coombs, 2007, Coombs & Holladay, 2002). A history of crisis suggests high consiste ncy, or an ongoing problem, and intensifies reputational threat (Coombs, 2004). Even if an organization is currently a victim, the presence of past crises maintains the heightened threat. Some crisis types can be moved further up the responsibility co ntinuum based on having a history of past crises, but if there is no historyor the history is not publ icly knownthere is little difference in stakeholder perceptions and therefore less th reat to organizationa l reputation (Coombs, 1998; 2004). It has been said that increased crisis histories sh ould result in strategies that result in greater acceptance of responsibility on the part of an organization. However, in
16 the case of the NHL lockout, one may argue th at the league actually worked to deflect blame upon its players association even t hough they entered the lockout with a known crisis history. Rounding out the second step, prior re putation and relationship history are assessed. The two concepts have slightly different effects and are both strongly associated with perceptions of crisis re sponsibility (Coombs & Holladay, 2001). Poor assessment of prior reputation/ relationship history and consistency of crisis history intensif[ies] attributions of crisis responsibility there by indirectly affecting the reputational threat (Coombs, 2007, p. 137). When considered, the two concepts essentially adjust the in itial assessment of the threat (Coombs, 2007, p. 137). Associated with the prior reput ation variable, a halo effect may occur whereby a positive prior reputation may shield or protect th e organization from reputational damage (Coombs & Holladay, 2006). The concept suggest s that reputational damage is deflected and stakeholders may give the organization the benefit of the doubt in the crisis situation. However, identifying a halo effect has been di fficult. The Velcro effect has been easier to identify, whereby negative performance histor y or reputation intensif ies attributions of responsibility (Coombs, 2006). In this cas e, the NHL was facing the 2004-05 lockout with a poor reputation and a prior history of similar crises. There was potential for the Velcro effect to put additional pressure on their reputation. Factoring these variables into their crisis communication appr oach arguably should have af fected their objectives and strategy.
17 Although a variety of additional factors such as legal or financial considerations may affect an organizations crisis response strategy, SCCT acts as a tool to help make decisions. By utilizing the cr isis responsibility continuum organizations can better prepare their crisis communication strategi es. As developed by Coombs and Holladay (2002), the continuum suggests th at the stronger reputational th reat is, or the stronger the personal control is, the more accommodative orga nizations should be with their response strategy. In contrast, the weaker the persona l control is in the situation, the more defensive an organization may be when responding to a crisis. Answering Coombs (2007) call to extend related literature, this case study analyzes the 2004-05 NHL lockout crisis situation, using SCCT as the guiding framework. Although the traditi on of using a case study will be continued, the data collection techniques will be expanded upon to gain a more thorough understanding of the overall situation. A mixed qualitative and quantitative newspaper content analysis will be used to examine how the 2004-05 NHL lockout was presented in the news. Additionally, a fan survey explores the reput ation of and attitudes toward the NHL. One team owner lamented that fans were the forgotten victims (T hompson, 2005) of the lockout when, as the primary revenue source of a league driven by gate revenues, the fans arguably should have been the NHLs primar y concern. This study fills a gap in the literature by looking at the long te rm effects of a crisis situ ation while assessing measures of attitude as they are related to SCCT variables. The research questions outlined earlier provide a baseline assessment of the 200405 NHL lockout as a crisis situation. In a ddition to positioning the 2004-05 NHL lockout
18 as a crisis and investigati ng league communication strategi es, these research questions investigate general fan perceptions of the NHL s reputation. However, this study seeks to extend understanding of SCCT, specifically th e variables that infl uence organizational reputation during a crisis situation and th e resulting attitudes of stakeholder groups. Therefore, the hypotheses outlined earlier ex plore SCCT factors as predictors of long-term effects on stakeholder attitudes following a crisis situation. Although SCCT literature suggests that all cr ises end (Coombs, 2007b), this study suggests that crises continue to affect organizations beyond the crisis situation. It is suggested that this effect can be addressed through meas urement of attitude and reputation. This is particularly important because Coombs (2007b) positions the time after a crisis as a time to measure and evaluate crisis st rategies. It seems reasonable to posit that an assessment of SCCT variables can help an organization gauge its reputation and stakeholder attitudes toward the organization, thus helping develop more effectiv e post-crisis communication strategies based on reputation management objectives. A central suggestion of this study is th at SCCT variables influence the NHLs current reputation, even five years after th e 2004-05 NHL lockout. It is also suggested that fan attitudes toward the NHL are less posit ive than their attitudes toward the sport of hockey. The sport of hockey, potentially identi fied as an overall industry, may arguably be safe from negative reputation effects in spite of crisis because of its nature as a general sport. However, the NHL may be viewed as an organizationan organization in crisis and thus negative attitudes may be held toward the league. Player re sponsibility is also addressed as it may give insight into results related to the NHLs pe rceived responsibility
19 for the crisis. Higher player responsib ility should result in lower NHL/owner responsibility, but may still negatively influence the NHLs reputation. The method section outlines the triangulat ed approach used in this exploratory case study. As will be discussed, a mixed-me thod quantitative and qualitative content analysis and a fan survey was used to inform this study. Categories and coding and measures and design for the two data collection approaches are outlined.
20 Chapter 3 Method Although case studies are sometimes consid ered less rigorous th an other research methods, it has been noted that case studies can be one of the most challenging social science approaches to research (Yin, 2009). Ar guments exist that case studies should be used as an exploratory research method, but it seems hard to disagree with Yins (2009) assertion that every research method can be used for exploratory, descriptive, and explanatory [purposes] (p. 7). The key to succ essful case study research is to approach it with rigor and maintain logical methodological designs. Th is study employs a case study approach in an effort to create a fuller exploratory snapshot of a pa rticular crisis event, the 2004-05 NHL lockout. Yin (2009) operationalizes the case study approach in two parts: (1) A case study is an empirical inquiry that investigates a contemporary phenomenon in depth and within its real-life context, especi ally when the boundaries between phenomenon and context are not clearly evident (p. 18) and (2) The cas e study inquiry copes with the technically distinctive situation in which there will be many more variables of interest than data points, and as one result relies on multiple so urces of evidence, with data needing to
21 converge in a triangulating fashion, and as anot her result benefits from prior development of theoretical propositions to guide data co llection and analysis (p. 18). Boiled down, he suggests that, when deciding on the appropr iate methodology for a study, a case study should answer a how or why question, the re searcher should have little to no control over the events being studied, and the event should be of a contemporaryas opposed to historicalnature. Basing the methodology of this study on th e preceding definition is appropriate for a variety of reasons. The 2004-05 NHL lo ckout can be considered a contemporary event. Per Yins (2009) definition, direct observation of the event was possible and people involved in or affected by the event were directly surveyed. Further, the event cannot be controlledit has already taken place and no variables can be manipulated. Combined, these two criteria point to the not ion that case studies allow researchers to understand a real-life phenomenon in depth (Yin, 2009, p. 18) and within context. Finally, the nature of the rese arch questions addressed in th is study are of the how and why variety. Most importa ntly, and following the opera tionalized definition, these questions were developed based on prior theo retical propositions of Situational Crisis Communication Theory (SCCT). SCCT was us ed to guide the data collection and analysis. A hallmark of good case study research, and the best way to maintain logic and rigor in data collection a nd analysis, is the use of triangulation This research used a variety of data collection techniques in an e ffort to meet high standa rds of logical design. Triangulation is a multiple method approach th at has been advocated by both qualitative
22 (Jankowski & Wester, 1991) and quantitati ve (Wimmer & Dominick, 2003; Yin, 2009) researchers. By definition, tri angulation refers to the use of both qualitative methods and quantitative methods to fully understand the nature of a research problem (Wimmer & Dominick, 2003, p. 48). The use of the two me thods helps creates depth of data, and provides more valid results than the use of a single research method (Jankowski & Wester, 1991). A basic assumption of triangulat ion is that weaknesses in single methods will be compensated by the c ounter-balancing strengths of a nother (cited in Jankowski & Wester, 1991, p. 62). Triangulation is seen as a strength, and a ma ndatory requirement, of case study research (Yin, 2009). In case studies, the most important advantage presented by using multiple sources of evidence is the development of converging lines of inquiry, a process of triangulation and corroboration (Yin, 2009, p. 115-116). For true triangulation, data should be collected from multiple sources in such a way that all results point to and support a single fact or conclusion. Add itionally, the potential problems of construct validity also can be addressed because the mu ltiple sources of evidence essentially provide multiple measure of the same phenomenon (Yin, 2009, p. 116). In an effort to meet the approach ou tlined by Yin (2009) while utilizing SCCT principles, this study employed methodologica l triangulation, approaching a single topic (the 2004-05 NHL lockout) through diverse data gathering methods and analyses. Mixed qualitative and quantitative newspaper analyses and a fan survey were used. Distribution of a communication manager survey and a player survey was unsu ccessful, but is still discussed as a data collection method. Each method will be discussed individually in
23 terms of participants (when applicable), de sign and materials, meas ures, and procedural categories. Content Analysis This study utilized a mixed quantitative and qualitative content analysis, analyzing newspaper content to address research questions about the 2004-05 NHL lockout as a crisis situation. Quantitative content analysis has been defined as a summarizing, quantitative analysis of message s that relies on the scientific method and is not limited to the type of variables th at may be measured or the context in which the messages are created or presented (Neuendorf, 2002, p. 10). The goal of the approach is to use the scientific method to analyze content to the ends of producing objective results (Poindexter & McCombs, 2000) Content analysis is often defined as systematic, objective, and quantitative (Stacks, 2002; Wimmer & Dominick, 2003; Wrench, Thomas-Maddox, Richmond & McCroskey, 2008). By being systematic, all of the collected content can be analyzed uni formly, and was assessed using SPSS and VBPro (Wimmer & Dominick, 2003). Objectivity s uggests that research biases should be avoided, achieved by having additional conten t coders utilize a set of operational definitions and coding criteria, thus giving validity to the content analysis results (Stacks, 2002). Qualitative analysis was simultaneously conducted and qualitative questions were built into the end of the codebook. Hsieh a nd Shannon (2005, p. 1278) define qualitative analysis as a research method for the subjectiv e interpretation of the content of text data
24 through the systematic classification pro cess of coding and identifying themes or patterns. Qualitative analysis served this study by providing examples and support for the quantitative analysis results, all of whic h spoke to the use of SCCT strategies during the lockout. The general flexibility of quali tative analyses has been criticized as a downfall of the method, so Hansens (1998) six-step model was employed. This model was similar to the quantitative approac h, and thus follows accordingly with few adjustments. To study the 2004-05 NHL lockout as a crisis situation, and thus position it within the framework of SCCT strategies, newspape rs were chosen because it has been noted that people become active information seekers in times of crisis, and research suggests that people turn to media for informati on (Wilcox, et. al, 2005). Three date ranges for article collection were chosen based on major turning points during the lockout including the first (September 15-17, 2004) and last (J une 13-15, 2005) days of the lockout, as well as the day that the season was officially canceled (February 16-18, 2005). It has been said that on February 16 fan and media frenzy wa s at its greatest height since September 15, 2004 (Podnieks, 2005, p. 55), mainly because meetings were still being scheduled between the NHL and NHLPA as they con tinued tryingbut faile dto salvage the season. These dates represent points in which the NHL would have been most diligent about maintaining a crisis communication stra tegy. Articles published tw o days after each turning point were analyzed in an effort to collect stories that incl uded news and editorial pieces.
25 Four American and three Canadian daily newspapers were analyzed for content, with the papers chosen based upon circulati on numbers. A benefit of using the most circulated papers for the content analysis is that they boast well-known and respected sports sections that cons istently carried content a bout the 2004-05 NHL lockout. Attempts were made to use reports from ES PN and TSN, the respective sports news leaders in the United States and Canada, but neither organization main tains an archive of articles dating back as far as the lockout. Individuals from both news outlets were reached to attempt gaining access to articles directly through the organizations, but both people pointed to the archiving issue. According to the Audit Bureau of Circul ations, the newspapers with the highest circulation numbers in 2006 included USA Today Wall Street Journal, New York Times Los Angeles Times and the Washington Post (Infoplease.com, 2008). The top five Canadian newspapers, based upon circula tion numbers compiled by the Canadian Newspaper Association (2008) in 2007, included Toronto Star The Globe and Mail Le Journal de Montreal La Presse and The National Post Due to the fact that two of the papers, Le Journal de Montreal and La Presse are French-language publications, content from only three of the five Cana dian newspapers was analyzed. Wall Street Journal was also removed from the sample due to the sma ll number of articles about the lockout. To find articles, the keyword national hockey league (in quotations) and the previously identified date ranges were se arched via Lexis-Nexis Academic. Los Angeles Times articles were pulled from ProQuest, as they were not available on Lexis-Nexis database. The search returned 320 articles. Duplications and articles consider ed newswire, news
26 roundup, notes or notebook, and transaction reports we re removed, reducing the sample size to 282 articles. Multip le units of analysis were us ed in this content analysis. Manifest units of analysis such as word counts and searches we re conducted (Stacks, 2002). A list of keywords were coded and analyz ed. Themes, or latent units of analysis, were also addressed. Articles were coded for positive and negative themes of newspaper coverage, patterns of coverage, a nd types of issues discussed. Intercoder reliability was tested in two ways. First, a guide was developed using wordand phrase-listings coded by five indi viduals. Next, approximately 15 percent ( n = 43) of the articles were c oded twice (see Appendix A). All articles were collected into a single database and assessed using the text analyzer at http://www.online-utility.org/ Sixty keywords that appeared 10 or more times in the articles were coded as positive, negativ e, or neutral by five coders. Neutral words, and words with less than 60 percent (3 out of 5) agreement, were removed from the sample, resulting in a search term list of 47 words, including 18 positive words and 29 negative words. Additionally, phrases were randomly chosen from the articles and were coded to provide a guiding framework for coders. Phrases were coded for negative/positive connotations and favorability toward the NHL, and coders were asked to identify words that gave positive or nega tive meaning to the phrases. Because coders were not guided beyond being asked to identif y words and label the phrases, the items that were agreed upon by the coders 3 or more items were incl uded in the codebook as samples. Of the 22 phrases coders were aske d to review, 19 phrases received agreement among three or more coders on both nega tive/positive connotations and favorability
27 toward the NHL. A total of 32 words were ra ndomly identified by coders as either being positive or negative in connotation, and were included as examples in the codebook. To further ensure the reliability of th e codebook used in this study, 15% of the articles were coded twice. The second inde pendent coder analyzed 43 randomly selected newspaper articles from the original 282 arti cles analyzed by the primary coder. Of the 43 articles, one article was dis carded from the sample due to irrelevance to the crisis situation. Intercoder relia bility was assessed using Holstis formula, as outlined by Stacks (2002). Holstis formula ranged from .714 on five items to 1.00 on three items, as shown in Table 1. These coefficients were consider ed acceptable for further data analysis and PASW Statistics 18 was used to analyze the data after discrepancies in coding were reconciled. Once intercoder reliability was established, the sample of 47 words was used in the VBPro text analysis program developed by Dr. M. Mark Miller of the University of Tennessee. Wild cards were used in the text analysis to improve the programs sentence recognition and word identifica tion. For example, the term cancel* was used to account for variations of the word cancel, includi ng canceled, cancelation and Canadian spelling differences including the use of th e double-L in words such as cancelled.
28 Table 1: Intercoder Reliability Analysis Variable N r Variable n r Story placement 35 0.83 SCCT: Labor history 41 0.98 Story type 36 0.86 Topics: Outside relationship 30 0.71 Lockout presentation 31 0.74 Topi cs: NHLPA relationship 30 0.71 Quoted: NHL Rep. 39 0.93 SCCT: Labor history 41 0.98 Quoted: NHLPA Rep. 42 1.00 SCCT: Fan concern 30 0.71 Quoted: Team/Org. Rep. 39 0.93 SCCT: Future growth 30 0.71 Quoted: Outside Partner Rep. 39 0.93 SCCT: Org. honesty 34 0.81 Quoted: Fan 40 0.95 SCCT : Org. soundness 37 0.88 Quoted: Other 39 0.93 SCCT: Management 32 0.76 Quoted: None 40 0.95 RELN: Fan 37 0.88 No. NHL reps. quoted 39 0.93 RELN: Player 38 0.91 No. NHLPA reps. quoted 38 0.91 RELN: Comm. Manager 42 1.00 Crisis presentation 38 0.91 RELN: Vendor 40 0.95 Topics: Crisis history 37 0.88 RELN: Media partner 37 0.88 Topics: Financial issues 35 0.83 RELN: Sponsor 42 1.00 Topics: Fan-related issues 30 0.71 RELN: Other 41 0.98 Categories and Coding For the codebook, designed to guide quantit ative and qualitative article-by-article coding, analytical categories were devel oped using SCCT as a guiding framework. The categories were developed from SCCT variable s. A previous qualitati ve content analysis uncovered themes that were expanded upon wi th additional qualitative and quantitative analysis (Formentin, 2009). Items were deve loped to code for how the lockout was portrayed by writers in an effort to position the lockout as a crisis Additionally, articles were coded to identify how the NHL as an organization was represented and how the
29 NHLs reputation was presented and discusse d. Reputation measures were modified from items developed by Coombs and Holladay ( 2002) and Sung and Yang (2008), as will be discussed in more detail in the survey resear ch section. Finally, arti cles were also coded for portrayals of crisis type and crisis history. Results re flect an analysis of the crisis situation based on each individual turning point, as well as three turning points combined. Survey Research To inform understanding of the content an alysis and test SCCT principles, survey research was conducted. Defined as a me thod of gathering relatively in-depth information about respondent attitudes and beliefs (Stacks, 2002, p. 175), surveys are measurement instruments aimed at gathering information about how a public perceives an issue or event or person (p. 175). Surveys are controlled, and can boast higher levels of measurable reliability and validity if cons tructed correctly. Specifi cally, an Internetbased survey was administered using the online survey tool, www.surveymonkey.com. Materials A survey was distributed to the under-st udied stakeholder group of NHL hockey fans. Although the NHL claims it has done prev ious research on fan perceptions at the time of the lockout, the information has been kept proprietary to date (Batchelor & Formentin, 2008). Users who frequent onlin e message boards are typically more interested in specific subject matter, thus onlin e distribution was used in an effort to gain a stronger response to the call for help with this study (Ridings & Gefen, 2004). Limitations to the generalizability of results are inherent due to th e online distribution of
30 the survey and the use of a convenience samp le. A variety of succe ssful and unsuccessful attempts were made to distribute the survey. A 52-question instrument (see Appendix B) was distributed to fans via the Webbased survey tool SurveyMonkey. Thirty NHL team communication managers were contacted for permission to post the survey s on individual team message boards. Even after a follow-up request, only one communicat ion manager suggested that posting the survey on the team message was allowed. Ho wever, this suggestion was made under the implication that the boards are controlled by the NHL. Of the managers reached for this study, 12 declined the request to distribute th e survey to their fans and 18 communication managers never responded to requests for assistance. Alternate methods of survey distribution were considered and it was determined that six sites may be appropriate for this study: NHL message boards ( www.nhl.com ), The NHL Arena fan message boards ( www.thenhlarena.com ), the Hockeys Future fan message boards ( www.hockeysfuture.com ), the National Hockey League Fan Association (NHLFA) (www.nhlfa.com ), the ESPN message boards (boards.espn.go.com), and TSN message boards. TSN is self-professed as C anadas sports leader (TSN.ca, 2010), and would have provided access to Canadian fans Unfortunately, it does not have a message board and was excluded from the conveni ence sample. Representatives from the NHL, The NHL Arena, Hockeys Future, and NHLFA were all contacted to request permission to share the survey link on th eir sites. Hockeys Future representatives declined the request citing that they dont permit [their] sites to be used for surveys. The NHLFA contact never responded to an initial request, and no additiona l attempts were made at
31 contacting the representative due to past e xperiences contacting the individual. An NHL representative was contacted, but no respons e was received. NHL Arena representatives agreed to grant permission to post the survey link on the main NHL discussion board. Additionally, the survey link was posted on ESPNs general message board. After posting the link on The NHL Arena boards, a fan expresse d interest in the survey and distributed the link in his hockey blog while sending the message to other people in his blogging network. In all, the surveys remained open for two full weeks, from Feb. 10 Feb. 24, 2010, generating 140 responses. Reminders requesting participation were made randomly during the two-week period, effectively bumpi ng the threads in which the survey link appeared and maintaining a ppropriate online dialogue. By the second week, reminders were not generating a significant number of responses and the survey was closed. Measures and Design The measures for this survey were deve loped from previous SCCT research and concepts that emerged from the newspaper content analysis. Sections of the survey represented SCCT propositions for reputati on management during crises. Specifically, items were created to measure six variables: current reputation, prio r reputation, attitude toward the NHL, attitude toward hockey, cr isis responsibility and crisis history. One goal of the survey was to measure the NHLs current reputation among fans, five years after the crisis, using SCCT vari ables. By measuring the leagues current reputation, attempts are made to determine if the leagues reputation has improved in the years following the 2004-05 NHL lockout crisis.
32 No known reputation measurements exist from the time of the lockout, and the survey instrument cannot precisely determine the leagues reputation at that time. Even so, to measure reputation, two sets of measures were developed. One set measures current reputation and the second set measures reputatio n five years ago (the time of the lockout). To assess current reputation, par ticipants were aske d to rate the statement Overall, my impression of the NHL is using a seven-point scale ranging from 1 ( very unfavorable) to 7 ( very favorable ). This one-item, global evaluation is identical to one used in previous reputation management studies by Coombs and Holladay (2006). The measure is intended to provide a general idea of how participants viewed organizational reputation, (Coombs & Holladay, 2006, p. 129). The instrument also included reputation measures developed from two previous studies. Sung and Yang (2008) defined reput ation as public pe rceptions of the organization shared by its multiple constituents over time (p. 363). The definition, based on their interpretation of multiple definitions of organizational reputation, suggests that outside stakeholder perceptions about cu rrent and future performance defines organizational reputation. For their own research, Sung and Yang (2008) modified reputation quotient measures develope d by Fombrun and Gardberg in 2000. The Reputation Quotient (RQ) has been developed to measure a companys reputation by examining how a representative group of stakeholders perceives companies on six dimensions of reputation (Fombrun & Gardbe rg, 2000). Used in nationwide surveys, RQ ranks companies by their reputations based on consumer feedback. Six dimensions of reputation have emerged including corporat e appeal, products and services, financial
33 performance, vision and leader ship, workplace environment, a nd social responsibility. In a study about university image, Sung and Yang (2008) modified only some of the RQ measures, thus arriving at five items that yielded a Cronbachs alpha of .88. Those items were adapted for this study. As an aside, in their 2000 article W hos Tops in Corporate Reputation? Fombrun and Gardberg refer to a forthcoming article when discussing RQ. This article is cited as appearing in the Journal of Brand Management ; however, no such article exists. Research into the matter revealed that the journal was launche d in 2000, and only two issues from that year are available via the University of South Fl orida Library and the publishers (Palgrave MacMillan) official si te. As such, although it would have been preferable to refer to the or iginal RQ questions developed by Fombrun and Gardberg, this survey will utilize modified versions of Sung and Yangs (2008) measures. Coombs and Holladays (2002) modified scale for measuring organizational reputation cites that trustworthiness is cons idered central to reput ation and is a common factor used in the commercial reputation measures (p. 175). The commercial reputation measures they refer to include Fombruns RQ. The Coombs and Holladay (2002) scale, adapted from their own 10-item Organizationa l Reputation Scale, used five items that produced Cronbachs alpha of .80 or higher in two prior studies. These items were modified for this study. Items such as The NHL looks like a sports league with strong prospects for future growth were used as vi sion and leadership measures, and The NHL is socially responsible addr essed social responsibility. A ll items were measured using
34 Likert-type scales ranging from 1 ( strongly disagree ) to 7 ( strongly agree ). Both Coombs and Holladays (2002) and Sung and Ya ngs (2008) items were used. To measure prior reputation, current reputa tion measures were altered to reflect a comparison to five years ago when the lo ckout happened. Respondents were instructed to recall their attitudes about th e NHL prior to the lockout to the best of their ability, generating self-reported comparisons. These it ems allowed for analysis of one of the three factors accounted for in SCCT: pr ior reputation/relationship history. Worth mentioning is that Sung and Yangs (2008) scale does not have an immediately obvious item for Corporate/ Emotional Appeal beyond the item The organization considers fans a top priority . Fombrun and Gardberg (2000) identify Corporate Appeal as How much the compa ny is liked, admired, and respected (p. 13), which can be addressed via atti tude measures. Attitudes were measured using a semantic differential scale in which respondents were aske d to rate their feelings on a 7-point scale comparing good to bad, positive to n egative, and favorable to unfavorable. Respondents were asked to rate their attit udes twice; once for th e NHL and once for the sport of hockey. A variety of newspaper artic les from the time of the lockout addressed the notion that fans differentiate between the two entities, which is a revealing feature of this study. To measure crisis responsibility, Coombs and Holladay (2002) utilized Griffin, Babin, and Dardens (1992) th ree-item scale for Blame (p. 175), yielding a Cronbachs alpha of .91. Other studies utilizing this sc ale have yielded Cronbachs alphas ranging from .80 to .86 (Coombs and Holladay, 2002). Items adapted for the present study
35 included, The blame for the lockout lies with the NHL, while items such as The blame for the lockout lies with the players, not the NHL, were added to reflect the nature of the specific crisis event. All items were measur ed using Likert-type s cales ranging from 1 ( strongly disagree) to 7 ( strongly agree). Additional crisis responsibility items were created for this study, intended to examine this particular crisis event in depth. Items intende d to determine responsibility for the lockout included I believe the NHL could have done more to prevent the lockout, which may provide some insight into other survey responses. Greed emerged as a theme of disapproval toward NHL players in the content analysis, so the item I believe the lockout was a result of greedy players was included in the instrument. When it comes to crisis situations, apologies are tr icky. Legal implications may make an apology an unreasonable response during a crisis, but if an apology is warrantedor believed to be warrantedperceptions of a cr isis situation can change. As such, an item such as I believe the NHL should have apologized for the lockout may ela borate other survey results. The SCCT variable for history was m easured using three items developed specifically for this study: I am aware that there have been two lockouts in NHL history; I am aware of the player strike that occurred in the early 90s; and I remember the 2004-05 lockout. The best way to address peoples knowledge of prior history of crisis within the organization was to simply ask about previous events that actually occurred. The use of a contemporary event for a case study allows for questions about actual crisis history.
36 Questions aimed at adding depth to the st udys findings were also included in the survey. Asking respondents to address I am proud to be an NHL fan directly addresses the inherent nature of being a fan of anythingif someone is not proud to be a fan of the sport, league, or a team they support, then this could arguably be a measure of reputation. The statement The NHL and the NHLPA are th e same organization was also included in the fan survey to identify if fans diffe rentiate between the league and the players association, the two groups most deeply entr enched in the contract bargaining agreement dispute. If fans think of the NHL in terms of the players and athletes that represent the league, then the NHL may inadvertently make itself look bad if it makes the players look bad during this type of crisis. Ultimately, placing blame on players may affect the leagues reputation, thus it is important to know if fans differentiate between the league and the players association. Additional Surveys Two additional groups were approached to participate in surveys developed for this case study: NHL communication managers and NHL players. A quick glance at club staffs shows that most communication de partments have between three to five communication managers. The highest ranki ng communication executive was identified for each of the 30 teams and contacted via email to request participation in the study. The goal was to obtain generalizable results with an emphasis plac ed on understanding the leagues reputation amongst organization-level communication managers.
37 Thirty NHL team communication managers were contacted for permission to distribute a 47-question survey instrument to team comm unication employees. A first wave of requests yielded one agreement to participate in the study and four rejections. Of the remaining 25 managers, no responses were received. A second wave of participation requests were sent out a week later, yielding four poten tial participants and four rejections. The remaining team communica tion managers never responded to any requests for participation. Of the potential participants identifie d following the second request, only one communication manager eventu ally completed the survey. For the team that agreed to distribute the survey to th eir communication managers, five individuals were contacted and three participated. The original requests were sent prior to the 2010 Winter Olympic break, posing some logistical issues with data collecti on. Because of the Olympic break and pending playoffs, the communication manager survey was removed from this study as it was deemed more appropriate to leave the manage rs alone so as not to risk losing their participation in future studies. Considering time constraints and a low response rate, a different study may be better suited for this survey and communication managers may be easier to reach at a different time of the year. A final group approached for participation in this study was that of the players. Athletes are arguably the selli ng point of sports leagues and the intention was to survey the players about their perceptions of the league and fans. Ini tial research into the lockout shows that player perceptions of the results of the crisis situation are sometimes hard to identify (Batchelor & Formentin, 2008). Arguably, this may have to do with the lack of
38 anonymity afforded by the public interviews the players are ofte n asked to conduct. Although attempts were made to reach out to this affected stakeholder group, it was not possible to secure access to the players. An NHLPA representative was contacted to help identify a decision-maker who could assist with survey dist ribution. The representative sugg ested that this access needed to be approved by the NHL and offered to help find the appropriate person while suggesting to be blunt, a survey of the pl ayers would have many challenges and would be very difficult to achieve [sic]. Alt hough additional contact was made with this person, no additional information was offered in regard to distributing the survey. A final attempt was made to distribute the player survey via an e-mail to the NHLs Vice President of Public Relations and Player Development. No response was received. Due to the timing of the survey distributionduring the Winter Olympic break, and toward the end of the seasona decision was made to remove the player survey from this study. The survey itself was shorter than the othe r surveys, featuring 25 questions including demographics, and was designed to gain some insight into player perceptions of the league. There was little intention to make the survey generalizable. The following section presents results of the content analysis and survey.
39 Chapter 4 Results As previously discussed, the following research questions and hypotheses are addressed in this study. The research ques tions and hypotheses are separated per the research method that informs them: Research Questions Content Analysis RQ1: Was the 2004-05 NHL lockout presented by major newspapers as an organizational crisis? RQ2: How was the 2004-05 NHL lockout presented by major newspapers in terms of SCCT variables? RQ3: What strategic communication strate gies did the NHL use during the 2004-05 lockout, as presented by major newspapers? Survey RQ1: Do fans have a positive per ception of the NHLs reputation? RQ2: Has the NHLs reputation improved in the five years since the lockout?
40 Hypotheses H1: NHL crisis responsibility, crisis hist ory, and prior reputation influence the NHLs current reputation. H2: Hockey fan perception of reputat ional threat variables (crisis responsibility, crisis history, and prior reputation) related to the 2004-05 NHL lockout are related to fa n attitude toward the NHL. H3: Hockey fan perception of reputat ional threat variables (crisis responsibility, crisis history, and prior reputation) related to the 2004-05 NHL lockout are not related to fan attitude toward hockey. H4: Among hockey fans, perception of high player responsibility for the 200405 NHL lockout is inversely rela ted to perception of NHL/owner responsibility for the 2004-05 NHL lockout. H5: Among hockey fans, perception of high player responsibility for the 200405 NHL lockout negatively influenc es perception of NHL reputation. H6: Hockey fan attitude toward the NHL is less positive than fan attitude toward hockey. Content Analysis For this study, 282 newspapers articles were coded from seven North American newspapers. Coded articles came from th ree turning points during the lockout. The majority of articles (47.2%, n = 133) were from the day the lockout was canceled, as shown in Table 2.
41 Of the coded articles, 78.8% of the arti cles came from Canadian newspapers ( n = 222). Papers from the United States acc ounted for 20.6% of the coded articles ( n = 58). Table 3 provides a breakdown of the number of articles coded from each paper. Table 2: Newspaper Frequencies, by Date Date n % Sept. 15-17 81 28.7 Feb. 16-18 133 47.2 June 13-15 66 23.6 Table 3: Newspaper Frequencies, by Source Source n % Los Angeles Times 16 5.7 New York Times 17 6.0 USA Today 14 5.0 Washington Post 11 3.9 Globe and Mail 97 34.4 National Post 76 27.0 Toronto Star 49 17.4 The majority of articles (46.8%) were 500-1000 words in length ( n = 132), as shown in Table 4, and appeared inside the s ports section of the respective newspapers (39.7%, n = 112), as shown in Table 5. News and feature pieces ( n = 152) made up 53.9% of the coded articles, as shown in Table 6.
42 Table 4: Newspaper Frequencies, by Word Length Article Length n % Less than 500 words 78 27.7 500-1000 words 132 46.8 More than 1000 50 17.7 Table 5: Newspaper Frequencies, by Story Placement Placement n % Front Page, paper 21 7.4 Front Page, sports section 51 18.1 Inside Sports Section 112 39.7 Other/Undetermined 76 27 Table 6: Newspaper Frequencies, by Article Type Type of Article n % News/Feature 152 53.9 Editorial/Opinion 53 18.8 Letter to the Editor 17 6.0 Other 38 13.5 Research Questions Research Question 1 asked if the 200405 NHL lockout was presented by major newspapers as an organizational cr isis. Although the word crisis ( n = 11) rarely appeared in the coded articlessometimes appearing because of the name of a cited crisis management consulting agencycoders indi cated that newspapers cast the 2004-05 NHL lockout in a negative light in 58.2% of the articles ( n = 164). Of the remaining articles, 30.1% were considered balanced pieces ( n = 85) and only 3.2% ( n = 9) positioned the lockout in a positive light, as shown in Ta ble 7. Additionally, words coded as positive ( n
43 = 17) and words coded as negative ( n = 29) were found 2,290 times and 2,162 times, respectively. Fans (n = 712), considered a positive word, accounted for 16% of the overall sample of words and 31.1% of the positively coded words. Lockout ( n = 478) was the most frequently appearing negativ e word, accounting for 10.74% of the of the overall sample and 22.12% of the negatively coded words. Terms such as cancel* ( n = 310) appeared more frequently during tim e frames that would encourage a logical increase. For example, 84.52% of occurrences the term cancel* ( n = 262) occurred at the time the season was canceled. Overall, the c ontent analysis seems to indicate that the 2004-05 NHL lockout was presented by major ne wspapers as an organizational crisis. Table 7: Overall portrayal of the NHL lockout NHL portrayal n % Negative 164 58.2 Positive 9 3.2 Balanced 85 30.1 Research Question 2 asked how th e 2004-05 NHL lockout was presented by major newspapers in terms of the SCCT variab les for reputation, crisis history, and crisis responsibility. Measures designed to mimic th e reputation measures found in the surveys (to be discussed) addressed topics including NHL concern for fans, NHL prospects for future growth, NHL honesty, financial s oundness of the NHL, and the NHL as wellmanaged. Due to lack of subject matter addr essing the topic, an item assessing the NHL as socially responsible was dropped from the an alysis. Coders were asked to assess each reputation measure in two parts: 1) Identify if the topic come up in the article, and 2) If the topic was addressed, identify if it was a ddressed in a positive or negative manner.
44 Frequency of reputation-based topics bei ng addressed was low and discussions were generally negative, as shown in Table 8. Table 8: Newspaper Analys is, Reputation Variables Variable Yes % No % Missing % Concern for Fans 13 4.6 34 12.1 215 76.2 Prospects for Future Growth 42 14.9 54 19.1 166 58.9 Honest 2 .7 41 14.5 219 77.7 Financially Sound 2 .7 123 43.6 137 48.6 Well-Managed 3 1.1 53 18.8 206 73.0 Crisis history was addressed by assessing history in two ways: 1) coders were asked to identify if a history of crisis was addressed, and 2) coders were asked to identify if the NHLs labor history was discussed. Hist ory of crisis could i nvolve any discussion of previous lockouts, strikes, or negati ve incidents related to the NHL. Negative incidents, even if unrelated to the 200405 NHL lockout, could potentially affect the NHLs reputation. As such, even negative incidents unrelated to the 2004-05 NHL lockout were considered in the content an alysis. For example, qualitative analysis revealed that at the time of the lockout the ongoing and highly pub licized court-battle between players Todd Bertuzzi and Steve M oore was taking place. Bertuzzi and other members of the Vancouver Canucks orga nization were being sued by Moore for involvement in an on-ice incident that ended Moores career. A lthough unrelated to the lockout, this qualifies as a ne gative incident. History of crisis came up in 10.3% ( n = 29) of the articles. Labor history was a point of discussion in 8.9% ( n = 25) of the articles. History of crisis was discussed most frequently ( n = 17) during the first turning point of
45 the lockout (Sept. 15-17, 2004). As shown in Table 9, the NHLs labor history was discussed most frequently ( n = 25) when the lockout began. A chi-square was conducted to assess whether discussion of labor histor y changed significantly based on the date of publication, yielding significant results: 2(6, N = 280) = 26.413, p < .000. Discussion of labor history was not signifi cant across specific newspapers. Table 9: Discussion of labor history Date of Publication Yes No Total Sept. 15-17 16 64 81 Feb. 16-18 5 116 133 June 13-15 4 53 66 Totals 25 233 2 Finally, an assessment of crisis responsibility is best informed by a qualitative analysis of the coded newspaper articles. Most reaction in the newspaper articles presented a view that the lockout wa s caused by both the NHL and NHLPA, although some writers or quoted individuals specified an organization that they felt was to blame for the 2004-05 NHL lockout. For exampl e, Rick Salutin (2004) of the Globe and Mail wrote Count me with the 21 per cent of Canadians who blame owners, not the 52 per cent who say its the players fault. He discussed fairness between rich guys and described Commissioner Gary Bettman as the owners mouthpiece. Writing for the National Post Jack Todd (2004) wrote that Bettmans strategy has been to deflect the blame by demonizing someone -in Bettmans case, the NHL Players Association. On the other hand, a National Post (2005) editorial argued not to blame the owners, saying instead that the NHLPA overplayed its hand and they cannot
46 reasonably expect the league to survive under the current economic system. Fans such as Kristen Mazza were quoted saying, I mostly blame the players. I don't have much sympathy for someone who makes more in one year than I'll make in 30 years (cited in Elliott, 2005). A subsequent Globe and Mail (2004) editorial suggested that both owners and players share blame for this mess. Fans can be forgiven for wishing a pox on both their greedy houses. Kings fan Dan Daniel gave up his season tickets citing, Im embarrassed to be a hockey fan at this point. Its the first major professional sports league in America to cancel a season because of greed (cited in Bolch, 2005). However, an employee laid off by the NHLs Dallas Stars was quoted as saying, I cant really be bitter at the Stars; they have to look at the economic situa tion. I loved my job and I would have loved to keep it. Its tough to blame either side (NHL or NHLPA), but somebody let it get to this point (cit ed in Zwolinski, 2004). High-profile people with a connection to the NHL also chose between sides. Disgraced former NHLPA director Alan Eagleson blamed both sides (Woods, 2004b), while legend Phil Esposito was quoted as saying he was simply angry and upset at the lack of compromise and So sick and tired of them calling each other names (cited in Woods, 2005). However, another legend, Guy La Fleur, sounded off against the players saying they lacked respect for the game, had taken advantage of the economic system for years, and that they are unrealistic in th eir quest to avoid a salary cap (Stubbs, 2005). Overall, the content analysis seems to suggest that the NHL is presented by newspaper writers as having a poor reputation. Newspaper writers did not cover the
47 NHLs history of crisis with great frequency, and were more likely to discuss previous labor disputes at the time the lockout was announced. Finally, newspaper writers presented responsibility for the lockout in two ways: 1) Players were blamed for the lockout, or 2) Both players and owners were blamed equally for the lockout as a cohesive pair representing the NHL as an individual whole. Finally, RQ3 asked what strategic communi cation strategies the NHL used during the 2004-05 NHL lockout. Both the announcement of the lockout and the cancellation of the season were formally announced during press conferences featuring Bettman. On each date, the NHL made its announcement before the NHLPA and did so from New York City near the leagues offices. At the beginning of the lockout, Bettman set a tone by frequently using cost certainty and linka ge as buzzwords to describe the needs of the NHL during the lockout. Cost certai nty was called a euphemism for a cap on salaries (Cole, 2004) and linkage was a name applied to the con cept of revenue-sharing. Additionally, Bettman positioned cost certainty as a way to achieve a better economic system and lower prices for fans (Duhatschek, 2004). He claimed that the future of the game was at stake and the owners were unite d in their stance. Bettman also began the lockout by pointing blame toward the NHLPA, suggesting they should have see[n] this bleak day approaching but didnt lift a finge r to prevent its arriva l. In the background of Bettmans press conferences, the NHL prep ared a Web site devoted to explaining its side of the dispute, a move described as a fierce public relations campaign designed to sway fan support (Woods, 2004).
48 When the season was formally canceled, the NHL distributed a full copy of a memo sent to NHLPA executive director Bob Goodenow outlining the leagues final offer. The letter suggested This offer is not an invitation to begin negotiations its too late for that, and was described as a fou r-paragraph missive with a sharp take-it-orleave-it tone (Wharnsby, 2005). Bettman opened the press conference announcing the cancellation by saying it was a S ad, regrettable day. He then went on to describe the NHLPA as an unwilling partner while suggesting everyone owes our fans an apology. When the lockout was lifted in mid-J une 2005, the NHL and NHLPA sent out a joint statement saying, "The National Hockey League and the National Hockey League Players Association have reached an agreement in principle on the terms of a new Collective Bargaining Agreement. It is antici pated that the ratificat ion process will be completed next week, at which time the parties will be prepared to discuss the details of the Agreement and plans for next season (cited in Wherry, 2005). Overall, the content analysis seems to indicate that the NHL used defensive communication strategies. Although it apologi zed to fans, its primary communication efforts were defensive. The NHL either blamed the players for forcing them into the lockout, or the NHL used its current economic stability to reason that the lockout was a necessity for the economic health of the league.
49 Fan Survey Respondents in this study included 140 hocke y fans contacted via the Internet. Of the participants, 14.3% were female ( n = 20) and 58.6% were male ( n = 82). Thirty-eight participants did not indicate gender. The age of respondents ranged from 17 to 57 years old ( M = 31.87, SD = 10.572). The majority of responde nts were between the age of 20 and 29 ( n = 40), accounting for 28.6% of participants as shown in Table 10. In addition, 67.9% of respondents identified themselves as Caucasian (n = 95) and 2.1% were Asian/Pacific Islander ( n = 3). No other ethnic choices were selected, with 30% ( n = 42) of respondents leaving the question bla nk. The majority of respondents (31.4%) identified themselves as having some college ( n = 44) education, while 24.3% are college graduates ( n = 34). Respondents were asked to report where they lived via an open-ended question and came from 22 states and four Canadian provinces. Of the states and provinces represented, eight of the stat es and one of the provinces do not have a NHL franchise. Specifically, 9.29% of respondents ( n = 13) were Canadian, a nd 60.71% lived in the U.S. ( n = 85). Of the respondents who lived in the U.S., the majority ( n = 15) were from Pennsylvania, accounting for 10.71% of overall respondents and 17.65% of U.S. respondents. Other states with 10 or mo re respondents incl uded Michigan (n = 11) and New York ( n = 11). This demonstrates a genera lly diverse participant sample.
50 Table 10: Respondent Demographics, Age Age Range N % 19 or younger 6 4.2 20-29 40 28.6 30-39 24 17.2 40-49 14 9.9 50 or older 9 6.3 Missing 47 33.6 Finally, respondents indicated that they ha ve been hockey fans from 3 to 53 years ( M = 23.12, SD = 11.909), with the majority of participants (26.43%) identifying themselves as hockey fans for 10 to 19 years ( n = 37), meaning most were fans prior to the lockout. Respondents identifying themse lves as fans for less than 10 years ( n = 8) made up 5.6% of participants. Of the participants, 32.9 ( n = 46) left the question blank. Respondents generally considered them selves to be diehard fans ( n = 105, M = 6.59, SD = .98) and strongly agreed that they remembered the 2004-05 NHL lockout (n = 109, M = 6.52, SD = .97). They were also aware that there have been two NHL lockouts (n = 105, M = 5.59, SD = 1.79), and were aware of the player strike that occu rred in the early 90s ( n = 108, M = 5.63, SD = 1.73). Research Questions Research Question 1 asked if NHL fans have a positive perception of the NHLs reputation. The one-item, gl obal evaluation used by Coombs and Holladay (2006) in previous reputation research revealed th at respondents had a generally favorable impression of the NHL ( n = 75, M = 5.63, SD = 1.32). In addition, the nine additional items used to measure current reputation in this study were collapsed into a composite
51 measure (a = .874), which yielded an average mean score of 4.28. Overall, survey results indicate that fans have a somewhat favor able perception of the NHLs reputation. Research Question 2 asked if the NHLs re putation has improved in the five years since the lockout. Paired t-tests were conducted on reputa tion measures addressing the leagues current reputation and its reputati on five years ago. Nine items previously developed by Coombs and Holladay (2006) and Sung and Yang (2008) were modified into two sets of items to assess current and past reputation perceptions. The matched items were paired and differences in item means were assessed. Of the nine pairs, six were statistica lly significant. Ther e was a significant difference between perceptions about the NHLs concern for the well-being of fans five years ago ( M = 4.06, SD = 1.46) and the NHLs current concern for fans (M = 4.44, SD = 1.43), t (107) = 2.341, p = .021. When measuring the NHL as a league with prospects for future growth, there was a significant diffe rence between perceptions five years ago ( M = 4.42, SD = 1.75) and current perceptions ( M = 4.90, SD = 1.40), t (106) = 2.487, p = .014. Five years ago the NHL was consider ed less socially responsible ( M = 4.22, SD = 1.52) than they are now (M = 4.58, SD = 1.47), t (107) = 2.324, p = .022. When measuring honesty, the item The NHL is basically dishonest was reversecoded, yielding a significant difference between perceptions of NHL honesty five years ago ( M = 3.83, SD = 1.538) and now ( M = 4.53, SD = 1.43), t (107) = 4.116, p = .000. Another item reverse-coded was The NHL is NOT concerned with the well-being of its fans, ( M = 4.80, SD = 1.44), and when paired with the item, Five years ago, the NHL was concerned with the well-being of its fans ( M = 4.16, SD = 1.429) a significant
52 difference was observed, t (104) = 4.267, p = .000. Finally, a significant difference was identified between respondent perceptions about the NHLs financial soundness five years ago ( M = 3.07, SD = 1.58) and current financial soundness ( M = 3.79, SD = 1.325), t (106) = 4.055, p = .000. Items that were not sign ificant addressed the leagues consideration of fans, how likely fans are to believe what the NHL says, and if fans believe the NHL as a well-managed organi zation. This indicate s that no significant changes exist for these reputat ion items from the time of the crisis five years ago. Overall, survey results indicate that th e NHLs reputation among fans is somewhat improved compared to its reputation five years ago. Tests of Hypotheses For purposes of conciseness, the remaining analysis in this st udy will refer to the prior reputation/relationship history variable simply as prior reputatio n. Prior to hypothesis testing, the multi-item measures used to assess crisis responsibility, crisis history, prior reputation, curre nt reputation, attitude toward NHL, and attitude toward hockey were assessed for intern al reliability. The alpha coe fficient and number of items in each composite measure are shown in Table 11. The alpha score for all composite measures indicated strong in ternal consistency, so data analysis proceeded with hypothesis testing. Although alpha values between .80 and 1.00 indicate high reliability (Berman, 2002), a lower limit of .70 is still considered a useful measure of constructs (Broom and Dozier, 1990; Stacks, 2002). The Cronbachs al phas for the single-item constructs ranged
53 from .79 to .96, with four items having coeffici ents of .87 or greater. This suggests that the items measuring SCCT variables de monstrate high internal reliability. Table 11: Reliability Analysis for Variable of interest Variable a n Crisis responsibility .81 4 Crisis history .79 3 Prior reputation .87 9 Current reputation .87 9 Attitude toward NHL .90 3 Attitude toward hockey .96 3 Hypothesis 1 posited that NHL crisis re sponsibility, crisis history, and prior reputation influence the NHLs current reputation. To te st this hypothesis, linear regression analysis was conducted in which the measures of crisis re sponsibility, crisis history, and prior reputation we re the predictor variables a nd current reputation was the criterion variable. The results of linear re gression analysis revealed that 14% of the unique variance in current reputation was due to the measures of crisis responsibility, crisis history, and prior reputation, R =.38, R2=.144, F (3,95)=5.176, p=.002. However, only crisis responsibility a nd prior reputation were unique predictors of current reputation, as shown in Table 12. Overall, th ese results provide pa rtial support for the propositions of SCCT, suggesti ng that perceptions of cris is responsibility and prior reputation positively affect perceptions of current reputation.
54 Table 12: Regression Model for Fan Perception of SCCT Reputational Threat Variables Predicting Attitude Toward NHL Variable B SE B Sig. Crisis responsibility -. 194 .092 -.215 .039 Crisis history -.030 .081 -.037 .715 Prior reputation -.254 .092 .270 .007 Hypothesis 2 posited that hoc key fan perception of reput ational threat variables (crisis responsibility, crisis history, and prior reputation) related to the 2004-05 NHL lockout influences fan attitude toward the NHL To test this hypothesi s, linear regression analysis was conducted in which the measures of crisis respon sibility, crisis history, and prior reputation were the pr edictor variables and attitude toward the NHL was the criterion variable. The results of the anal ysis were not significant, suggesting that, overall, these results do not provi de support for the notion that threat variables are related to attitude toward the NHL as an organization. Hypothesis 3 posited that hoc key fan perception of reput ational threat variables (crisis responsibility, crisis history, and prior reputation) related to the 2004-05 NHL lockout do not influence fan attitude toward hockey. Linear regressi on analysis was used to test this hypothesis. Results i ndicated, no significant overall effect, R =.29, R2=.087, F (3, 83)=2.529 p=.063; however, crisis history was a unique predictor of attitude toward hockey, =-.080, t (82)=2.198, p=.031. These results suggest th at only the crisis history threat variable is related to attitude toward the sport of hockey.
55 Hypothesis 4 posited that among hockey fans, perceptions of high player responsibility for the 2004-05 NHL lockout ar e inversely related to perceptions of NHL/owner responsibility for the 2004-05 NHL lockout. A scale analysis of all responsibility measuresinclu ding those placing blame on th e players and those placing blame on the NHLwas not, as exp ected, internally reliable ( a = .219). As such, items were separated into groups depending on wh ether blame was being placed on players or the league. Two modified measures used in previous studies included The blame for the lockout lies with the NHL ( M = 4.58, SD = 1.60) and The blame for the lockout lies with the players, not the NHL (M = 3.85, SD = 1.39). A third item, The NHL is not to blame for the lockout ( M = 3.02, SD = 1.35), was also reverse-coded ( M = 4.98, SD = 1.35) to be included in both the NHL-responsib ility and player-responsibility variables. All the responsibility items we re collapsed with responsibil ity measures developed for this study, creating single variables for NHL Responsibility and Player Responsibility. Items measuring blame toward the NHL showed high internal reliability ( a = .868). Items included the reverse coded The NHL is not to blame for the lockout, I believe the NHL could have done more to prevent the lockout, I believe the NHL should have apologized for the lockout, The blame for the lockout lies with the NHL, I believe the owners were at fault for the NHL lockout, and I believe the NHL should have taken responsibility for the lockout. The scale M = 29.77 and SD = 6.70 over six items shows that respondents somewhat agree that the NHL ( M = 4.96, SD = 1.12) is to blame for the lockout. A single NHL Responsib ility variable was cr eated for additional statistical analysis.
56 For a single player blame variable, the item The NHL is not to blame for the lockout was included in an effort to see how its inclusion in both scales would affect results. Internal reliability increases ( a = .837) without this item, but it will be retained for comparative purposes. Items include, I beli eve the lockout was a result of greedy players, I believe the players should have taken responsib ility for the lockout, The blame for the lockout lies with the players, not the NHL, and The NHL is not to blame for the lockout. The scale M = 15.06 and SD = 4.47 over four items shows that respondents somewhat disagree that the players ( M = 3.77, SD = 1.12) are to blame for the lockout. Thus, a single Player Respons ibility variable was created for additional statistical analysis. A paired t -test of the collapsed NHL Responsibility and Player Responsibility variables showed a significant difference of perceived responsibility between the two groups. Respondents were more likely to blame the NHL ( M = 4.96, SD = 1.12) than the players ( M = 3.77, SD = 1.12) for the 2004-05 lockout, t (104) = 6.039, p = .000. Overall, survey results seem to indicate that fan pe rceptions of high player responsibility are inversely related to perceptions of NHL/owner responsibility. Hypothesis 5 posited that hocke y fan perception of high pl ayer responsibility for the 2004-05 NHL lockout negatively influences perceptions of NHL reputation. To test this hypothesis, linear regressi on analysis was conducted in wh ich the measures of player responsibility, crisis history, a nd prior reputation were the pred ictor variables and attitude toward the NHL was the criterion variable The results of the analysis were not
57 significant and, overall, suggests that fan perceptions of high player responsibility do not negatively influence perceptions of the NHLs reputation. Hypothesis 6 posited that fan attitudes towa rd the NHL are less positive than their attitudes toward the sport of hockey. As previously discusse d, items measuring attitudes toward both the NHL ( a = .901) and hockey ( a = .957) had high internal consistency and were collapsed into unique measures. Thes e measures were analyzed using a paired ttest, yielding a significant difference between means. Paired items ( n = 111) revealed that respondents were more likely to have positive attitudes toward hockey (M = 6.86, SD = .43) than they were toward the NHL ( M = 6.31, SD = .88), t (110) = 6.614, p = .000. Overall, this suggests that fan attitude toward the NHL is less positive than fan attitude toward the sport of hockey. The following section provides a discussion of the results of the content analysis and survey, as well as a discussion of the c onclusions of this study. The conclusion wraps up with limitations of this study, suggestions for recommendations, and suggestions for future research.
58 Chapter 5 Discussion and Conclusion Results of the content analys is and survey help inform this study via research questions and hypotheses. The 2004-05 NHL lockout was presented in the coded newspapers as an organizational crisis. Pertai ning to SCCT variables, newspaper writers presented the NHL as having a poor reputation but did not frequently discuss the leagues crisis history. Responsibility for the lockout was split in two ways: Either the players and owners were presented as a cohesive unit equally responsible for the 2004-05 NHL lockout, or blame was placed on the players. Th e content analysis also suggests that the NHL used defensive, blame-based communication strategies when communicating during the 2004-05 NHL lockout. Results of the fan survey indi cate that participants in this study have a somewhat favorable perception of the NHLs reputation; a reputation that has somewhat improved in the previous five years. Overall, results provide partial support for SCCT propositions. Perceptions of crisis responsib ility and prior reput ation positively aff ect perceptions of current reputation, and crisis hi story is related to attitude toward the sport of hockey. The three SCCT threat variables do not significan tly relate to attitude toward the NHL.
59 Additionally, survey results indicate there is an inverse relationship between player responsibility and NHL /owner responsibility, although re sults did not indicate if high perceptions of player re sponsibility negativ ely influence perceptions of the NHLs reputation. However, survey resu lts suggest that participant a ttitudes toward the sport of are more favorable than their attitudes toward the NHL. Discussion By the time the lockout wa s lifted in June 2005, interest in the labor dispute had significantly waned. As one writer put it, the news yesterday that the two sides had tentatively agreed to a new collective bargaini ng agreement didn't exactly arrive as a bolt of lightning. We all knew it was coming ... eventually (Cox, 2005). From the dates analyzed, the lockout got the bulk of its coverage at th e time the season was canceled an unprecedented move in North American sports history; one that gave the NHLs 200405 lockout what Cox (2005) called the dubious distinction of bei ng the longest labour brawl in the history of pro s ports. Most stories (39.7%) we re buried inside the sports section of newspapers. Interest, however, never seemed to be an issue south of the Canadian border; there really wasnt any to begin with. Although f our papers from the United States were analyzed for this study, they accounted for only 20.6% of the articles analyzed. Three Canadian papers made up 78.8% of the code d articles, boosting the not uncommon notion that the NHL garners only minimal interest in the United States. Canadas game ( Globe and Mail, 2004) thrived even without hockey in the form of the NHL. Canadians returned to their outdoor rinks and mi nor league games, essentially taking hockey back to its
60 grassroots beginnings (Kuitenbr ouwer, 2005). In the United States, however, fans simply tuned out. It was reported that replacement programmi ng on ESPNpoker and bowling, for exampledoubled the NHLs average ratings (Penner, 2005). At a time of year when the NHL previously averaged a 0.2 ra ting on ESPN2, replacement programming averaged a 0.4 rating. These and other ongoing issues helped position the 2004-05 NHL lockout was an organizational crisis. It wa s league Commissioner Gary Bettman who came out at the beginning of the lockout suggesting, We owe it to hockeys fans to achieve an economic system that will result in affordable ticket prices and stable franchises. The very future of our game is at stake and the NHL's owners are united We have no other choice, (cited in Duhatschek, 2004). With television ratings already suffering going into the lockout, ESPN exercised its option to walk away from the final year of a $60 million partnership with the NHL before the end of the lockout. Sports pundit J. A. Adande (2005) suggested the NHL cant be considered one of the Big Four professional sports anymore. At the time of the lockout, the National Football wa s spreading $4 billion of television revenues among its 32 teams (Garber, 2005). The NHLs overall revenue in 2003-04 was reported at $2.1 billion per year. The lockout was easily positioned as a reputation-damaging event for the NHL, already considered a niche sport (Heath & El-Bashir, 2004; Foster & Dillman, 2005). As one writer put it, the record 301-day lockout crippled an already fragile industry (Duhatschek, 2005). Situational Crisis Communication Theory (SCCT) variables (crisis responsibility, crisis history, and reputation) were presented in a somewhat negative way in the coded
61 newspaper articles. Items measuring reputati on were not mentioned frequently, and when they were mentioned they were mostly discussed negatively. Financial soundness (n = 145) was the most frequently appearing re putation variable, note d in 51.42% of the articles coded, and coded negatively ( n = 123) 84.83% of those instances. The reputation variable of prospects for future growth accounted for 25.3% (n = 42) of the positivelycoded variables. However, discussion with the intercoder revealed that prospects for future growth often coded positively because th e NHL was in such a poor state that it is almost as though there was nowhere for th e league to go but up. The NHL was not considered well-managed, with 53 of 56 inst ances (94.64%) of management discussions coding negatively. Items measuring crisis history showed th at newspaper writers did not necessarily focus on the leagues labor history. Disc ussions about previous lockouts were significantly higher at the time the lockout was announced an d were typically used as back-story or explanation for the ongoing ev ents. Writers tended to discuss only the ongoing lockout, suggesting that crisis hist ory was not a point of focus during the lockout. This is a positive thing for the NHL, as knowledge of crisis history can significantly heighten reputat ional threats of crises. Finally, crisis responsibility was harder to pinpoint via an analysis of newspaper articles. Although the NHL locked out its play ers and incited the lo ckout, players were still considered responsible fo r the lockout. Results of the content analysis suggest that the crisis can be considered intentional if th e owners and players are viewed as a single entity. At times, both the NHL and NHLPA received blame for the lockout, and the
62 blame appeared to be spread equally. Argume nts were presented s upporting both sides of the labor dispute, and a variety of stakeholder groupsfans, former players, laid-off employees, restaurant and bar ownerseither blamed the sides equally or were equally countered by people with opposing opinions. In keeping this discussion pointed toward the NHL, however, analyzing the leagues communication strategy may round out discussion about crisis responsibility. From the beginning to the end of the lockout, the NHL used de fensive communication strategies. In addition to outright blaming the players association, Bettman used buzzwords and terms aimed at making the lo ckout seem like an unfortunate necessity. While never accepting blame for the state and reputation of the NHL, Bettmans announcements focused on obtaining cost cert ainty in order to make the league successful. The term caught on enough that ev en fans used it when discussing the lockout. In North Carolina, fan Chris Baker blame[d] players for not realizing some kind of cost certainty is necessary for th e NHL to survive (cited in Allen, 2004). The defensive strategy of the league was boosted by regular apologies to hockey fans coupled with suggestions that the NHLPA could have do ne more to prevent the crisis. A Web site was developed to offer the leagues side of the storyit was not reported if the NHLPA did something similar. When the lockout was lifted, Stephen Brunt ( 2005) wrote that both the NHL and NHLPA would still make plenty of money from the new agreement, and suggested that Resentment of that long-standing truth, espe cially in Canada coupled with a brilliant public relations campaign waged by the league has been at the heart of fans siding with the owne rs in this dispute.
63 Overall, these defensive communication stra tegies may have served to separate the NHL from its players association. In doing soand if truly done effectivelythe NHL arguably made themselves victims instead of intentional actors of the crisis. Both of the scenarios were presented in newspapers ar ticles, thus making it hard to determine if the NHL was ultimately presented as being responsible for the lockout. Per SCCT, greater responsibility creates greater reputational threat. In the case of the NHL, and by virtue of reports about the leagues communication strategy, newspaper articles positioned the NHL as both responsible for the lockout and as victims of the lockout. Because research suggests that people turn to media for information (Wilcox, et al., 2005), further research during an ongoing sports labor agreement crisis may show if and how such presentations in the medi a affect stakeholder perceptions. In discussing stakeholders, it is argu able that fans are the NHLs primary stakeholder group. Although res pondents to the fan survey did not comprise a truly representative sample, and results are not generalizable, the survey still yielded significant and interest ing results. Noteworthy issues with the sample of respondents included size of the sample (n = 140) and gender discrepancies. Females have been a growing part of hockeys fan base, highlight ed by numerous female-friendly events hosted by teams (i.e.: the Tampa Bay Lightnings Hockey n Heels events), an influx of womens merchandise in recent years, and in itiatives such as the Washington Capitals Scarlet Caps Hockey for Women Web site (Washington Capitals, 2010). With nearly 60% of respondents identif ying themselves as male, future research would need to find ways to tap into the female hockey fan populatio n in order to gain a truly representative
64 sample of hockey fans. Additionally, respondents strongly agreed with the assessment that they were diehard fans ( n = 105, M = 6.59, SD = .978) and disagreed with the assessment that they were casual fans ( n = 109, M = 2.42, SD = 2.006). This suggests that survey responses and results may be skewed because of involvement with the sport, the league, and the crisis situation. Despite those shortcomings, demographics such as age and location were fairly diverse, thus supporting the use of the onlin e survey tool SurveyMonkey. Hockey fans from a wide range of locations offered their opinions, something an experiment could not have accomplished. For example, an experiment would have limited the diversity of the ages and locations of respondents (especially at a major uni versity in Florida), and the current methodology allowed actual NHL stakehol ders to offer their opinions about a crisis situation that actually took place. All these factors cons idered, the research questions and hypotheses yielde d some interesting findings. In general, RQ1 revealed that partic ipants in this study have a somewhat favorable perception of the NHLs reputati on, an improvement from the time of the lockout. Addressing RQ2, the leagues reputati on improved on six of nine measures of reputation including measures rela ted to concern for fans and prospects for future growth. The NHL is also considered more socially re sponsible than it was five years ago, at the time of the lockout. Although stat istically significant changes were achieved, it should be noted that the mean scores of the reputati on measures never increased by more than .7 points on the provided 7-point Likert-type scales. The most dramatic increase was related to perceptions of NHL honesty. Participants in this survey somewhat agreed with the
65 notion that currently the NHL is honest ( M = 4.53, SD = 1.43), although they identified the NHL as somewhat dishonest ( M = 3.83, SD = 1.54) at the time of the lockout. The slight increases are signs of im provement for the NHL, especially in light of the fact that newspapers portrayed the lockout as highly damaging for the NHL and its stature among fans. Of potential concern for an organization such as the NHL, however, is that no significant changes were noted for items addre ssing if participants in this study believe what the NHL says, if the NHL is considerat e of fans, and if the NHL is well-managed. A lack of significant change shows that opinions on these issues have not changed since the time of the lockout. Furthermore, reputation sc ores were extremely low on the items that showed no significant changes. Participants in this study di sagreed with the notion that the NHL is well managed ( n = 108, M = 2.94, SD = 1.4), do not believe or disbelieve what the NHL says ( n = 107, M = 3.93, SD = 1.31), and dont have an opinion on whether fans are consid ered a top priority (n = 109, M = 4.36, SD = 1.57). Tested hypotheses yielded results that both supported and did not support SCCT propositions examined in this study. Results of a linear regressi on showed support for H1, revealing that SCCT variab les can predict attitude toward the NHL, but crisis history was not a unique predictor of the attitude va riable. This suggests that the NHLs history of labor disputes does not impact fa n attitudes toward the organization. Hypothesis 2 posited that participant per ceptions of SCCT re putational threat variables related to the 2004-05 NHL lockout influence attitudes toward the NHL. A linear regression analysis was not significant, thus not lending support for this hypothesis.
66 Hypothesis 3 suggested that perception of SCCT reputational threat variables do not influence attitude toward hockey, which was su pported by the lack of a significant overall effect of SCCT variables in a linear regression. However, in this analysis crisis history emerged as a unique predictor of attitude toward hockey. An overall analysis of the findings related to H1-H3 suggests that the NHL s current reputation is driven by factors not addressed by SCCT. It shoul d be noted that crisis hist ory did not affect attitude toward the NHL as an organization, but was the only variable that affected attitude toward the sport of hockey. As will be discussed, this may have implications for expanding SCCT literature. As suggested by H4, perceptions of player responsibility were inversely related to perceptions of NHL responsib ility for the lockout. The existence of an inverse relationship suggests that there may be value to the content analysis findings that the NHL was portrayed as both victims and intentional actors of th e lockout. Without knowing why one side is blamed more than the other, it is hard make a connection between responsibility and por trayals of the NHLs responsibility in newspapers. A suggestion is that diehard fans may have been more likely to view the NHL and NHLPA as a single entity, thus sugge sting intentional crisis res ponsibility. On the other hand, casual fans may have been more likely to vi ew the NHL and NHLPA as separate entities in the context of the lockout, thus positioning the NHL as a victim in the crisis. Future research would be needed to test such hypotheses and expand on the finding that an inverse relationship was identified for responsibility for the crisis.
67 The direction of that re lationship, however, was the opposite of what was expected. Respondents of this survey were more likely to blame the owners for the lockout, making it hard to adequately test H5. It is possible that the makeup of the sample (diehard, male hockey fans) impacted the overall assessment of responsibility for the 2004-05 NHL lockout. As such, a linear regres sion analysis of player responsibility, crisis history, and prior reputa tion did not affect attitude to ward the NHL. Had the current survey yielded results that placed more blam e on the players, this study would have been better able to assess the pr oposition set forth with H5. Finally, H6 was supported as attitudes to ward the NHL were less positive than attitudes toward the sport of hockey. Although participant attitudes toward the NHL were positive ( M = 6.31, SD = .88), their attitudes toward hockey were very positive ( M = 6.86, SD = .43). The mean difference, though small, indicates that par ticipants in this study may differentiate between the NHL as an organization and the sport of hockey as an industry. This is also supported by the cont ent analysis findings that suggest the sport of hockey was alive and well during the loc kout in a hockey hotbed such as Canada. What this finding does not show is how casual fans view the sport. Future research may survey members of the general populationma ybe even through use of an experiment to show how closely related disinterest in the sport is related to disinterest in the league. This may be better informed by a study involving the National Football League, for example, as they are a league with a st rong reputation even among people who are not interested in the sport of football.
68 Conclusion At the time of the lockout, concern about the damage sustained by the game of hockey was high. Everyone from managers to players to fans wondered if the National Hockey League (NHL) would be able to rebound from losing a full year of play. Using Situational Crisis Communication Theory (S CCT) reputational threat variables this highly exploratory case study show s that the 2004-05 NHL lockout was a crisis event that may still be affecting stakeholder attitude s toward the NHL. This study, of course, presents only a preliminary pretest that sugge sts this line of inquiry and extension of SCCT should be the focus of future, more rigorous, research me thods. A discussion of such limitations, and suggestions for future studies follows. This study reveals that the NHL faces challe nges related to interest in the sport. In Canada, and among diehard fans, fans are hi ghly involved, active in formation seekers, who are opinionated about the sport. Greater coverage of the even t in Canada suggests these fans are more informed and simply more interested in knowing the details and side effects of something such as a labor disput e. Coverage in the United States was much different, and information presented was more generalized, suggesting a general lack of interest in the dispute. When the NHL or similar organization creates a communication strategy during such a crisis, it seems reasonabl e to suggest that they have to be keenly aware of such discrepancies among stakeholder groups. In the case of the NHL, the league still suffers from the stigma of being a niche sport. Although the leagues re putation has improved in the years since the lockout, it still does not boast a good reputation. How much of th at has to do with the lockout, however,
69 is up for debate. Lingering reputation issues from before the lockout may be a bigger culprit in the overall state of the NHLs re putation than the 200405 NHL lockout. It is possible that the lockout nega tively affected the leagues reputation, but the foundation of a poor reputation had already been set. Of course, this mimics Coombs and Holladays Velcro Effect (2006). Overall, the league has shown signs of improvement even though they still rate poorly on reputa tion variables such as financ ial stability an d management. Even so, the slight improvements mean it is possible that the lockout may have helped. As the president of one sports industry c onsulting firm suggested, You know the old saying, You never get a second chance to make a first impression? The NHL has one of those very rare opportunities fo r a second chance to make a first impression. At a terrible cost, but they have that opportunity, (cit ed in Penner, 2005). Additionally, there has been an influx of young and stellar talentSidney Crosby and Alexander Ovechkin that likely contributes to the im proving reputation of the sport. The use of a multi-method approach also helped inform a variety of questions posed by this study. Use of the content analysis allowed for more in-depth analysis of survey findings. Additionally, the content analys is shed light on some issues related to the crisis, such as which group or organizati on was assigned responsibility for the crisis situation. Particularly, it illuminates the noti on that adjusting SCCT variables can truly alter the way an organization communicati ons during a crisis. Looking at reputation measures and their effects on attitude also s howed what some fans and writers expressed during the lockout: The game of hockey is viewed more favorably than the NHL, an organization in the business or industry of hockey. A combination of the content analysis
70 and the fan survey was what truly informed not only the hypothesis, but the findings. Further, the method of survey distribu tionvia online communitiesallowed this study to be informed by a wide range of res pondents. Although obvious limitations exist, as will be discussed, the most posi tive result of testing this di stribution method is that it reached real fans with opinions about a real crisis event. Ultimately, the 2004-05 NHL lockoutand othe r sports labor disputescan be approached from a variety of different angl es in an effort to expand and build SCCT literature. It should be of part icular interest to dig deeper into the issue of player blame versus league blame because of the nature of sports organizations. Although owners may provide the financial footing fo r organizations, the true produ ct of a professional sports league is its players. If fa ns are unhappy with the players or have negative attitudes toward them, they may not be willing to s upport the athletes. By vi rtue of not supporting the athletes, they ultimately withdraw their support from individual teams or even the league as a whole. This obviously has the po tential to put dents in to the owners bottom lines. Future studies may also work to establish profiles of Canadian and American hockey fans. Like other internationally ope rating organizations, the NHL operates in multiple countries, something unique for a No rth American sports organization. Attitudes and beliefs about hockey held by Canadian s are much different than their U.S. counterparts. Differences may be identified a nd revealed not only th rough more in-depth analysis of newspaper articles about the NHL and hockey, but through additional surveys.
71 In-depth analysis of existing data such as previously administered surveys or polls may also inform such a study. Finally, and most important for developm ent of SCCT, results of this study suggest a variety of areas of future intere st for development or understanding of the theory. Due to the exploratory, case study-base d nature of this st udy, it is acknowledged that the research is fraught with problems. Mo re rigorous future research is necessary to assert whether the following can truly ex tend SCCT development. However, future research may provide insight into some of the findings of this exploratory study. Additional variables of 1) A ttitude and 2) Industry Reputa tion may better inform future development of SCCT principles and propositio ns. Reputation is a variable that affects attitude. Because of the approach SCCT ta kes to reputation management, it is possible that use of the theorys re putational threat variables ma y better inform communication and crisis managers about stakeholder attit udes. This can help communication and crisis managers to not only develop strategies to appr oach crises, but also to develop strategies to approach stakeholde rs following a crisis. Industry Reputation also emerges as a pot ential reputational th reat variable for Coombs theory. For example, and as shown in this study, the NHL organization suffered reputational damage from a cris is event. However, the organizations industrythe sport of hockeydid not suffer reputational damage. Reports were that the lockout may have actually helped improve fan relationships w ith the sport by getting them playing hockey or watching hockey at different, more grassroots levels. Of course, this example applies only to those fans that have an establishe d interest in the game. A better example to
72 inform how a variable such as Industry Reput ation might fit into SCCT may be of the ongoing Toyota recall crisis. It seems reasonabl e to ask, if the automotive industry was not facing widespread reputat ional issues, would Toyota n eed to establish different communication objectives and strategies? It is also worth noting that crisis history was the only va riable that did not affect attitude toward the NHL as an organizati on, but was the only vari able that affected attitude toward the sport of hockey. This sugge sts more research may be needed to assess how crisis history affects organizational re putation. If a variab le such as Industry Reputation is introduced to SCCT, it may link cl osely to crisis histor y, thus affecting the usefulness and development of the theory. Finally, results of this study simultaneously show the strengths and limitations of a theory such as SCCT. The main assumption of SCCT is that a relationship exists between crisis responsibility and organizational reputation (C oombs & Holladay, 2002). As suggested by SCCT, blame and defl ection techniques are appropriate for organizations that are victims in a crisis. This illustrates a strength of the theory because, per previous findings, the NHL used defens ive communication strategies to match the suggested strategies for a victim in a crisis However, it is actually unclear whether the NHL was presented by newspaper writers as be ing responsible as victims or intentional actors during the 2004-05 NHL lockout. This illustrates a limitation in the current understanding of SCCT. Instead of following the steps of SCCT to choose a response strategy during a crisis, it is possible that the NHL reve rsed the steps of the theory by choosing a response strategy to affect perceptions of crisis responsibility. This is a natural
73 reaction for an organization in crisis, however as organizations in crisis often choose communication strategies to manipulate public opinion or perceptions of level of responsibility. It appears as t hough stakeholder groups that were more likely to believe or side with the NHLs messages were also more likely to see the NHL as a victim of the crisis. This presents a theore tical challenge that should be investigated in future SCCT studies because it implies that organizations can manage their reputation by picking a specific response strategy as opposed to si mply following the steps offered by SCCT. This case study, although large in scope, does not extend SCCT literature directly. Instead, it acts as explorat ory, pre-test case study resear ch aimed at showing what additional research may be considered to st rengthen the theory. In all, this case study supports SCCT propositions and suggests additi onal measures may be added to future studies of the theory in an effort to be tter prepare organizations in the face of a reputation-damaging crisis. Limitations Despite best attempts to create a triangulated, multi-method study, this exploratory case study research is fraught with obvious limit ations. The yielded results provide only a partial snap shot of one particular organizationthe NHLand one particular crisis eventthe 2004-05 NHL lockout. Foremost, results of this study are not generalizable. This study cannot directly extend SCCT, although it can provide suggestions for future research. Findings of this study should be examined in future, more rigorous studies.
74 A flaw of the literature review for this study is a lack of assessment about the nature of fandom and fan attitudes. Fan att itudes unique to sports can greatly inform studies linking SCCT to sport-re lated crises. Future research should carefully analyze this stakeholder group and its high rele vancy to sport organizations. A major limitation of this study, related to data collec tion, is the time frame in which the crisis was analyzed. More than five years have passe d since the 2004-05 NHL lockout was announced, creating a handful of data collecti on issues. Due to archiving flaws, it is hard to know if the newspape r articles yielded fo r the study are a truly representative sample of newspaper cove rage of the 2004-05 NHL lockout. Additionally, fans were asked to self-repor t their beliefs from five year s ago. It is hard to know if respondents truly were able to recall accurately their opin ions about the NHL and the crisis from five years ago, thus presen ting a limitation of the survey results. Such flaws in the research methods se verely limit this study. For the content analysis, this study framed the codebook a nd coding guidelines around SCCT principles and propositions. This significantly limited the da ta yielded for analysis as it limited the type of information being coded. The inhere nt nature of the institution of sports journalism also creates issues. Framing or bias es might be present in the articles coded for analysis. It is also impossible to generalize surv ey data collected for this study. Online distribution of surveys has a variety of lim itations. Data collection cannot be controlled for duplicated responses, and only a limited number of stakeholders may be reached. Flaws of such online convenience sampling were illustrated in this study. Noteworthy
75 issues with the sample of respondent s included the size of the sample ( n = 140) and gender discrepancies, with males making up mo re than half of th e respondents (58.6%). Additionally, the survey suffered from attrition, as less than 100 respondents fully completed the survey. Combined, the two data collection methods were not strongly triangulated. Two failed surveysto communication managers a nd playerslimited the potential to create a more holistic assessment of the NHLs ma jor stakeholder groups. And, even if the surveys had been distributed, there are still a variety of stakeholders such as broadcast and merchandising partners who may be survey ed for their perceptions about the 2004-05 NHL lockout and the effects of that crisis situation. Additionally, although the content analysis informed the survey results, the lack of generalizability makes it difficult to confidently assert the strength of the findi ngs without future, more rigorous research. Suggestions for such research follow. Future Research Future research based on this explorat ory case study should ad dress the possibility of linking attitude to SCCT variables. Organizational reputation may be closely linked to attitude toward the organizati on during a crisis, which can be assessed via an experiment or with a survey of stakeholders during an organizational crisis. The National Football League (NFL), for example, may be facing a lockout as its current collect bargaining agreement expires in 2011. A survey of NFL fans may lend insight into how the unique stakeholder group of sports fans reacts to a labor dispute in a hi ghly reputable league.
76 Additional studies may examine the possi bility of adding an industry reputation variable to SCCT. An experiment assessi ng industry reputation as it relates to organizational reputation may help inform SCCT strategies. For example, can the reputation of an industry protect or help an organization in crisis? Or much like Coombs Velcro effect, can a poor industry reputation ma ke an organizational more susceptible to poor perceptions of reputation? As as sessment of current SCCT variables responsibility, crisis history, prior reputation/relationship historyas they relate to organizational reputation, indus try reputation, or both, may al so inform development of SCCT. For example, per findings of this e xploratory study, an investigation into the connection between crisis hist ory and attitude toward an industry may be addressed. Additionally, an experiment may position a re putable organization within the framework of a disreputable industry, and a disreputab le organization within the framework of a reputable industry, to gauge differences and similarities between the two scenarios. Finally, an additional case study ma y use interviews and surveys of communication managers to assess how SCCT propositions are put into practice professionally. Exploratory results of this study suggest the NHL successfully chose their crisis response strategy to manipulate perceptions of their level of responsibility. Their response strategy, matched up with previously developed crisis types and appeared to generate perceptions of responsibility that align with the comm unication strategies suggested by SCCT. This shows support for the theory, but suggest s a limitation of how the theory is applied. A survey of communi cation managers who ha ve experienced an organizational crisis may yield insight into how crisis response strategies are typically
77 chosen, and how those procedures match S CCT steps. Coombs (2007) has previously suggested that communication managers ar e an understudied group, and such a study could fill that gap. Ultimately, development of a theory such as SCCT is for practical use in professional settings. The following recommenda tions suggest ways in which the current study may inform sports-related organizations during a crisis. Recommendations Although fans and fandom were not accura tely defined for this study, it is inherent that sports fans are a primary stak eholder group of sports organizations. Results of this study suggest that during a labor dispute such as the 2004-05 NHL lockout, the organization should maintain a communication strategy that addresses fans and makes them a primary concern. Although the NHL br iefly apologized to fans, the newspaper content analysis revealed that hockey fans felt underappreciated and ignored during the crisis. This left a poor impression on the stakeholder group, creating a sense that the NHL and the NHLPA only cared about their personal stakes in the dispute. For the NHL in particular, gate revenueand t hus, fansis a primary revenue source. In future labor disputes, a sports league should take care to address its fans regularly while acknowledging the strife such a self-created dispute creates. During a labor dispute with a players association, leagues should also choose a communication strategy that does not completely blame the athletes. In the business of sports, players are both the employee and th e product. Placing t oo much blame on the
78 players may alienate fans who do not wish to supportemotionally or financiallya group that is painted as greedy and selfish. L ack of support for the players may translate into lack of support for the sport, thus damaging the reputation and bottom line of a league even if its positioned as a victim in the crisis. In all, this exploratory case study sh ows a variety of ways in which SCCT propositions may be developed. Suggestions fo r future research are aimed at extending SCCT, especially as it can be used with spor ts leagues. Although the results of this study are not generalizable, they offer a brief sn apshot of a crisis eventthe 2004-05 NHL lockoutand how it was handled and received by the NHL, various sports writers and fans.
79 References Adande, J. A. (2005, July 14). NHLs credibility on thin ice. Los Angeles Times p. D8. Allen, K. (2004, Sept. 16). Owners put players on ice. USA Today p. 1C. Batchelor, R., & Formentin, M. (2008). Re -branding the NHL: Building the league through the My NHL integrated marketing campaign. Public Relations Review, 34(2008), 156-160. Benoit, W. L. (1997). Image repair discourse and cris is communication. Public Relations Review, 23 (2), 177-186. Bolch, B. (2005, Feb. 17). Fans Spread Blame fo r Lost Season; They say that greed by both sides in the labor dispute caused schedule to be wiped out. Los Angeles Times p. D9. Brunt, S. (2005, July 14). Puck Peace at last. The Globe and Mail, p. S1. Business Wire. (2006, June 9). National Hockey League Communications Executive Receives Public Relations Pro of the Year Honors from PRSA. Business Wire. Retrieved November 13, 2008, from http://findarticles.com/p/articles/mi_m0EIN/is_2006_June_9/ai_n26892279/pg_1 ?tag=artBody;col1.
80 Canadian Newspaper Association. (2008). 2007 Daily Newspaper Paid Circulation Data. Retrieved September 30, 2008, from http://www.cnaacj.ca/en/aboutnewspapers/circulation. Cancel, A. E., Cameron, G. T., Sallot, L. M., & Mitrook, M. A. (1997). It depends: A contingency theory of accommodation in public relations. Journal of Public Relations Research, 9 (1), 31-63. Cole, C. (2004, Sept. 15). Lockout looms over victory party. National Post p. A1. Coombs, W. T. (1995). Choosing the right word s: The development of guidelines for the selection of the appropriat e crisis-response strategies. Management Communication Quarterly, 8 (4), 447-476. Coombs, W. T. (1998). An analyt ical framework for crisis situations: Better responses from a better understanding of the situation. Journal of Public Relations Research,10 (3), 177-191. Coombs, W. T. (2004). Impact of past crises on current crisis co mmunication: Insights from situational crisis communication theory. Journal of Business Communication, 41 (3), 265-289. Coombs, W. T. (2006). Ongoing Crisis Communication: Planning, Managing, and Responding, 2nd Ed. Thousand Oaks, CA: Sage Publications. Coombs, W. T. (2007). Attribution theory as a guide for post-crisis communication research. Public Relations Review,33(2), 135-139.
81 Coombs, W. T. (2007a). Prot ecting organization reputati ons during a crisis: The development and application of situ ational crisis communication theory. Corporate Reputation Review, 10 (3), 163-176. Coombs, W. T. (2007b). Crisis management: A communicative approach. In C. Botan & V. Hazelton (Eds.), Public Relations Theory II (pp. 171-197). Mahwah, NJ: Lawrence Erlbaum Associates. Coombs, W. T., & Holladay, S. J. (1996). Communi cation and attributions in a crisis: An experimental study in crisis communication. Journal of Public Relations Research, 8(4), 279-295. Coombs, W. T., & Holladay, S. J. (2001). An extended examination of the crisis situations: A fusion of relational management and symbolic approaches. Journal of Public Relations Research. 13 (4), 321-340. Coombs, W. T., & Holladay, S. J. (2002). Help ing crisis managers protect reputational assets: Initial tests of the situational crisis communication theory. Management Communication Quarterly, 16 (2), 165-186. Coombs, W. T., & Holladay, S. J. (2004). Reas oned action in crisis communication: An attribution theory-based approach to crisis management. In D. P. Millar, & R. L. Heath (Eds.), Responding to crisis: A rhetorical approach to crisis communication (95-115). Mahwah, NJ: Lawrence Erlbaum Associates. Coombs, W. T., & Holladay, S. J. (2006). U npacking the halo effect: Reputation and crisis management. Journal of Communication Management, 10(2), 123-137. Cox, D. (2005, July 14). NHL Deal. The Toronto Star p. A01.
82 Duhatschek, E. (2004, Sept. 16). Fans back owners in NHL lockout, poll finds; Season in jeopardy over league push for 'cost certainty'. The Globe and Mail p. A1. Duhatschek, E. (2005, July 14). Players gi ve and receive in new deal: Age for unrestricted free agency will fall from 31 to 27 by 2007. The Globe and Mail p. S3. Elliott, H. (2005, Feb. 17). NHL ends faceoff by canceling season; Negotiations come to a standstill after owners fail to get concessions they say are essential. Los Angeles Times p. A1. Fearn-Banks, K. (2007). Crisis Communication: A Casebook Approach, 3rd Edition Mahwah, NJ: Lawrence Erlbaum Associates. Fink, S. (1986). Crisis Management: Planning for the Inevitable New York, NY: American Management Association. Fombrun, C. J., & Gardberg, N. (2000). W hos tops in corporate communication? Corporate Reputation Review 3(1), 13-17. Formentin, M. (2009, April). The NHL lockout crisis: A qualitative content analysis of newspaper coverage. Panel presentation at th e Graduate Communication Association Interdisciplinary Research Symposium, "Speaking of Method," Tampa, FL. Foster, C., & Dillman, L. (2005, July 14). Gettin g fans back will be a tough sell: King and Mighty Duck officials are optimistic th at their supporters will return once the puck in dropped on a new season. Los Angeles Times, p. D9.
83 Garber, G. (2005, Oct. 26). Unions cu t, revenue sharing center of holdup. ESPN.com Retrieved from http://www.espn.go.com. The Globe and Mail (2004, Sept. 16). Canada's great game stands at a crossroads. The Globe and Mail p. A20. Hagan, L. M. (2007). For reputations sake: Managing Crisis Communi cation. In E. L. Toth (Eds.), The future of excellence in public relations and communication management: Challenges for the next generation (413-440). Mahwah, NJ: Lawrence Erlbaum Associates. Heath, T., & Bashir, E. (2004, Sept. 16). As e xpected, NHL players locked out: Lack of new agreement puts season in jeopardy. The Washington Post p. D1. Houston, W. (2004, September 18). NHL takes th e early lead in public-relations spin game; NHLPA needs better defence if its to catch the well-funded league. The Globe and Mail p. S1. Hsieh, H., & Shannon, S. E. (2005). Three appro aches to qualitative content analysis. Qualitative Health Research, 15 (9), 1277-1288. Infoplease.com. (2008). Top 100 Newspapers in the United States. Retrieved September 30, 2008, from http://www.infoplease.com/ipea/A0004420.html. Jankowski, N. W., & Wester, F. (1991). The qualitative tradi tional in social science inquiry; Contributions to ma ss communication research. In K. B. Jensen & N. W. Jankowski (Eds.), A Handbook of Qualitative Methodologies for Mass Communication Research (44-74). New York: Routledge.
84 Kuitenbrouwer, P. (2005, Feb. 17). On backyard rinks, dream lives: Shinny is breaking out all over the country. National Post p. A3. Mitroff, I., Pauchant, T., & Shrivastava, P. (1989). Can your company handle a crisis? Business and Health 7(5), 41-44. National Post (2005, Feb. 17). Dont blame the owners. National Post p. A20. Neuendorf, K. A. (2002). The Content Analysis Guidebook. Thousand Oaks, CA: Sage Publications, Inc. Pauchant, T.C., & Mitroff, I. (1992). Transforming the Crisis-Prone Organization. Preventing Individual, Organizational and Environmental Tragedies San Francisco, CA: Jossey-Bass Publishers. Penner, M. (2005, Feb. 17). ESPN cant be too upset this reality show is over. Los Angeles Times, p. D2. Podnieks, A. (2005). The Lost Season Bolton, Ontario, Canada: Fenn Publishing Company Ltd. Poindexter, P. M., & McCombs, M. E. (2000). Research in Mass Communication: A Practical Guide Boston, MA: Bedford/St. Martins. Ridings, C. M., & Gefen, D. (2004). Virtual community attraction: Why people hang out online. Journal of Computer-Mediated Communication 10 (1), Art. 4. Rosen, S., & SISanderson, A. (2001). La bour markets in professional sports. The Economic Journal 111 (Feb.), F47-F68. Salutin, R. (2004, Sept. 17). Hockey naught in Canada. The Globe and Mail p. A17. Schilling, J. (2006). On the pragmatics of qualitative assessment: Designing the process for content analysis. European journal of psychological assessment, 22 (1), 28-37.
85 Schmidt, M. B., & Berri, D. J. (2004). The im pact of labor strikes on consumer demand: An application to professional sports. The American Economic Review 94(1), 344-357. Stacks, D. W. (2002). Primer of Public Relations Research New York: The Guilford Press. Stocker, K. P. (1997). A strategic approach to crisis management. In C. L. Caywood (Eds.), The handbook of strategic public rela tions and integrated communications (189-203). New York: Mcgraw-Hill. Stubbs, D. (2005, Feb. 17). Lafleur sounds off agai nst today's players: 'We played with passion'. National Post p. S6. Sung, M. & Yang, S. (2008). Toward a model of university image: The influence of brand personality, external prestige, and reputation. Journal of Public Relations Research 20 (4), 357-376. Thompson, R. (2005, Feb. 17). Cancellation could cost economy $170-million: Most dramatic impact is on draught beer sales. National Post p. A5. Todd, J. (2004, Sept. 16). NHL failed to get TV deal, now it's paying the price: No network contract means less cash, bitter feelings. National Post p. S5. Toronto Star (2004, Sept. 15). Don t weep for NHL or players. Toronto Star p. A26. Wharnsby, T. (2005, Feb. 16). NHL: Hopes flicker; Deadline day: The gap narrows a bit as hockey world mourns a legendary father. The Globe and Mail p. A1. Wherry, A. (2005, Feb. 17). Sides agree on one thi ng: 'it's a sad day': 'We get this close': Owners, players wonder how fans will respond. National Post p. S2.
86 Wilcox, D. L., Cameron, G. T., Ault, P. H., & Agee, W. K. (2005). Public Relations Strategies and Tactics (7th ed.). Boston: Pearson Education, Inc. Wimmer, R. D., & Dominick, J. R. (2003). Mass Media Research: An Introduction (7th edition). Belmont, CA: Wadsworth-Thomson. Woods, A. (2004, Sept. 15). Time runs out to sa ve NHL season: Lockout expected today: Players, owners well prepared for prolonged stoppage. National Post p. B8. Woods, A. (2004, Sept. 16). NHLs future at st ake as players locked out: Current financial system is broken, Bettman says. National Post p. A1. Woods, A. (2004b, Sept. 17). NHL impasse could have been avoided: Eagleson: 'I blame both of them'. National Post p. B9. Woods, A. (2005, Feb. 17). Game over: NHL officially cancels season. National Post p. A1. Wrench, J. S., Thomas-Maddox, C., Richm ond, V. P., & McCroskey, J. C. (2008). Quantitative Research Methods for Communication: A Hands-On Approach New York, NY: Oxford University Press. Yin, R. K. (2009). Case Study Research: Design and Methods (4th edition). Thousand Oaks, CA: SAGE Inc. Zwolinski, M. (2004, Sept. 16). Casualties of labour war are spread far and wide. Toronto Star p. D04.
88 Appendix A: Content Analysis Codebook QUANTITATIVE ANALYSIS 1.) Indicate if the paper is from an American or Canadian newspaper: American 1 Canadian 2 1a.) Please provide the autho rs name, if available: _______________________ 1b.) Please provide the t itle of the article: _______________________________ 2.) Indicate the date of publication: Sept. 15 17 (Lockout announced) 1 Feb. 16 18 (Season canceled) 2 June 13 15 (Agreement announced) 3 3.) Indicate length of story in words: Less than 500 words 1 500-1000 words 2 More than 1000 words 3 4.) Indicate placement of story: Front Page 1 Front Page Sports Sec. 2 Inside Sports Sec 3 Other/Undetermined 4 5.) Indicate source of story: LA Times 1 New York Times 2 USA Today 3 Washington Post 4 The Globe and Mail 5 The National Post 6 Toronto Star 7
89 Appendix A (Continued) 6.) Indicate what type of news piece the story was: News/Feature 1 Editorial/Opinion 2 Letter to the Editor 3 Other 4 7.) To the best of your ability, indicate whether the newspaper story casts the NHL lockout in an overall positive, negative, or balanced light. Please use the provided list of words as a guide in making your assessment. Positive indicates that the lockout is good for the NHL and game of hockey. Negative indicates the lockout is bad for the NHL and game of hockey. Balanced indicates that the story does not necessarily address pros and cons. Not Sure indicates the coder is unsure of the tone of the article. Positive 1 Negative 2 Balanced 3 Not Sure 4 8.) Indicate which sources the author quot ed in the story. Chec k all that apply: NHL Representative = anyone who represents the NHL such as the Commissioner or communication managers. NHLPA Representative = anyone who represents the NHL such as the Executive Director or players. Team/Organization Representative = anyone who represents one of the 30 NHL teams, including general mana gers, communication managers, etc. Outside Partner Representative = anyone who represents a partner organization, such as vendors, broa dcast partners, othe r leagues, etc. Fan = anyone identified as a fan. Other = anyone quoted that does not fit in the abov e categories and has no specific affiliation with the league including experts, lawyers, etc. None = no quotes were used. NHL Rep. 1 NHLPA Rep. 2 Team/Org. Rep. 3 Outside Partner Rep. 4 Fan 5 Other 6 None 7
90 Appendix A (Continued) 9.) Please identify how many individual NHL re presentatives were quoted in this story. None 0 One 1 Two 2 Three 3 Four 4 Five or more 5 10.) Please identify how many individual NHLPA representatives were quoted in this story. None 0 One 1 Two 2 Three 3 Four 4 Five or more 5 SCCT: 11.) Was the situation directly referr ed to as a crisis by the writer? Yes 1 No 2 12.) Please identify which, if any, of the followi ng topics were discussed (check all that apply): History of crisis : involves any discussion of pr evious lockouts, strikes, or negative incidents related to the NHL. Financial issues : involves any discussion of issu es related to league finances, revenue, etc. Fan-related issues : involves any discussion of fans the lockouts effect of fans, fan opinions/reactions. Relationship with NHLPA : involves any discussion of the leagues relationship with NHLPA, including past and present dealings, issues, mutual successes, etc. Relationships with outside organizations: involves discussion of the leagues relationship, past or present, with an y partner organizations. This may include broadcast partners, food and beverage serv ice industry partners, vendors, etc. None of the above: Please note if none of the above issues were discussed. (Continued on next page)
91 Appendix A (Continued) History of crisis 1 Financial issues 2 Fan-related issues 3 Relationship with NHLPA 5 Relationships with outside organization 6 None of the above 99 13.) Was the NHLs labor history addressed? This can include mention of previous strikes, lockouts, or labor disagreements. Yes 1 No 2 N/A 99 14.) Was the NHL portrayed as being concerned for fans? Yes 1 No 2 N/A 99 15.) Was the NHL portrayed as a league with prospects for future growth? Yes 1 No 2 N/A 99 16.) Was the NHL portrayed as socially responsible? Yes 1 No 2 N/A 99 17.) Was the NHL portrayed as being an honest organization? Yes 1 No 2 N/A 99
92 Appendix A (Continued) 18.) Was the NHL portrayed as a financially sound organization? Yes 1 No 2 N/A 99 19.) Was the NHL portrayed as being a well-managed organization? Yes 1 No 2 N/A 99 20.) If the NHL was discuss as having a st rained relationship at any point during the article, please identify if any of the follo wing stakeholder groups were specifically mentioned: Fans 1 Players 2 Comm. Managers 3 Vendors 4 Media Partners 5 Sponsors 6 Other 7________ N/A 99
93 Appendix A (Continued) QUALITATIVE ANALYSIS 1.) Using the provided list of words and examples of positive/negative phrases as a guide, indicate if the NHL as an organization was portrayed positively or negatively. If the NHL was portrayed in a neutral manner, or you are not sure, choose neither. Positive 1 Negative 2 Neither 3 1a.) Please briefly describe the main reason the NHL was portrayed either positively or negatively: 2.) If the NHLs communication strategy was addressed, what was the major point of discussion? Communication strategy may involve descriptions of how the NHL reached out to the NHLPA, the media, fans, or other entities, during the lockout. This can involve mention of press conferences, written communication, interviews, etc. SCCT: 3.) If the NHLs labor history was a ddressed, what history was discussed? 4.) If the NHLs reputation was addressed, what was discussed? The following are topics that elaborate on reputation: Was the NHL discussed as being: concerned for fans a league with prospects for future growth socially responsible trustworthy or honest financially sound well-managed?
94 Appendix A (Continued) CODING GUIDE Positive/Negative Word Lists The following are words that have been identified as either positive or negative in meaning. Positive Negative able Accepted battle bitter agreeing Committed blame broken fan Fixed canceled cancellation helped Lucrative damage death negotiating Negotiations difficult dispute passion Power failed frustrating profit Progress greedy impasse successful Survive impose lies trust Win lockout lose lost problem refused rejected strike suffer tragedy trouble unable uncertain war
95 Appendix A (Continued) Positive/Negative Phrases The following phrases have been identified as favorable or unfavorable toward the NHL When coding, please keep in mind th at most questions are asking about how the NHL is being portrayed during the 2004-05 lockout. Sometimes, a phrase may be positive or negative in connotation, but it may not pertain to the NHL as an organization. Examples of 1. positive and favorable, 2. negative and unfavorable, and 3. both positive/negative and neither favorable /unfavorable statements have been provided. Additionally, words with agreed upon positiv e and negative connotations have been identified to help put phrases in context: Bolded words have been identified as having negative connotations Underlined words have been identified as having positive connotations. Positive/Favorable to NHL Cluster: 1. Bettman assured fans that the league would rise from the ashes of a lost season 2. more fan-friendly entertainment and an earnest attempt at creating more offence 3. Their talk of cooperation [between the NHL and NHLPA] was a stark contrast to their old philosophical differences 4. The hockey world will soon be able to celebrate the return of NHL hockey. 5. focus on partnership and cooperation with both sides apparently committed to mutually improving marketing and sustaining a l eague that will now have to rebuild what was a damaged gamed before it disappeared. 6. live from New York, in a professi onally orchestrated, televised press conference setting. [Mr. Bettmans ] opening statement was assertive Negative/Unfavorable to NHL Cluster: 1. The lockout will certainly mean an adjustment for hockey fans and business. 2. Its likely that such a move would sour relations with the leagues corporate sponsors and self-described broadcast partners 3. The NHLPA has rejected the leagues proposals, fearing it could result in a ceiling on salaries.
96 Appendix A (Continued) 4. After failed labor negotiations that le d to the NHL seasons being canceled Wednesday, officials from both local t eams said they expect holding on to business to be a struggle 5. Both sides clearly failed . 6. The owners have screwed this league up so badly under Bettmans watch that only a lockout of a season or more could righ t all that they have made wrong since their last lockout in 1994. 7. Joe Fan is the big loser in the NHLs economic battle. Neither Favorable nor Unfavorable to NHL Cluster: 1. Positive: The league also will unveil rules changes that could include allowing two-line passes and intro duction of an overtime shoot out to eliminate regularseason ties. 2. Negative: Both the league and its players union said they are prepared to endure a dispute for months or even years rather than move off their entrenched positions. 3. Negative: The union said 228 players voted in person and the rest by e-mail. Their defeat on many issues appeared to leave Goodenows position untenable 4. Negative: the lockout is particularly distressing Neither Positive nor Negative Statement Cluster: 1. Favorable: somehow, hearing the owners say Were tired of losing money was more palatable 2. Unfavorable: For a decade Goodenow pulverized Bettman in every meaningful aspect of hock eys economic picture.
97 Appendix B: Fan Survey Welcome The following survey is designed to measure your attitudes about the NHL. The results of this survey will be used in a study analyz ing the NHL's reputation following the 2004-05 lockout. The survey will take less than 15 minutes to complete. Your input is invaluable to my research, and your participat ion is highly appreciated. Any collected survey data will remain conf idential and anonymous. If you have any questions, please feel free to contact me at firstname.lastname@example.org Thanks so much for your time, and I look forward to hearing from you. Sincerely, Melanie Formentin You are invited to participate in a research study about the National Hockey Leagues (NHL) reputation following the 2004-05 NHL lockout. If you agree to be part of the research study, you will be asked to answer a series of questions about your attitude towa rd the NHL and hockey. Bene fits of this research include a better understanding of th e NHLs reputation. Participants of this study will not face any known risks or discomforts. Participating in this study is completely voluntary. Even if you decide to participate now, you may change your mind and stop at any time. You may choose not to answer any survey question for any reason. If you have questions about this research study, you may contact Dr. Kelly Page Werder at the University of South Florida School of Mass Communications at email@example.com or University of South Florida, School of Mass Communications, 4202 E. Fowler Ave. CPR 107, Tampa, FL 33620. The University of South Florida Institutional Review Board has determined that this study is exempt from IRB oversight. By continuing to the following pages, I agree to participate in the study. Survey Questions: Part 1 A Using the 7-point scales, please select the num ber that best corresponds with your attitude toward the NHL. 1. My attitude toward the NHL is: Good Bad Positive Negative Favorable Unfavorable
98 Appendix B (Continued) Survey Questions: Part 1 B Using the 7-point scales, please select the num ber that best corresponds with your attitude toward the sport of hockey. 2. My attitude toward hockey is: Good Bad Positive Negative Favorable Unfavorable Survey Questions: Part 2 The following questions refer to your attitude about the NHL prior to the 2004-05 NHL lockout. To the best of your ability, please tr y to recall your attitude s about the NHL prior to the lockout Using the following 7-point scale, please indicate your overall agreement with the following statements: (Using Likert-type scale of (1) Strongly Disagree, (2) Disagree, (3) Somewhat Disagree, (4) Neither Agree/Disagree, (5) Somewhat Agree, (6) Agree, (7) Strongly Agree.) 1. Five years ago, the NHL was concerne d with the well-being of its fans. 2. Five years ago, the NHL looked like a sports league with strong prospects for future growth. 3. Five years ago, the NHL was socially responsible. 4. Five years ago, the NHL was honest. 5. Five years ago, I was likely to believe what the NHL said. 6. Five years ago, the NHL was concerne d with the well-being of its fans. 7. Five years ago, the NHL was considerate of fans. 8. Five years ago, the NHL was financially sound. 9. Five years ago, the NHL was managed well. Survey Questions: Part 3 The following questions refer to your attitude about the NHL TODAY. To the best of your ability, please respond to how you feel about the NHL now. Using the following 7-point scale, please indicate your overall agreement with the following statements: (Using Likert-type scale of (1) Strongly Disagree, (2) Disagree, (3) Somewhat Disagree, (4) Neither Agree/Disagree, (5) Somewhat Agree, (6) Agree, (7) Strongly Agree.) 1. The NHL considers fans a top priority. 2. The NHL looks like a sports league with strong prospects for future growth. 3. The NHL is well-managed. 4. The NHL is socially responsible. 5. The NHL is financially sound.
99 Appendix B (Continued) 6. The NHL is concerned with the well-being of its fans. 7. The NHL is basically dishonest. 8. Under most circumstances, I would be likely to believe what the NHL says. 9. The NHL is NOT concerned with the well-being of its fans. Survey Questions: Part 4 Youre almost done with the su rvey! Thanks for your patience. The following are general questions about the NHL and the 2004-05 lockout. To the best of your ability, please indicate your level of agreement with the following statements. This section continues on th e next page of the survey. Using the following 7-point scale, please indicate your overall agreement with the following statements: (Using Likert-type scale of (1) Strongly Disagree, (2) Disagree, (3) Somewhat Disagree, (4) Neither Agree/Disagree, (5) Somewhat Agree, (6) Agree, (7) Strongly Agree.) 1. I remember the 2004-05 lockout. 2. I consider myself a casual hockey fan. 3. I am more excited about the NHL than I was five years ago. 4. I believe the NHL could have done more to prevent the lockout. 5. I believe the lockout was a result of greedy players. 6. I am aware of the player strike that occurred in the early 90s. 7. I believe the NHL should have apologized for the lockout. 8. I am more supportive of the NHL than I was five years ago. 9. The blame for the lockout lies with the NHL. 10. I believe the owners were at fault for the NHL lockout. 11. I believe the players should have ta ken responsibility for the lockout. 12. I believe the NHL should have taken responsibility for the lockout. 13. The blame for the lockout lies with the players, not the NHL. 14. The NHL is not to blame for the lockout. 15. I consider myself a diehard hockey fan. 16. I am aware that there have b een two lockouts in NHL history. 17. The NHL and the NHLPA are the same organization. Using the 7-point scale, please indicate your overal l impression of the NHL: 18. Overall, my impression of the NHL is: (1) Very Unfavorable, (2) Unfavorable, (3) Somewhat Unfavorable, (4) Neither Favorable/Unfavorable, (5) Somewhat Favorable, (6) Favorable, (7) Very Favorable
100 Appendix B (Continued) Demographics 1. Age: 18 or younger 19-24 25-34 35-44 45-54 55-64 64+ 2. Gender: Male Female 3. Race/Ethnicity: African American Asian/Pacific Islander Caucasian Hispanic/Latin Native American Other ____________ 4. Education: High School Some college College graduate Some postgraduate work or higher 5. Income: $24,999 or less $25,000 $39,999 $40,000 $54,999 $55,000 $69,999 $70,000 $84,999 $85,000 $99,999 $100,000 or more 6. State/Province of Residence: _______________ 7. I have been a hockey fan for ____ years. 8. Where did you find this survey? ____________