SELF - EFFICACY AND PERFORMANCE IN VOLLEYBALL REFEREES By Benjamin D. Spencer A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Kinesiology - Master of Science 2015 ABSTRACT SELF - EFFICACY AND PERFORMANCE IN VOLLEYBALL REFEREES By Benjamin D. Spencer Sport officials are an under - researched subpopulation in the sport psychology literature. Particularly little is known about psychological factors that may predict officiating performance. Feltz and Guillen (2011) suggested that self - efficacy may influence performance in the refereeing context, as it does in many others. Myers, Feltz, Guillen, and Dithurbide (2012) indicated that referee self - efficacy is composed of four dimensions: Game Knowledge, Decision - Making, Pressure, and Communication. The current s tudy sought to evaluate the relationship of these various dimensions to performance in several aspects of officiating. A secondary purpose was to evaluate proposed sources of referee efficacy as predictors of referee efficacy dimensions and performance. Volleyball referees ( N = 76) who were candidates for USA Volleyball (USAV) National or Junior National badges completed a survey which measured experience as an official, experience playing and coaching volleyball, referee self - efficacy, and sources of ref eree self - efficacy. Following administration of the survey, participants completed the USAV referee performance evaluation protocol. No relationship was found between self - efficacy and performance in high - level volleyball officials. Little was found relati experience and performance in their evaluations, and few connections were identified between previously established sources of referee confidence and dimensions of referee self - efficacy. These null findings may be due to lack of varia nce in ability and confidence on the part of the referees, or produced by an evaluation system which is designed to teach candidates, and pass most of them, rather than explic itly evaluate their performance . iv For my wife, Meghan, who has made so many sacrifices so that I may continue to learn, and without whom I never would have finished Figure 2; And f or my father, who set me on this path long ago, though I doubt either of us would have guessed it would look like this. v ACKNOWLEDGMENTS First, I would like to acknowledge the support and guidance of my advisor, Dr. Deborah Feltz, in the development of this thesis and during my two years at Michigan State. I must also thank the members of my committee, Dr. Al Smith and Dr. Francisco Villarr uel, for their patience and their helpful feedback. The support of my class - and lab - mates was priceless to me as I worked through this Lincoln , Tayo Moss , Steve Samendinger , Alison Ede, Emery Max, and Ch ristel Beverly: I could not have done this without you. Special mention must be given to Anthony Delli Paoli and Dr. Andy Driska, for helping me get moving had a n egative experience with anyone in the Kinesiology department, which is astonishing to think about. As always, I am nothing without my family. The love and support of my parents , Dave and Pat; my brother , Nick; and my wife , Meghan are , as ever, the foundat ion upon which my life is built. our heads and food on the table, after I dragged her halfway across the country to be unemployed. Finally, I must thank the wonderful pe ople at USA Volleyball, in particular Michelle I felt welcomed and was accommodated during this project beyond my wildest expectations. To all of you, thank you. vi TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ..................... vi i LIST OF FIGURES ................................ ................................ ................................ ................... viii CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ..................... 1 Nature o f t he Problem ................................ ................................ ................................ .... 1 Purpose o f t he Study ................................ ................................ ................................ ...... 6 Proposed Model ................................ ................................ ................................ ............. 7 Delimitations ................................ ................................ ................................ .................. 7 Limitations ................................ ................................ ................................ ..................... 8 Definitions ................................ ................................ ................................ ...................... 8 CHAPTER 2 REVIEW OF LITERATURE ................................ ................................ ................................ .... 9 Self - E fficacy T heory ................................ ................................ ................................ ...... 9 Sources of E fficacy I nformation ................................ ................................ .................... 11 Self - E fficacy and P erformance ................................ ................................ ...................... 13 Self - E fficacy in Sport O ffic i als ................................ ................................ ..................... 14 CHAPTER 3 METHOD ................................ ................................ ................................ ................................ .. 18 Participants ................................ ................................ ................................ ..................... 18 Instruments ................................ ................................ ................................ ..................... 18 Data Collection Procedures ................................ ................................ ............................ 20 Data Analyses ................................ ................................ ................................ ................ 21 CHAPTER 4 RESULTS ................................ ................................ ................................ ................................ .. 2 3 Preliminary Analyses ................................ ................................ ................................ ..... 2 3 Descriptive Statistics, Gender Differences, and Differences between Can didate Levels ................................ ................................ ................................ ............ 2 5 Correlations among the Variables ................................ ................................ .................. 2 6 Tests of Components of the Proposed Model ................................ ................................ 2 9 CHAPTER 5 DISCUSSION ................................ ................................ ................................ ............................ 3 2 Limitations and Future Directions ................................ ................................ ................. 3 7 Conclusion ................................ ................................ ................................ ..................... 3 8 APPENDICES ................................ ................................ ................................ ........................... 3 9 Appendix A : Consent Form ................................ ................................ ........................... 40 vii Appendix B : Demographics and Sport Official Self - Rating Scale ................................ 41 Appendix C : Refficacy Questionnaire ................................ ................................ ........... 43 Appendix D : USAV/PAVO Rating Sheets ................................ ................................ .... 45 Appendix E : Practical Rating Sheet Instructions ................................ ........................... 49 Appendix F : PAVO/USAV Rating Criteria ................................ ................................ ... 50 REFERENCES ................................ ................................ ................................ .......................... 5 2 viii LIST OF TABLES Table 1. Adjusted ite ms in the Referee Self - Efficacy Scale ................................ ...................... 19 Table 2. Summary Statistics for Normality and Internal Consistency ................................ ....... 23 Table 3. Descriptive Statistics for Variables ................................ ................................ .............. 2 5 Table 4. Pearson Correlation Coefficients between Sources and Dimensions of Referee Self - Efficacy ................................ ................................ ................................ ................. 2 7 Table 5 . Pearson Correlation Coefficients between Performance Predictors and Performance Ratings ................................ ................................ ................................ .......... 2 7 Table 6. The Predictability of Referee Self - Efficacy Sources and Experience on Referee Self - Efficacy Dimensions ................................ ................................ ........................ 2 9 Table 7 . The Predic tability of Referee Self - Efficacy Sources and Dimensions on Performance ................................ ................................ ................................ .......................... 30 ix LIST OF FIGURES Figure 1 : Conceptual Model of Referee Self - Efficacy (Guillen & Feltz, 2011 ) ....................... 5 Figure 2: The Hypothesized Model ................................ ................................ ........................... 7 Figure 3: USAV/PAVO First Referee Rating Sheet ................................ ................................ . 45 Figure 4: USAV/PAVO Second Referee Rating Sheet ................................ ............................. 47 Figure 5: Practical Rating Sheet Instructions ................................ ................................ ............ 49 Figure 6 : PAVO/USAV First Referee Rating Criteria ................................ .............................. 50 Figure 7 : PAVO/USAV Second Referee Rating Criteria ................................ ......................... 51 1 CHAPTER 1 INTRODUCTION Nature of the Problem Research in sport psychology is traditionally focused on coaches and athletes. Referees are a sub - population with important roles in sports, but are largely ignored i n the current literature. McInman (1997) analyzed four major sports psychology journals over a 10 - year period, and found that only 1.12% of articles addressed or involved officiating. The frequency with which players, coaches, and spectators blame offici implies that an increased understanding of the factors underlying referee performance would be productive for nearly everyone involved in sports at any level. The most important predictor of referee performance is experience. Years of experience officiating, number of matches officiated, and hours of practice are positively related to skill (Catteeuw, Helsen, Gilis, & Wagemanns, 2011), although officiating is somewhat unique in that deliberate practice is difficult (MacMahon, Helsen, Starks, & Weston, 2007). The best way to practice is simply to referee more matches. Experience playing or watching the game may also be helpful (Pizzera & Raab, 2012), and in some sports may be able to substitute for refereeing expe valuable than extensive experience as a referee. This means that for some sports, officials should cease competitive play and specialize as a referee early, while in o thers fast - tracking former high - level competitors may be the most productive avenue for producing high level referees fast - tracked competitors make the best officials, but it appears that former competitors tend to 2 make better referees in very specific, technical sports such as judo and trampolining, while more traditional sports (like volleyball) tend to favor early specialists. Defining and measuring referee performance can be challenging. While athletes and coaches can often be evaluated on objective, observable outcomes wins and losses, times, scores, batting averages, etc referees must be evaluated on frequent and often subjective interpretations of incid ents during play. Rather than producing a quantifiable end result, sports officials facilitate a safe and fair contest for the participants to a greater or lesser degree. This broad directive results in two general approaches to assessing or studying ref eree performance: an incident - by - incident, right - or - wrong objective analysis of specific decisions during the course of a contest; and a more holistic appraisal of more subjective elements of officiating such as match control and personnel management. Rev iewing this literature is complicated by the reality that specific officiating tasks vary from sport to sport, however it is reasonable to assume that some commonalities can be inferred. The first approach lends itself well to laboratory - setting study of p erceptual and decision - making processes. MacMahon et al. (2007) showed that referees were more accurate than soccer players in judging videotaped challenges, implying that incidental decision - making is a skill that referees develop through experience and practice. Catteuw et al. (2011) went a step further, finding that soccer referees and assistant referees (linesmen) each perform better on tasks specific to their role. The oft - studied officiating biases are apparent in individual decisions. Multiple st upon officials. This effect has been identified across multiple sports, leagues, and countries, including the English Premier League (Boyko, Boyko, & Boyko, 2007) and the NBA (Lehman & Reifman, 2001). Other studies have noted racial biases (e.g. Wagner - Egger, Gygax, & 3 Ribordy, 2012), differences when officiating men versus women (Souchon et al. 2004, 2009a, 2009b, 2010), and even a bias against taller athletes (van Quaqueb eke & Leissner, 2010). The second, more holistic approach to evaluation is more difficult and less common in research. Game management is sometimes quantified as use of the oft - - which can be conceptualized as another form of bias. NCAA basketball referees are more likely to call fouls on teams which so far have fewer fouls than their opponents, and that tendency increases in strength as the gap widens (Anderson & Pierce, 2009). Lopez and Snyder (2013) noted that NHL referees will issue make - up calls over the course of a game to even out the number of penalties assessed to each team, with the aim of achieving perceptions of balance and a penalty to a team that has already received one, and more likely to award a penalty to a team who has previously had one given against them (Plessner & Betch, 2001). Referees manage the flow of a competition by adjusting their decision - making depending on the context. For example, soccer officials who view recorded incidents in the context of a match award fewer yellow cards than when watching incidents in random sequence, and are more likely to award yellow cards when told that an incident is late in t he match rather than early (Unkelbach & Memmert, 2008). Despite the difficulties and complexities of evaluating referee performance, leagues or for assignmen t, and to assure quality for the competition. Accordingly, they must either find assessment tools in the literature or create their own. Anshel (1995) developed the Behaviorally Anchored Rating Scale for Basketball Referees (BARS - BR) in an attempt to pro vide an objective measure of officiating performance. Unfortunately, application of the measure is, as 4 Behavior (Trudel, Cote & Sylvestre, 1996) quantifies the tim e that referees spend in various activities (e.g. monitoring without interaction, intervening verbally or with gestures), but does not provide a judgment of performance quality. Leagues and referee organizations attempt to evaluate their officials through observation, approximately objective scoring and feedback from veteran and expert referees. The population of interest in the present study is USA Volleyball (USAV) referees. U SAV uses their own rating sheets to evaluate performance in candidates for Junior National or National certification. Experienced, high level referees rate candidates on multiple dimensions of performance, including judgment (e.g., consistency), mechanics/signals (e.g., scanning the court before beckoning for the serve), positioning/focus (e.g., watching each ball contact), match control (e.g., warm - up administration), communication with match participants (e.g., demeanor and approachability), communication with officiating team (e.g., interactions with line judges, scorekeeper, and second referee), and professionalism (e.g., appearance). These ratings serve as the performance outcome measure in the present study. Notably, this instrument attempts to score candidates on both the subjective, match control tasks and the objective, single - incident tasks. Guillen and Feltz (2011) proposed a conceptual model of referee efficacy, which they defined as the extent to which referees believe they have the capacity to perform successfully in - efficacy, referee self - efficacy was suggested to include six dimensions: game knowledge, decision - making skills, psychol ogical skills, strategic skills, communication/control of the game, and physical fitness. Guillen and Feltz proposed four sources of referee self - efficacy based on the Sources of Sport Confidence Scale ( Vealey, Hayashi, Garner - Holman, & Giacobbi , 1998). These sources include mastery 5 experience, significant others, physical and mental preparation, and partner qualifications. Finally, the authors speculated that self - efficacy would influence referee behavior, satisfaction, stress, and performance, as well as athlete rule violations and coach behavior. Figure 1: Conceptual Model of Referee Self - Efficacy (Guillen & Feltz, 2011) Myers, Feltz, Guillen, and Dithurbide (2012) expanded on that theoretical framework to develop a specialized measure for evaluati ng self - efficacy in referees. The Referee Self - Efficacy Scale (REFS) consists of 39 items which measure four factors of referee self - efficacy: game knowledge, decision making, pressure, and communication. Game knowledge was defined as 6 the confidence that referees have in their knowledge of their sport, including rules, officiating mechanics, and basic game strategy. Decision making was defined as the confidence that referees have in their ability to quickly and firmly make decisions during competition. Pressure was defined as the confidence that referees have in their ability to be uninfluenced by pressure from players, spectators, and coaches. Communication was defined as the confidence that referees have in their ability to communicate effectively wit h other referees, coaches, players, and auxiliary personnel. Myers et al. demonstrated factorial validity for the four - dimension REFS with a large sample of referees representing 15 different team sports from the US and Spain. In addition, the authors show ed support for the sources of referee self - efficacy as significant predictors of the four dimensions of REFS. While Myers et al. (2012) provided preliminary support for the first part of the Guillen and Feltz (2011) model, no research has examined the ou tcomes of self - efficacy in referees such as performance. Myers et al. suggested that such an investigation could make an important contribution to the literature, especially if examined simultaneously with proposed sources of referee efficacy (e.g., explor ing dimensions of referee efficacy as mediators ). Purpose of the Study The purpose of the present study was to examine the predictive strength of referee self - efficacy on officiating performanc e in volleyball referees. This research represents a clear gap in the existing literature: self - efficacy and performance have never been studied together within the referee population. Additionally, this research sought to explore the mediati ng role of the dimensions of referee efficacy between referee self - efficacy s ources and performance in volleyball referees. 7 Proposed Model The study tested the hypothesized relationships in the proposed model (see Figure 2 ). Figure 2 : The H ypothesized M odel . The proposed sources of referee efficacy are hypothesized to predict the dimensions of referee efficacy identified in Myers et al. (2012) as well as the various dimensions of performance on which volleyball referees Delimitations 1. The population was delim ited to USAV volleyball referees who are candidates to become National or Junior National officials. 2. The referee self - efficacy of the officials was measured by the Referee Self - Efficacy Scale (REFS) (Myers et al., 2012). 3. self - efficacy was measured by the Sport Officials Self - Rating Scale (Guillen & Feltz, 2011) . 4. Referee performance was measured by the USAV First Referee Rating Sheet . 8 Limitations 1. Myers et al. (2012) suggest ed a minimum sample size of 300 for use of the REFS . 2. Not all possible determinants of referee performance were measured in this study. Definitions 1. Referee efficacy : The extent to which a referee believes that he or she has the ability to successfully officiate a competition (Myers et al., 2012). 2. Referee, o fficial, sport official : These are used interchangeably throughout this manuscript, and can also be assumed to include other sport - specific titles such as umpire, judge, or technical official, which refer to an authority figure responsible for presiding over a sport competition a nd enforcing the rules from a neutral point of view. 3. Self - efficacy : Academically, self - and execute the courses of action required to produce given attainments (Bandura, 1997). 9 CHAPTER 2 REVIEW OF LITE RATURE Self - E fficacy Theory The theory of self - efficacy (Bandura, 1977, 1997, 2001) was developed within the framework of social cognitive theory. Social cognitive theory view s behavior from an agentic pers pective (Bandura, 2001); that is, individuals use forethought, self - reflection, and self - regulation to influence their own functioning rather than passively react to their environment. These agentic behaviors interact along with personal factors and environmental conditions to determine motivation and be havior (Bandura, 1986, 1997). Self - referent thought mediates the capabilities and self - perceptions of efficacy affect th eir motivation and performance. Self - efficacy is , according to Bandura, the most influential form of self - belief. Bandura defines self - perform a task at given levels. Efficacy involves not only knowing what behavior is appropriate for a si tuation, but also organizing cognitive, social, and behavioral strategies and skills to produce the correct action. Thus, judgments of efficacy are not based on skills alone, but instead on what an individual can do with the skills they possess. Efficacy e xpectations, according to Bandura, should not be confused with outcome expectations. While efficacy expectations reflect consequence of a behavior, such as recogniti on, rejection, rewards, or punishment. The critical distinction is that an individual might believe that certain behavior will result in a desired outcome, but their execution of that behavior in the end will be more dependent upon their beliefs in their c apability than on their beliefs in regard to outcome. 10 Self - efficacy beliefs vary and are measured across three dimensions: level, strength, and generality (Bandura, 1997) - efficacy reflects their expected performance at a give n level of difficulty. Volleyball referees with disparate levels of self - of such situations they could assess correctly in a given number of opportunities (e. g., 1 out of 10, 5 out of 10, 10 out of 10). Strength of self - certainty that they can attain a level of performance. Two referees might both believe that they can correctly call 10 out of 10 possible ballhandl ing infractions, but one might be much more certain about their ability to do so than the other. Generality refers to the degree to which an individual considers themselves efficacious in numerous tasks or domains, or to transfer efficacy judgments from on e task to another. A referee with a large degree of generality in their efficacy beliefs might be able to transfer their efficacy for calling ballhandling infractions into other decisions, or even other sports. This is important, because self - efficacy beli efs are not a global trait for an individual; rather they are specific to distinct domains of functioning and even specific aspects of a given domain. For example, a referee might have high levels of efficacy for judging infractions, but low efficacy for c ommunication with coaches and players. Feltz, Short, and Sullivan (2008) identify several distinct types of efficacy beliefs relevant to sport, many of which may be importa nt to the referee subpopulation. The most straightforward type is perhaps task self - efficacy , beliefs about performing a particular task with graded levels of difficulty. The previously mentioned ballhandling infractions are an example of task self - efficacy. Ameliorative or coping efficacy bility to manage perceived threats. For referees, these threats might appear in the forms of stress from pressure applied by match participants, coaches, and specatators, distractions or difficulties 11 presented by the tournament environment, anxiety resulti ng from the evaluation process, or a need to recover from mistakes. Performance efficacy successfully complete a task at a specific time or in specific context, rather than in general. An official might feel go od about their ability to judge ballhandling typically, but for whatever reason not feel so confident on the day of a tournament. Collective efficacy shared beliefs in its ability to perform at a given level. For officials in many sport s, including volleyball, communication and teamwork among a group of referees is crucial for successful performance. Other types of efficacy beliefs less pertinent to the current study include self - regulatory efficacy , learning efficacy , and competitive ef ficacy . Sources of Efficacy Information Bandura (1997) proposes four sources of efficacy information: past performance, vicarious experience, verbal persuasion, and physiological/emotional states. These sources impact (such as persistence, effort, and choice) and thought patterns (such as goals and attributions) indirectly via their influence on efficacy expectations. A person may draw on one or more of these sources to form their efficacy beliefs, and the salience of the var ious sources may differ across individuals or tasks. Sources of past performance efficacy information mastery experiences and accomplishments. Usually, efficacy expectations will increase with past performance that the ind ividual perceives as successful , and lower with perceived unsuccessful exper iences. The efficacy value of performance experiences is also impacted by task difficulty, temporal pattern of success and failure, effort expended, and guidance received (Bandura, 1982). Success with minimal effort on tasks that are considered difficult implies high ability and enhances efficacy beliefs. 12 conception of an ability as an inherent aptitude as opposed to an acquired skill (Bandura, 1997). Bandura considers past performance to be the most influential source of efficacy information. Vicarious experience involves observation and comparison of oneself to others performing a task. Watching or simply visualizing others succeed can increase self - efficacy for a task, while seeing others fail to perform successfully can lower efficacy expectations. This modeling effect is strongest when the observer is similar to the model. Efficacy beliefs can also be affected by so cial comparison with others. Weinberg, Gould, and Jackson (1979) showed that self - efficacy could be manipulated by portraying a competitor as competent or incompetent. Self - - effi cacy (Dowrick & Dove, 1980). This may extend to imaginal experiences in which an individual visualizes themselves or others behaving successfully or successfully, though Maddux (1995) argues that imaginal experiences should be considered a distinct source of efficacy information, rather than being included in vicarious experience. While vicarious experiences are believed to be a weaker source of self - efficacy than past performance, they remain influential. When an individual is convinced through vicarious e xperience that they lack efficacy for a task, they may act in ways that confirm that notion. Verbal persuasion comes in feedback as well as self - talk, positive imagery, and other cognitive strate gies . According to Bandura, it is easier to lower efficacy beliefs through criticism than it is to raise them through iness to the individual receiving the feedback, as well as the realistic nature of the information imparted. Individuals can regulate their thought processes 13 to convince themselves that they can perform at a certain level through self - talk (Feltz et al., 2 008). Physiological and emotional states which influence self - efficacy include autonomic arousal, pain, fear, fatigue, and stress, among others. In general, positive states and emotions, or the lack of negative states, will increase efficacy expectations, while negative physiological states will lower efficacy expectations. Individuals evaluate and interpret these states differently; for example, one person might interpret the feeling of butterflies in their stomach as anxiety and an indication that they a re not prepared for a coming experience, while another individual might associate that same feeling with readiness to perform. Feltz et al. (2008) separated physiological information and emotional states into independent categories for sources of self - efficacy in sport contexts because they relate to different aspects of performance. With regard to the current study: physiologica l information is a more important efficacy source for physical tasks as opposed to non - physically demanding performances, and officiating a volleyball match is not inherently a physically demanding task, so affective states may be of more interest. Self - E f ficacy and P erformance Bandura (1986) suggests that self - efficacy contributes to behavior in multiple ways. Self - efficacy beliefs influence how people behave, their emotional reactions to events, and their thought patterns in various situations. People te nd to avoid situations which they do not believe they are capable of succeeding in, and their level of self - efficacy helps determine the degree of effort and persistence they will show when facing failure. Emotional reactions and thoughts are affected by o self - efficacy can focus on tasks at hand and produce more effort in comparison to people with low self - efficacy, who may be anxious and divert attention from possible solut ions. Notably, 14 efficacy judgments only a major determinant of behavior when requisite skills and proper incentives are present; a referee without any experience is unlikely to be successful if thrown into a high - pressure situation, regardless of their beli ef in their capability. Self - efficacy affects people in innumerable domains, from education, to politics, to sport performance. High self - efficacy not only yields better performance, but efficacious individuals are less afraid to set challenging goals and persevere through failure (Feltz et al., 2008). Though complacency can set in during periods of success, a meta - analysis conducted by Moritz , Feltz, Fahrbach, and Mack (2000) showed self - efficacy to have an average correlation of performance of .38 when i ndividuals have incentive to act on their efficacy beliefs, when they possess the requisite skills, when the nature of the task is clear, and when the measure used for obtaining efficacy and performance data is unambiguous. Thus, self - efficacy has a modera tely positive effect on subsequent sport performance. Self - E fficacy in S port O fficials Only recently have researchers begun studying self - efficacy in the sport officiating context. Because the officiating role is distinct from others in sports in its non - competitive nature, its unique pressures, and its tasks which are very different than those of players or coaches, models used to study efficacy beliefs in other populations (such as players and coaches) are not suitable for application to referees. Guille n and Feltz (2011) offered a conceptual framework for referee efficacy, which included referee - specific sources of efficacy information as well as effects or outcomes of efficacy in referees. This model was developed as a result of a focus group of Midwest ern soccer referees with various levels of experience. Participants were asked to identify what they believed to be the key areas of referee efficacy needed to perform their job as an official, the sources of their efficacy, and the influence of those effi cacy beliefs on 15 Based on these discussions, referee self - efficacy was suggested to include six dimensions: game knowledge, decision - making skills, psychological skills, strategi c skills, communication/control of the game, and ph ysical fitness. Game knowledge reflects the sport they officiate. Decision - making skills refers to making critical decisions accurately during competition, Psychological skills need to focus attention and concentration, recover from bad calls, and demonstrate poise. Strategic skills include proper positioning on the area of play, consistent prope r mechanics and signals, and anticipation of game actions. Communication/control of game relates to communication with players, coaches, and other officials, as well as resolving disputes and making necessary adjustments in their behavior and decisions to maintain control of the game. Physical fitness was deemed important in sports where referees engage in a lot of physical exercise, as good fitness is a requirement to stay with the play. Vealey, Hayashi, Garner - Holman, and Giacobbi ( 1998) identified nine sources of sport confidence in athletes: mastery, demonstration of ability, physical and mental preparation, physical self - environmental comfort, and situational favorableness. Sport confidence, as defined by Vealey et fits with self - efficacy because both describe what people perceive they can do, rather than what they have or what they are (Feltz et al., 2008). Using these sources of confidence , combined with sources of self - efficacy , Guillen and Feltz (2011) proposed four major sources of self - efficacy for referees. M astery experience involves years of referee experience, past performance, mentored experience, and knowledge of the rules, 16 and as with the performance accomplishments source of self - theory, was expected to be the strongest source of efficacy. S ignificant others refers to the - support from players, coaches, spectators, peers and partners, and evaluators or administrators, as well as social comparison with other referees. P hysical and mental preparation r elates to goal - setting, arousal regulation, self - talk, visualization of good performance, and readiness for maximum effort. The final source, partner qualifications, and with qualified, able, familiar partners. Myers, Feltz, Guillen, and Dithurbide (2012) developed an instrument to measure and quantify the framework proposed by Guillen and Feltz (2011) . The Referee Self - Efficacy Scale (REFS) was developed based on themes f rom a focus group of soccer referees as well as relevant conceptual and measurement literature. It is composed of 39 items which measure four factors of referee self - efficacy: game knowledge, decision makin g, pressure, and communication. After initial test ing on several hundred soccer referees in the United States and Spain, t wo of the dimensions proposed in Guillen and Feltz (2011), psychological skills and control of the game, were collapsed into the pressure dimension on the REFS. This decision was suppo rted by a single - group exploratory structural equation model, in which a five - factor solution did not significantly improve upon the four factor solution accepted as the final model. Myers et al. showed validity for the REFS across two countries (Un ited St ates and Spain), levels of competition (youth, high school, and elite, which included collegiate, professional, and international referees), team gender, and sports (soccer and basketball) . In addition, the authors showed support for the most of the source s of referee self - efficacy as significant predictors of the four dimensions of referee efficacy . While social support did not predict any of the four 17 dimensions, physical/mental preparation predicted all four, environmental comfort and situational favorabl eness predicted decision making, pressure, and communication, and past accomplishments and vicarious experience predicted game knowledge, decision making, and pressure. To my knowledge, self - efficacy . This represents a gap in the literature which the current study s ought to fill. 18 CHAPTER 3 METHOD Participants The participants in this study were American volleyball referees at the National and Junior National levels. The total number of registered volleyball referees at these levels in the United States is about 700. P articipants for this study were 76 referees: 30 applying for National certification, 46 applying for Junior National certification, which represents about 10% of the targeted populati on. Of the 60 participants who reported their gender, 40 (66.7%) were male, 20 (33.3%) female. 59 (77.6%) of participants described themselves white/Caucasian, 5 (6.6%) chose black/African - American, 5 (6.6%) reported themselves as Hispanic, 2 (2.6%) were A candidates who reported their age ( N =19), ages ranged from 23 to 63, with a mean of 44. Instruments Sports Officials Self - Rating Scale (A modified version of the Sources of Sport Confidence Scale; See Appendix B) ( Vealey, Hayashi, Garner - Holman, & Giacobbi, 1998 ) : The Sports Officials Self - Rating Scale (Guillen & Feltz, 2011) was used to evaluate sources of self - efficacy in the participating referees. Officials a re asked to indicate how important various events are in giving them confidence in officiating their sport. The measure has 25 items on a 7 - point scale. - point scale is as follows : 1=Not at all important, 2=Not very important, 3=Slightly important, 4=Of average importance, 5=Very important, 6=Extremely important, 7=Of highest importance. Each item relates to one of six sources of self - efficacy: social support (e.g., Get positive f eedback from other officials ), physical or mental preparation (e.g., Keep my focus on the game ), environmental 19 comfort (e.g., Officiate in a venue I like ), situational favorableness (e.g., Am familiar with officials I will officiate with ), past accomplishm ents (e.g., Performed well in previous contests ), and vicarious experience (e.g., See successful officiating by other officials in my sport ). One to improve its rele vance to indoor volleyball specifically. Included with the Sports Officials Self - Rating Scale was a short questionnaire that collects demographics and background information about the participant. This information includes age, gender, USAV region, years of experience playing, coaching, and officiating volleyball, highest level playing, coaching, and officiating volleyball, other sports refereed (if any), number of matches officiated in the past year, and number of training sessions or clinics attended in the past year. Referee Self - Efficacy Scale (REFS; See Appendix C): The REFS (Myers, Feltz, Guillen, & Dithurbide, 2012) was used to evaluate self - efficacy in the participating referees. The REFS has 39 items on a 5 - elation to the primary more of four dimensions of referee self - efficacy: game knowledge (e.g., understand the basic strategy of the game ), decision making (e.g., make critical decisions during competition ), pressure (e.g., uninfluenced by pressure from players ), and communication (e.g., communicate effectively with coaches ). Many items are related to more than one dimension: for example, the Table 1. Adjusted items in the Referee Self - Efficacy Scale Item # Original item Adjusted item 2 Know when and how to call more fouls/penalties to control the flow of the game Know when and how to call more or fewer faults/infractions to control the flow of the game 9 Get in proper positions for making decisions Focus on the right area for making decisions 10 Be in the proper angles for decisions Maintain the proper viewing angle for decisions 20 item make critical decisions during competition relates to all four. Three items were adjusted to improve their relevance to volleyball. USAV/PAVO First Referee Rating Sheet (See App endi ces D , E, & F ): The rating sheets were officials use the rating sheets to evaluate the performance of referees who are candidates for National or Junior National bad ges. Each candidate is evaluated three times by multiple raters. The officials are rated on seven dimensions of performance: judgment (e.g., consistency), mechanics/signals (e.g., scanning the court before beckoning for the serve), positioning/focus (e.g ., watching each ball contact), match control (e.g., warm - up administration), communication with match participants (e.g., demeanor and approachability), communication with officiating team (e.g., interactions with line judges, scorekeeper, and second refe ree), and professionalism (e.g., appearance). Raters assign points for each dimension of performance (0 - 15 for each dimension except professionalism, which is 0 - 10), which is then summed for a total score out of a possible 100 points. Raters also assign a recommended level of play - . The reliability and validity of this instrument have not been investigated in previous research. Data Collection Procedures Participants completed a paper survey p acket. Permission to use human subjects for this study was obtained from the Institutional Review Board at Michigan State University. I contacted USAV administrators to garner support and cooperation in conducting this research. Participants completed t he survey measures at the 2014 USAV Girls Championships during a time set aside by tournament administrators. This time was arranged at first 21 roun d of performance evaluations. The lead author w as present to conduct the consent process (See Appendix A), administer the measurement instruments and answer any questi ons the participants may have. Referees were encouraged by USAV to participate, but will not be offered any monetary or material incentive. Participants were given a subject number with their survey packet and instructed to write this number on each of their rating sheets so that their surveys c ould be matched to their performance evaluation s without revealing their identity. The lead researcher captured and entered data from the performance evaluations as the raters turned them in following debriefings with the candidates. Data Analyses Data from the paper surveys were entered into excel a nd double - checked by the researcher and a colleague. Subsequently, all data were loaded into SPSS. Subscale scores for the dimensions and sources of referee efficacy were calculated. Preliminary data analyses, such as multivariate normality, homoscedastici ty, univariate normality, outliers, and multicollinearity were conducted as required to screen data before examination (Kline, 1998). Bootstrapping and square root transformations of the data were each attempted after significant skewness and kurtosis were identified for many variables. Although these procedures were successful in producing normality, the results of subsequent analyses did not substantially differ for the adjusted versus the unmodified data. The data was analyzed in two steps. In the first step, descriptive statistics were calculat ed for the included variables. Pearson correlations were calculated for all sources and dimensions of officiating efficacy, dimensions of performance, and demographic variables to determine the existence of relati onships between key variables. In the second step , multiple regression was used to test the predictive strength of the theorized sources of referee efficacy for dimensions of 22 referee efficacy, as well as the predictive strength of the sources and dimension s of referee efficacy for the measured dimensions of referee performance. Originally, it was intended to test the model using structural equation modeling. Due to the lack of significant relationships identified in prior analyses, this final step was aban doned. 23 CHAPTER 4 RESULTS The results are presented in three sections. In the first section, the preliminary analys e s are presented to evaluate the accuracy and norma lity of the variables. The second section presents descriptive information of the variables and correlations among the variables. The final section presents the results of testing the individual components of the proposed model using multiple regression. Preliminary A nalyses The preliminary analyses were conducted to assess the normal ity and reliability of the variables. To test the assumption of normality of variables, skewness and kurtosis values for each variable were assessed ( see Table 2 ). The assumption of normality can be made if the value of skewness ranges from - 1 to +1, and the value of kurtosis ranges from - 1 to +2 (Huck, 2004). For the Sport Officials Self - rating Scale, the skewness values of each subscale ranged from - .856 (social support) to .06 (environmental comfort), while kurtosis values ranged from - .892 (environme ntal comfort) to .72 (vicarious experience). For the Referee Efficacy Scale subscales, skewness values ranged from - 1.33 (communication) to - 1.02 (pressure), while kurtosis values ranged from .93 (pressure) to 3.35 (communication). For the dimensions of performance on the USAV/PAVO Rating Sheet, sk ewness values ranged from - 1.97 (communicatio n with match officials) to - .28 (positioning/focus), while kurtosis values ranged from - . 62 (positioning/focus) to 9.30 (communication with match officials). Reasona ble assumptions about normality could be established for each source of referee self - efficacy and for the positioning/focus and communication with match participants dimensions of performance. The assumption of normality could not be met for the remaining dimensions of performance or any of the 24 dimensions of referee self - efficacy. Attempts to normalize the data using square root transformations and bootstrapping were both successful, but did not substantially alter the results of subsequent statistical te sts. Therefore, for ease of interpretation, the statistics reported in this manuscript are based on the original non - transformed, non - bootstrapped data. Table 2 . Summary Statistics for Normality and Internal Consistency Variables N Skewness Kurtosis a Judgment 76 - 1.0 4 0.99 0.6 7 Mechanics 76 - 1.3 3 3.6 3 0.74 Positioning/Focus 76 - 0.2 8 - 0.6 2 - 0.1 2 Match Control 76 - 1.0 9 2.29 0.8 4 Communication with match participants 76 - 0.73 0.54 0.4 3 Communication with officials 76 - 1.9 7 9.30 0.29 Professionalism 76 - 1.5 2 4.3 3 0.54 Overall performance 76 - 0.52 0.2 8 0.6 1 Social support 76 - 0.8 6 0. 50 0.87 Preparation 76 - 0.36 - 0.12 0.7 6 Environmental comfort 76 0.0 6 - 0.89 0.8 3 Situational favorableness 76 - 0.22 - 0.2 6 0.7 1 Past experience 76 - 0.80 0.58 0.73 Vicarious experience 76 - 0. 50 0.72 0.9 2 Game knowledge 7 6 - 1.27 2.2 6 0.67 Decision making 7 6 - 1.27 2.6 2 0.79 Pressure 7 6 - 1.0 2 0.9 3 0.8 2 Communication 7 6 - 1.3 3 3.35 0.69 alpha values were calculated to evaluate the internal consistency of each variable ( see Table 2 ). Values for the dimensions and sources of referee self - efficacy ranged from .67 (game knowledge dimension of referee self - efficacy) to .92 (vicarious experie nce as a source of referee self - efficacy). All scale variables met or nearly met Nunnaly (1978) standard of .7 2 to be an acceptable reliability coefficient . F or the dimensions only match control (.84) and mechanics (.74 25 match within the evaluation period, or because s tandards are inconsistent across raters. Descriptive S tatistic s , G ender D ifferences, and D ifferences between C andidate L evels Descriptive statistics for the dimensions of performance, sources and dimensions of referee self - efficacy, and other potential p redictors of performance are presented in Table 3 . Mean evaluation scores were between 13 and 14 for all dimensions of performance except communication with match participants, which was the highest - scoring dimension at 14.25, and professionalism, which was scored on a 10 - point scale rather than 15 - point. Candidates rated social support, preparation, and past experience as the most important sources of confidence in their officiating ability, with mean scores above 5, while environmental comfort was the only source rated below the scale midpoint at 3.9. Mean scores for the four Table 3. Descriptive Statistics for Variables Variables N M SD Minimum Maximum Judgment 76 13.29 0.64 11.20 14.25 Mechanics 76 13.66 0.57 11.20 14.17 Positioning/Focus 76 13.35 0.51 12.00 14.17 Match Control 76 13.27 0.74 10.33 14.60 Communication with match participants 76 14.25 0.38 13.17 15.00 Communication with officials 76 13.84 0.54 11.00 15.00 Professionalism 76 9.54 0.39 7.80 10.00 Overall performance 76 91.30 1.47 86.67 93.83 Social support 76 5.55 0.93 3.20 7.00 Preparation 76 5.77 0.74 3.80 7.00 Environmental comfort 76 3.91 1.47 1.00 7.00 Situational favorableness 76 4.99 0.97 3.00 7.00 Past experience 76 5.47 1.06 2.00 7.00 Vicarious experience 76 4.78 1.16 1.00 7.00 Game knowledge 75 4.50 0.39 3.00 5.00 Decision making 75 4.40 0.49 2.40 5.00 Pressure 73 4.46 0.50 2.80 5.00 Communication 76 4.35 0.43 2.57 5.00 Years officiating 75 14.24 7.35 4.00 35.00 Years coaching 75 13.49 11.35 0.00 38.00 Years playing 75 5.57 7.73 0.00 30.00 Matches officiated last 12 months 75 320.04 219.14 40.00 1500.00 Clinics attended last 12 months 75 4.77 3.50 1.00 20.00 26 dimensions of referee self - efficacy were all between 4 and the s cale maximum of 5, indicating that the candidates were highly confident in all aspects of their ability. More variation was observed in the reported volleyball experience of the candidates. Some candidates indicated playing, coaching, and/or officiating v olleyball for over 30 years, while others reported as few as 4 years officiating and no experience playing or coaching. Match and training experience over the last year varied widely as well, with a range of 40 to 1500 matches worked and 1 to 20 clinics o r training events attended. One - way ANOVA tests did not reveal any significant differences between genders for any of the variables. However, National - than Junior National - mensions of match control ( F (1, 74) = 9.05 , p = .004 ), communication with match participants ( F (1, 74) = 6.48 , p = .013), communication with match officials ( F (1, 74) = 5.14 , p = .026), and professionalism ( F (1, 74) = 11.12 , p = .001), as well as overall ( F (1, 74) = 11.61 , p = .001). Compared to Junior National - level candidates, National - level candidates also reported more years playing volleyball ( F (1, 7 3 ) = 7.63 , p = .007), but fewer matches officiated in the past 12 months ( F (1, 74) = 4.98 , p = .029). C orrelations among the V ariables Pearson correlations were calculated in order to find which sources of referee self - efficacy were related to dimensions of self - efficacy ( see Table 4 ), and which potential predictors of referee performance were related to t he various dimensions of performance upon which the candidates were evaluated (see Table 5 ). Table 4 displays the correlations of the sources of referee self - efficacy with the four dimensions of referee self - efficacy. Only one significant relationship eme rged: preparation was 27 significantly, positively related to game knowledge, r = .24, p = .039. There may also be a modest relationship between preparation and pressure, but it was not statistically significant, r = .23 , p = .055. Individuals reporting high self - efficacy in the game knowledge dimension reported drawing confidence from their preparation. No other significant relationships or strong trends were observed. Table 5 . Pearson Correlation Coefficients between Performance Predictors and Performance Ratings Predictors Ovr Judg Mec Pos MC ComP ComO Prof Social support - .18 .02 - .03 .15 - .3 9 ** - .1 4 - .06 - .01 Preparation - .17 .06 .00 - .06 - .2 9 * .01 - .11 - .13 Environmental comfort - .21 1 - .13 - .00 - .12 - .2 7 * - .02 - .10 .02 Situational favorableness - .02 .06 - .08 .08 - .2 3 1 - .11 .07 .12 Past experience - .15 - .01 - .03 - .07 - .10 - .12 - .11 - .18 Vicarious experience - .16 - .03 - .06 - .09 - .15 - .03 - .19 .04 Game k nowledge .19 - .11 .02 .14 - .04 . 20 .13 - .04 Decision m aking . 20 1 - .07 .03 .09 .04 .19 .15 - .05 Pressure .12 - .05 .04 .10 .01 .10 .18 - .11 Communication .12 - .09 .04 .10 - .04 .12 .11 - .03 Years officiating - .11 - .10 - .1 7 .00 - .07 .10 .11 .10 Years playing .08 - .07 - .15 - .03 .06 .19 - .07 .02 Years coaching .02 - .08 - .2 1 1 .01 - .05 .26 * - .20 1 - .10 Matches last 12 months - .05 .00 - .05 - .00 .03 - .0 6 - .0 7 .01 Clinics last 12 months .12 - .10 .1 3 - .02 .03 .2 1 1 .00 .2 4 * ** p <.01, * p < .05 , 1 p < .10 (2 - tailed) Note: Ovr = Overall, Judg = Judgment, Mec = Mechanics /Signals , Pos = Positioning/Focus, MC = Match Control, ComP = Communication with Match Participants, ComO = Communication with Officiating Team, Prof = Professionalism Table 4 . Pearson Correlation Coefficients between Sou rces and Dimensions of Referee Self - Efficacy Sources of Efficacy Game Knowledge Decision - Making Pressure Communication Social Support - .02 - .09 - .06 - .01 Preparation .24* .19 .23 1 .18 Environmental Comfort - .06 - .04 - .04 - .00 Situational Favorableness - .02 - .07 - .05 .04 Past Experience .09 .11 .15 .15 Vicarious Experience .14 - .02 .08 .12 Years Officiating .12 .14 .11 .15 Years Playing .13 .20 .17 .14 Years Coaching .16 .17 .15 .15 Matches last 12 months .10 .03 - .07 - .03 Clinics last 12 months .00 .08 .05 .01 * p <.05, 1 p = .055 (2 - tailed) 28 Table 5 displays the correlations of the sources of referee self - efficacy, dimensions of referee self - efficacy, and other potential performance predictors with the various dimensions of There were no statistically significant relationships between any predictors and the overall performance score, but two predictors showed statistically insignificant trends: environmental comfort as a source of self - efficacy trended negatively with overall performance score, r = - .21 , p = .066, while the decision - making efficacy dimension trended positively, r = .20 , p = .094. Referees who dre w more confidence from environmental comfort may have performed worse overall, while candidates who reported high levels of self - efficacy for d ecision - making may have performed better. The match control dimension of performance was significantly, negatively correlated with three sources of referee efficacy: social support, r = - .38 , p < .001, preparation, r = - .28 , p = .013, and environmental co mfort, r = - .27 , p = .020. There was also a negative trend between match control and situational favorableness, r = - .23 , p = .051. Candidates who reported draw ing more confidence from these sources tended to score lower on the match control dimension of performance. Years of experience coaching volleyball was significantly, positively correlated to performance in the communication with match participants dimension, r = .26 , p = .024. There were also negative trends for experience coaching with the mech anics dimension of performance, r = - .21 , p = .076, and communication with other officials, r = - .20 , p = .083. Referees with more coaching experience performed better when working with players and coaches, but may have showed weaker officiating mechanics and ability to communicate with their fellow officials. Clinics attended in the past 12 months was significantly, positively related to professional performance, r = .24 , p = .043, and showed a positive trend for communication with match participants, r = .21 , p = .076. Candidates who attended more clinics and training 29 events were graded higher on their professionalism during matches, and may have showed better ability to communicate with players and coaches. No other relationships or strong trends were observed between sources and dimensions of referee self - efficacy, experience and training, and performance ratings. Tests of C omponents of the Proposed M odel The first stage of the model was tested by four univariate multiple regression s using years of experience, matches and clinics/training events in the past year, and the six sources of referee self - efficacy as predictors and the four dimensions of referee self - efficacy as criterion variables. The results of the multiple regression ana lys i s for each REFS subscale are summarized in Table 6 . Table 6 . The Predictability of Referee Self - Efficacy Sources and Experience on Referee Self - Efficacy Dimensions Predictors Game Knowledge Decision - Making Pressure Communication Social Support ( ) - .12 - .17 - .16 - .15 Preparation ( ) .2 7 1 .29 * .38 * .18 Environmental Comfort ( ) - .21 - .15 - .20 - .21 Situational Favorableness ( ) - .06 - .10 - .05 .05 Past Experience ( ) .18 .30 1 .34 * .20 Vicarious Experience ( ) .07 - .15 - .29 1 .02 Years Officiating ( ) .17 .17 .14 .15 Years Playing ( ) .04 .10 .06 .08 Years Coaching ( ) .09 .06 .06 .09 Matches last 12 months ( ) .11 .05 - .06 - .04 Clinics last 12 months ( ) - .02 .09 .11 .03 * p <.05, 1 p < .10 (2 - tailed) The predictors in the regression model s did not significantly predict referee self - efficacy for the dimensions of game knowledge ( F (11, 62) = 1.06 , p = .406), decision - making ( F (11, 62) = 1.2 3, p = .285), pressure ( F (11, 60) = 1.57 , p = .130), or communication ( F (11, 63) = .74 , p = .698). With regard to individual predictors, preparation significantly predicted decision - making ( = .29 , p = .046) and pressure ( = .38 , p = .009), and may predict game knowledge ( = .27 , p = .059). Past experience predicted pressure ( = .34 , p = .039) and may predict decision - making ( 30 = .30 , p = .058), and vicarious experience may predict pressure ( = - .29 , p = .053). No other significant or near - significant beta weights were observed. A simplified regression model including only the self - efficacy sources preparation, past experience, and vicarious experience did significantly predict the self - efficacy dimension pressure, ( F (3, 69) = 3.60 , p = .018), accounting for 9.8% of variance based on adjusted R 2 . Thus, the first stage of the propose d model was not well - supported by the data. While one dimension of referee self - efficacy may be weakly associated with a handful of predictors, by and large the proposed predictors (experience and sources of referee self - efficacy) were not observed to rel ate to the four dimensions of referee self - efficacy. The second part of the model was tested by eight univariate multiple regressions using the dimensions of referee self - efficacy and the experience variables (years of experience, matches in the last 12 m onths, and clinics in the last 12 months) as predictors and the performance dimensions and overall performance score as criterion variables. The results are displayed in Table 7 . Table 7 . The Predictability of Referee Self - E fficacy Dimensions and Experience on Performance Predictors Ovr Judg Mec Pos MC ComP ComO Prof Game k nowledge .07 - .21 - .10 .18 - .33 .10 .13 - .06 Decision m aking .62 .12 .12 - .22 .56 .59 - .04 .29 Pressure - .40 - .04 - .04 .21 - .32 - .50 1 .24 - .36 Communication - .08 .06 .13 .01 .08 - .06 - .12 .03 Years officiating - .17 - .11 - .14 .01 - .1 2 .01 .13 .14 Years playing .05 - .01 - .09 - .11 .08 .08 - .08 .00 Years coaching - .04 - .04 - .20 .04 - .08 .17 - .22 - .14 Matches last 12 months - .16 .00 - .06 .03 - .06 - .2 2 1 - .02 - .02 Clinics last 12 months .11 - .11 .15 .00 . 01 .20 .05 .27 * * p < .05 , 1 p < .10 (2 - tailed) Note: Ovr = Overall, Judg = Judgment, Mec = Mechanics/Signals, Pos = Positioning/Focus, MC = Match Control, ComP = Communication with Match Participants, ComO = Communication with Officiating Team, Prof = Professionalism 31 The predictors in the regression model did not significantly predict referee performance overall ( F (9, 62) = .99 , p = .458) or for the dimensions judgment ( F (9, 62) = .25 , p = .984), mechanics ( F (9, 62) = .79 , p = .627), positioning/focus ( F (9, 62) = .28 , p = .978), match control ( F (9, 62) = .38 , p = .937) communication with match participants ( F ( 9 , 62) = 1.85 , p = . 077 ), communication with other officials ( F ( 9 , 62) = .82 , p = . 600 ), or professionalism ( F ( 9 , 6 2 ) = .90 , p = . 531). Because communication with match participants was near significance, a simplified regression model including only the game knowledge, decision - making, and pressure dimensions of referee self - efficac y , with communication with match participants as the criterion variable, was attempted. This trimmed model was also insignificant, ( F (3, 69) = 1.72 , p = .171). Only three predictors had standardized coefficients for individual dimensions of performance t hat were significant or approached significance: pressure, for communication with match participants ( = - .50 , p = .090); matches in the last 12 months, also for communication with match participants ( = - .2 2 , p = .074); and clinics/training events atten ded in the last twelve months, for professionalism ( = .27 , p = .039). Due to the lack of functional connections between and within stages of the model, no further modeling techniques or tests were conducted. 32 CHAPTER 5 DISCUSSION Despite the importance participants and other actors in competitive contexts, little is known about the root causes of that performance. Past research has identified experience as a primary predictor of performance in r eferees, but until recently no psychological factors had been scientifically associated with performance. Guillen and Feltz (2011) suggested that self - efficacy might be a psychological construct of interest for referees, and Myers et al. (2012) provided a n instrument for measuring self - efficacy specifically in sport officials. This research is the first attempt at evaluating the relationship between self - efficacy and performance in sport officials. The main purpose of this study was to examine the predict ive strength of referee self - efficacy on officiating performance in volleyball referees. None of the dimensions of performance were significantly predicted by regression models including the four dimensions of referee self - efficacy (game knowledge, decisi on - making, pressure, and communication) and the experience variables collected in the study (years of experience officiating, coaching, and playing, matches officiated in the last 12 months, and clinics or training events attended in the last 12 months). There was a significant relationship between clinics attended and the professionalism dimension of performance. This is a rather curious result, as the impression used to get the overall score to a point where the rater was happy after they could not find points to deduct in other categories. One explanation may be that o fficials who are more professional may invest more time and money into attending , and this prof essional approach may come out in their demeanor and performance in matches. As raters often scored candidates after debriefing 33 them and providing feedback, the professionalism score may also have been influenced by the evaluations. F i nally, this result may be the result of bias on w ith, or seek to reward candidates who may have shown more investment in their own development. However, the rating team generally seemed to avoid assigning raters to evaluate individuals from their own home region, so this explanation may not apply in mos t cases. There were reasonably strong trends of association between the pressure dimension of self - efficacy and matches officiated, and the communication with match participants dimension of performance. Communication skills developing with experience is no surprise, but at first glance the relationship between pressure and communication might seem unintuitive: why would - efficacy for pressure, rather than communication, influence their ability to communicate with coaches and players , and wh y is that relationship negative ? T he USAV rating sheet lists the following items of evaluation for communication with match participants: 1. Respectful, dignified manner 2. Demeanor, approachability 3. Communication with team members 4. Acknowledgement of coaches It may be that players and coaches are the primary source of pressure on referees in this ability to communicate outward to the match participants. The negative direction of the relationship may reflect a 34 of their behavior; perhaps some referees who think they are skilled in handling pressure do so by displaying the attributes and behaviors for which the raters are instructed to look. A secondary purpose of t his research was to investigate the connections between the sources and the dimensions of referee self - efficacy. Two sources significantly predicted one or more dimensions: preparation was related to decision - making and pressure efficacy, while past experi ence was only significantly related to pressure efficacy. Predictive relationships may also exist between preparation and game knowledge efficacy, past experience and decision - making efficacy, and vicarious experience and pressure efficacy, but these were not statistically significant. It should not be surprising that, in data light on strong associations, preparation and past experience emerged as the only notable predictors: in developing the Sources of Sport Confidence Questionnaire, Vealey et al. (199 8) identified preparation as the most salient source of self - confidence in athletes, and preparation and experience are consistently forwarded as two of the most important efficacy sources (e.g. Bandura, 1977, 1986). In the Myers et al. (2012) study, prepa ration predicted all four dimensions of referee efficacy, environmental comfort and situational favorableness predicted decision - making, pressure, and communication efficacy, and past accomplishments and vicarious experience predicted game knowledge, decis ion - making, and pressure efficacy. The results were similar in that social support was not found to significantly predict any dimension. Myers and colleagues found that years of experience and highest level of experience also predicted all four dimensions of referee self - efficacy, but there was no such relationship for the experience measures in the current study. 35 One explanation for the lack of significant results from this study is the absence of diversity in the sample. The candidates were of similar sk ill levels hence their congruent candidacies and the vast majority of them passed their evaluations. My impression was that the regions actively winnow out local referees who they do not feel are ready to move up yet, forwarding to the evaluation process o nly those whom they believe to be skilled and experienced enough to advance. The spread of performance evaluation scores was relatively narrow, overall performance scores low enough to fail the evaluation were by some measures statistical outliers. Additionally, the candidates as a group were very confident in their abilities, perhaps as a result of the aforementioned winnowing process. Scores on the REFS dimensions averaged well over 4 on a 1 - 5 scale, with standard deviations of less than 0.5. Including a more diverse range of ability levels might have generated more significant results. Efficacy scores in the Myers et al. (2012) study were similarly high, but that study made no attempt to use them to discriminate between outcomes like performance. Problems with the outcome measure might also explain the lack of significant findings. The USAV rating sheets were used as the outcome measure in this study because they provided an opportunity to collect a standardized performance score for a relatively large number of referees. Introduction of a more scientifically validated measure would have been optimal, but impractical. I did not anticipate the degree to which the ev aluation process would be compromised. While all raters used the rubric and made comments relating to individual dimensions of performance, some did not assign category scores, choosing instead to simply assign an overall score out of 100. As mentioned bef ore, some that did give category scores used 36 professionalism as something of a dummy score, docking a point if they thought the overall score too high after summing the other dimensions. Raters would frequently discuss their observations with the next rater scheduled to evaluate a of the rating sheet as a valid metric for performance. This was symptomatic of what I view as a larger problem: the evaluation process was intertwined with training of the candidates in ways which damaged the usefulness of the evaluation. Candidates were debriefed by their raters following each t wo - match round of evaluations, and then in subsequent matches judged at least in part on their efforts to address issues brought up during the debriefing, rather than their performance in that match alone. This phenomenon was most evident in nightly meetin gs of the rating team, in which the evaluators would discuss each candidate one by one, noting their current standing with regard to passing or failing, and warning fellow raters what to look out for the next day. Because opportunities for training of the candidates are limited, it may be outcome variables, which were intend ed to represent objective performance, instead to some Finally, the context in which the referees were evaluated must be considered. The tournament was staged in a convention center with several dozen courts in a single exhibition hall. The event space was cramped, with parents and spectators very close to the playing area, and with the number of whistles and shouting athletes it was extremely loud. Referees worked long days in this environment, and eve ning social events and unfamiliar sleeping arrangements 37 may have limited their quality rest. This was a stressful environment for the candidates, compounded by their knowledge that they were being evaluated for professional advancement. With that in mind, - efficacy for coping with pressure was one of the more salient (if statistically insignificant) predictors of some performance dimensions. yball referees at this level while referees often work high school or college contests , they also frequently call matches at large - scale events like this one. In that case, perhaps the evaluation context was appropriate to the demands of the avocation. The working situation for the evaluation team should also be considered: the raters worked for even longer hours than the candidates, and it is unclear whether their scoring may have changed over the course of several days of evaluations. The stress and distr actions of the event space may impact the raters as well, and all of the scores are confounded by variation in level of play and difficulty of officiating unique events from match to match . Limitations and Future Directions The primary limitation of this study was its small number of participants. Myers et al. (2012) suggested N =250 to use most parameters of the REFS, and N =400 to use all parameters. Seventy - six referees participated in the study, representing nearly all of the candidates for the year. A s imilar study conducted with officials from another sport might incorporate a larger measure for performance might have been a strength; unfortunately, the compromi sed objectivity of that measure may mean that the lack of other measures to interpret is instead a limitation. Replication with a larger group might necessitate incorporation of multiple performance measures or, more ambitiously, development of a more broa dly applicable instrument. Another 38 limitation of this study was the homogeneity of participants with regard to skill and confidence levels. Future research might involve officials at a wider variety of skill and experience levels. As this project was obs ervational in nature, causal inference regarding the relationship between self - efficacy and performance cannot be drawn from the results. Future experimental - efficacy in a laboratory - reproduci ble task, such as calling balls and strikes, or making offside decisions based on video from a soccer match. Tasks of this nature have been shown to discriminate between referee skill levels or specializations in the past, but never used in conjunction wit h manipulation of psychological constructs like self - efficacy. Conclusion No relationship was found in this thesis between self - efficacy and performance in high - level v olleyball officials. There also ex perience and performance in their evaluations, and few connections between previously identified sources of referee confidence and dimensions of referee self - efficacy. These null findings may be due to lack of variance in ability and confidence on the part of the referees, or produced by an evaluation system which is designed to teach candidates, and pass most of them, rather than explicitly evaluate their performance. The results of this study do replicate those of the Myers et al. (2012) study in identify ing preparation and past experience as the most important sources of referee efficacy. Future research might incorporate a more objective evaluation system, a wider variety of referee skill and experience levels, officials from different sports, and/or exp erimental manipulation of self - efficacy during simulated officiating tasks in a laboratory setting. 39 APPENDICES 40 Appendix A: Consent Form Volleyball Officials Research Study Participant Consent Form You are being asked to participate supervision of Deborah Feltz, Ph.D. from Michigan State University. This study is to investigate the contribution of self - confidence to match performance in volleyball official s. You have been identified as a potential participant in this study because you are a volleyball referee. You must be at least 18 years old to participate in this research. Procedure: As part of this research, you will be asked to complete a short packet of surveys designed to provide background information about yourself as an official, your self - confidence for the officiating task, and your sources of self - confidence. It should take approximately 10 minutes to complete. Additionally, copies of y our completed rating sheets will be taken as a performance measure. Benefits: We believe that the study results will have practical applications for volleyball officials to improve their performance. Additionally, the information gained from this study wi ll increase our understanding of different aspects of sports officiating. Risks: There are no known physical, legal, or economic risks associated with this study. None of the questions address sensitive issues regarding personal beliefs, behaviors, experi ences or attitudes. Voluntary Participation: Participation in this research project is completely voluntary. You have the right to say no. You may change your mind at any time and withdraw. You may choose not to answer specific questions or to stop partic ipating at any time. Whether you choose to participate or not will have no affect on your evaluation. Confidentiality: Your participation in this study will remain confidential. The principal investigator, secondary investigator, and the IRB will have acce ss to the research data. It will be kept in a locked file cabinet and on a password protected computer. All collected data will be de - identified and analyzed at the group level to ensure the confidentiality of individual responses. Your confidentiality wil l be protected to the maximum extent allowable by law. Contact and Questions: If you have concerns or questions about this study, such as scientific issues, how to do any part of it, or to report an injury, please contact the researchers: Ben Spencer ( spenc291@msu.edu ), (402) 429 - 1500; or Deborah L. Feltz, Ph.D. ( dfeltz@msu.edu ), (517) 355 - 4732, or by regular mail at: Michigan State University, 134 IM Circle, East Lansing, MI 48824. If you have questions or concerns about your role and rights as a research participant, would like to obtain information or offer input, or would like to register a complaint about this study, you may contact, anonymously if you wish, the Michigan State Un - 355 - 2180, Fax 517 - 432 - 4503, or e - mail irb@msu.edu or regular mail at 207 Olds Hall, MSU, East Lansing, MI 48824. Statement of Assent/Cons ent: Your signature below means that you voluntarily agree to participate in this research study and release copies of your rating sheets to be used in this research. ________________________________________ _____________________________ Signature Date 41 Appendix B: Demographics and Sport Officia l Self - Rating Scale Sport Officials Self - Rating Scale Think back to times when you felt very confident when officiating in your sport. What things made you feel confident? What things helped you believe in your abilities and gave you confidence that you would perform successfully? Participant ID #: ____________________ Race: White/Caucasian Age : Gender: Female Male USAV Region: ____________________ American Indian/Alaskan Native Other Is volleyball the primary sport you officiate? : Yes No Do you referee other sports? : Yes No List: _________________________________ Years experience officiating volleyball : ___ playing volleyball: ___ coaching volleyball: ___ Highest level as referee: ___ as player: ___ as coach: ___ Approximately how many matc hes have you officiated in the past 12 months? ___ In how many clinics or training events have you participated in the last 12 months? ___ Listed below are some things that may help officials feel confident in performance situations. For each statement, check the number that indicates HOW IMPORTANT THAT IS IN HELPING YOU FEEL CONFIDENT IN OFFICIATING. Please respond to every question even though they may seem repetitive. There are no right or wrong answers because every official is different. Please be honest your answers will be kept completely confidential. I gain confidence in officiating when I... Not at all important 1 Not very important 2 Slightly important 3 Of average importance 4 Very important 5 Extremely important 6 Of highest importance 7 1. Get positive feedback from other officials 1 2 3 4 5 6 7 2. Keep my focus on the game 1 2 3 4 5 6 7 3. Officiate in a venue I like 1 2 3 4 5 6 7 4. Am familiar with officials I will officiate with 1 2 3 4 5 6 7 5. Performed well in previous contests 1 2 3 4 5 6 7 6. Know I have support from other officials in my sport 1 2 3 4 5 6 7 7. See successful officiating by other officials in my sport 1 2 3 4 5 6 7 8. 1 2 3 4 5 6 7 9. Watch another official I admire perform successfully 1 2 3 4 5 6 7 10. Am assigned a match/game I feel qualified for 1 2 3 4 5 6 7 11. 1 2 3 4 5 6 7 12. Made good decisions in previous contests 1 2 3 4 5 6 7 42 13. Am encouraged by other officials 1 2 3 4 5 6 7 14. Watch another official perform well 1 2 3 4 5 6 7 15. Venue conditions are favourable 1 2 3 4 5 6 7 16. Prepare myself physically and mentally for a contest 1 2 3 4 5 6 7 17. Like the venue where I am officiating 1 2 3 4 5 6 7 18. Have performed well in difficult contests 1 2 3 4 5 6 7 19. Get positive feedback from evaluators of my officiating 1 2 3 4 5 6 7 20. Watch well - officiated contests 1 2 3 4 5 6 7 21. Believe in my ability to give maximum concentration in a contest 1 2 3 4 5 6 7 22. Receive support and encouragement from other officials 1 2 3 4 5 6 7 23. Watch officials who are at my level perform well 1 2 3 4 5 6 7 24. Am assigned to officiate with a qualified partner 1 2 3 4 5 6 7 25. Am in good physical condition 1 2 3 4 5 6 7 43 Appendix C: Refficacy Questionnaire REFFICACY QUESTIONNAIRE Referee confidence refers to the extent to which referees believe that they have the capacity to perform successfully in their job. Think about how self - confident you are when you officiating. Truthfully respond to the questions below based on how confident you feel about officiating. There are no correct answers. Please be honest your answers will be kept completely confidential. Circle the number which corresponds to your feelings of self - confidence. In the context of performing your re feree job, how confident are you in your ability to Low Medium High 1. Understand the rules of your sport 1 2 3 4 5 2. Know when and how to call more or fewer faults/infractions to control the flow of the game 1 2 3 4 5 3. Demonstrate poise under pressure 1 2 3 4 5 4. Communicate effectively with coaches 1 2 3 4 5 5. Stay up with the play 1 2 3 4 5 6. Think and respond successfully during competition 1 2 3 4 5 7. Resolve disputes 1 2 3 4 5 8. Apply the rules accurately 1 2 3 4 5 9. Focus on the right area for making decisions 1 2 3 4 5 10. Maintain the proper viewing angle for decisions 1 2 3 4 5 11. Make critical decisions during match (game/competition) 1 2 3 4 5 12. Be in control of the game 1 2 3 4 5 13. Be successful as a referee at your current level 1 2 3 4 5 14. Concentrate well enough to be successful 1 2 3 4 5 15. Communicate effectively with partners 1 2 3 4 5 16. 1 2 3 4 5 17. Consistently be successful in making correct decisions 1 2 3 4 5 18. Uninfluenced by pressure from players 1 2 3 4 5 19. Handle unexpected situations 1 2 3 4 5 20. Demonstrate effective teamwork with partners 1 2 3 4 5 21. Recognize your own mistakes 1 2 3 4 5 22. Uninfluenced by pressure from spectators 1 2 3 4 5 23. Adapt to different game situations and still be successful 1 2 3 4 5 24. Achieve your professional goals as a referee 1 2 3 4 5 25. Know and understand the basic strategy of the game 1 2 3 4 5 26. Communicate effectively with players 1 2 3 4 5 27. Be in good physical condition 1 2 3 4 5 28. Handle challenges about decisions appropriately 1 2 3 4 5 29. Demonstrate decisiveness 1 2 3 4 5 30. Anticipate game situations 1 2 3 4 5 31. Communicate effectively with auxiliary game personnel ( e.g., video reviewer, scorekeepers, timekeepers, goal judges, etc ) 1 2 3 4 5 44 32. Be firm in your decisions 1 2 3 4 5 33. Know and understand proper officiating mechanics 1 2 3 4 5 34. Know all the rules of your sport 1 2 3 4 5 35. Make quick decisions 1 2 3 4 5 36. Not let a bad call affect your next call 1 2 3 4 5 37. Demonstrate accurate judgement 1 2 3 4 5 38. Be successful even when the crowd is against you 1 2 3 4 5 39. Uninfluenced by pressure from coaches 1 2 3 4 5 45 Appendix D: USAV/PAVO Referee Rating Sheets Figure 3: USAV/PAVO First Referee Rating Sheet 46 47 Figure 4: USAV/PAVO Second Referee Rating Sheet 48 49 Appendix E: Practical Rating Sheet Instructions Figure 5: Practical Rating Sheet Instructions 50 Appendix F: PAVO/USAV Rating Criteria Figure 6 : PAVO/USAV First Referee Rating Criteria 51 Figure 7 : PAVO/USAV Second Referee Rating Criteria 52 REFERENCES 53 REFERENCES Anderson, K. J., & Pierce, D. A. (2009). Officiating bias: The effect of foul differential on foul calls in NCAA basketball. Journal of Sports Sciences , 27 (7), 687 694. Anshel, M. H. (1995). Development of a r ating s cale for d etermining c ompetence in b asketball r eferees: Implications for s port p sychology. The Sport Psychologist , 9 , 4 28. Band ura, A. (1977). Self - efficacy: T oward a unifying theory of behavioral change. Psychological Review , 84 (2), 191. Bandura, A . (1982). Self - efficacy in human agency. American Psychology, 37 , 122 - 147. Bandura, A. (1986) . Social foundations of thought and action: A social cognitive theory . Englewood Cliffs, New Jersey: Prentice - Hall, Inc. Bandura, A. (1997). Self - efficacy: The exercise of control . New York: Freeman. Bandura, A. (2001). Social - cognitive theory: An agentic perspective. Annual Review of Psychology, 52 , 1 - 26. Boyko, R. H., Bo yko, A. R., & Boyko, M. G. (2007). Referee bias contributes to home advantage in English Premiership football. Journal of Sports Sciences , 25 (11), 1185 1194. Catteeuw, P., Helsen, W., Gilis, B., & Wagemans, J. (2009). Decision - making skills, role specifi city, and deliberate practice in association football refereeing. Journal of Sports Sciences , 27 (11), 1125 1136. Dosseville, F., Laborde, S., Raab, M., & others. (2011). Contextual and personal motor Sport Psychologist , 25 (1), 67. Dowrick, P.W., & Dove, C. (1980). The use of modeling to improve the swimming performance of spina bifida children. Journal of Applied Behavior Analysis, 13, 51 - 56. Faul, F., Erdfelder, E., Buchner, A., & Lang, A. - G. (2009). Stat istical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods , 41 , 1149 - 1160. Feltz, D.L., Short, S.E., & Sullivan, P.J. (2008). Self - efficacy in sport . Champaign, IL: Human Kinetics. Guillén, F., & Feltz, D. L. (2011). A c onceptual m odel of r eferee e fficacy. Frontiers in Psychology , 2 , 25 . 54 Huck, S. (2004). Reading statistics and research (4 th ed.) . Upper Saddle River, NJ: Ally & Bacon. Kline, R. B. (1998). Principles and practice of structural equa tion modeling . New York: Guilford Press. Lehman, D. R., & Reifman, A. (2001). Spectator i nfluence on b asketball o fficiating. The Journal of Social Psychology , 127 (6), 673 675. Lopez, M. J., & Snyder, K. (2013). Biased impartiality among N ational H ockey L eague referees. International Journal of Sport Finance , 8 , 208 223. MacMahon, C., Helsen, W. F., Starkes, J. L., & Weston, M. (2007). Decision - making skills and deliberate practice in elite association football referees. Journal of Sports Sciences , 25 (1), 65 78. Maddux, J.E. (1995). Self - efficacy theory: An introduction. In J.E. Maddux (Ed.), Self - efficacy, adaptation, and adjustment: Theory, research, and application . New York: Plenum Press. McInman, A. D. (1997). Where are all the sport psychology umpi re studies? Presented at the 32nd Annual Conference of the Australian Psychological Society, Cairns, Australia. Moritz, S.E., Feltz, D.L., Fahrbach, K.R., & Mack, D.E. (2000). The relation of self - efficacy measures to sport performance: A meta - analytical review. Research Quarterly for Exercise and Sport, 71, 280 - 294. Myers, N. D., Feltz, D. L., Guillén, F., Dithurbide, L., & others. (2012). Development of, and i nitial v alidity e vidence for, the r eferee s elf - e fficacy s cale: A m ultistudy r eport. Journal of Sport and Exercise Psychology , 34 (6), 737. Nunnaly, J. (1978). Psychometric theory. New York: McGraw - Hill. Pizzera, A., & Raab, M. (2012a). Does m otor or v isual e xperience e nhance the d etection of d eceptive m ovements in f ootball? International Journal of Sports Science and Coaching , 7 (2), 269 284. Pizzera, A., & Raab, M. (2012b). Perceptual j udgments of s ports o fficials are i nfluenced by their m otor and v isual e xperience. Journal of Applied Sport Psychology , 24 (1), 59 72. Plessner, H., & Betsch, T. (200 1). Sequential e ffects in i mportant r eferee d ecisions: The c ase of p enalties in s occer. Journal of Sport and Exercise Psychology , 23 , 254 259. Souchon, N., Cabagno, G., Rascle, O., Traclet, A., Dosseville, F., & Maio, G. R. (2009). o f player gender a t t he highest national level. Psychology of Women Quarterly , 33 (4), 445 452. 55 Souchon, N., Cabagno, G., Traclet, A., Dosseville, F., Livingstone, A., Jones, M., & Maio, G. R. - making and player gender: the moderating role of the type of situation. Journal of Applied Sport Psychology , 22 (1), 1 16. heuristics: The moderating impact of standard of competition. Journal of Sports Sciences , 27 (7), 695 700. Souchon, N., Coulomb - Cabagno, G., Traclet, A., & Rascle, O. (2004). making in handball and transgressive behaviors: Influence of stereotypes about gender of players? Sex Roles , 51 (7 - 8), 445 453. Trudel, P., Cote, J., & Sylvestre, F. (1996). Systematic Observation of Ice Hockey Referees During Games. Jo urnal of Sport Behavior , 19 (1), 66 81. Unkelbach, C., Memmert, D., & others. (2008). Game management, context effects, and calibration: The case of yellow cards in soccer. Journal of Sport and Exercise Psychology , 30 (1), 95. v an Quaquebeke, N., & Giessne r, S. R. (2010). How embodied cognitions affect judgments: Height - related attributi on bias in football foul calls. Journal of Sport & Exercise Psychology , 32 , 3 22 . Vealey, R. S., Hayashi, S. W., Garner - Holman, M., & Giacobbi, P. (1998). Sources of sport - confidence: conceptualization and instrument development. Journal of Sport & Exercise Psychology , 20 , 54 80. Wagner - Egger, P., Gygax, P., & Ribordy, F. (2012). Racism i n soccer? Perception o f challenges o f black a nd white players b y white referees, soccer players, a nd fans. Perceptual and Motor Skills , 114 (1), 275 289. Weinberg, R., Gould, D., & Jackson, A. (1979). Expectations and performance: An empirical test - efficacy theory. Journal of Sport Psychology, 3, 345 - 354.