llllHHllHlllliIIIIIIUHIHIJIIIIJHIIHIIHHIJIIHHIHII 293 00794 9757 This is to certify that the thesis entitled Relationship between competitive trait anxiety and sport performance: A meta-analysis Date 0-7639 presented by Daniel F. Wagman has been accepted towards fulfillment of the requirements for Masters degree in Sport Psych. Major professor S /J\“(/ C; ). MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE “£3 3}? my; MSU is An Affirmative Action/Equal Opportunity Institution czbimm rims-p. 1 RELATIONSHIP BETWEEN COMPETITIVE TRAIT ANXIETY AND SPORT PERFORMANCE: A META-ANALYSIS BY Daniel F. Wagman A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Physical Education and Exercise Science 1992 ABSTRACT Relationship Between Competitive Trait Anxiety and Sport Performance: A Meta-Analysis By Daniel Frederick Wagman The purpose of this meta-analysis was to investigate the overall relationship between competitive trait anxiety, as measured by SCAT, and sport performance. In addition, various variables believed to possibly modify the competitive trait anxiety/sport performance relationship (i .e. , gender, ability, age, and type of sport) were examined. Twelve studies, yielding l4 Pearson correlation coefficients, were identified as meeting the requirements necessary for this synthesis of studies. Various statistical procedures as outlined under meta- analytic synthesis were followed in order to normalize the data, weight the data, investigate variability of the correlation coefficients under' review, anui determine homogeneity. Results indicated that all study features investigated were highly variable. This variability was maintained when various study feature combinations were considered. Thus, it is concluded that the competitive trait anxiety/sport performance relationship was not moderated by any of the features investigated in this study. I would like to dedicate this research to my mother who through her faith in me allowed me to realize that any goal can indeed be achieved. ACKNOWLEDGMENTS I would like to thank Dr. Deb Feltz for her guidance, Dr. Marty Ewing for her challenging ideas, and especially Dr. Cathy Lirgg for helping in the statistical analysis of this thesis. iv TABLE OF CONTENTS List of Tables ........................................ vii List of Figures ...................................... viii Chapter I. INTRODUCTION ...................................... 1 Statement of the Problem ........................ 9 Questions that Prompted this Study ............ 10 Delimitations .................................. 10 Definitions .................................... 11 Limitations .................................... 12 II. A REVIEW OF THE LITERATURE ...................... 13 Introduction ................................... l3 Defining Arousal and Anxiety ................... 14 Arousal Theories ............................... 15 Drive Theory ................................ 15 The Inverted-U Hypothesis .................. .16 A Cognitive Model ........................... 17 The Optimum Level Theory .................... 18 Summary ..................................... 20 Competitive Anxiety ............................ 21 The Competitive Anxiety Model ............... 24 Development of SCAT ............................ 33 Reliability and Validity Issues ............. 34 A Review of Meta-Analysis ...................... 38 Introduction ................................ 38 Stages of Meta—Analysis ..................... 41 Problem formation ........................ 41 Data collection .......................... 42 Data evaluation .......................... 42 Data analysis and interpretation ......... 42 Public presentation of results ........... 42 Measuring Study Findings .................... 43 Limitations, Criticisms ..................... 44 Summary ........................................ 47 III. METHOD ......................................... 50 Collection of Studies .......................... 50 Coding of Study Features ....................... 51 Subject Variables ........................... 51 Task Variables .............................. 53 Treatment of the Data ................ . ......... 54 Chapter IV. RESULTS ......................................... 56 Tests of Homogeneity ............................ 57 Overall Test of Homogeneity .................. 57 The Homogeneity Test of Study Features ....... 59 Gender .................................... 59 Ability ................................... 61 Age ....................................... 61 Type of Sport ............................. 63 V. DISCUSSION AND SUMMARY ........................... 67 Discussion ....................................... 67 Summary .......................................... 77 APPENDICES .......................................... 80 Appendix A: The Sport Competition Anxiety Test, Forms A and C .............................. 80 Appendix B: ANOVA-like comparisons for study feature combinations ............................. 82 Appendix C: Studies not included in this meta-analysis .................................... 87 REFERENCES .......................................... 96 vi LIST OF TABLES Table Page 1. Features of studies ................................ 52 2. Summary of characteristics and correlations for competitive trait anxiety/sport performance studies ............................................ 58 3. Summary of analyses for gender ..................... 6O 4. Summary of analyses for ability .................... 62 5. Summary of analyses for age ........................ 64 6. Summary of analyses for skill ...................... 65 7. Summary of analyses for gender by age .............. 82 8. Summary of analyses for gender by ability .......... 83 9. Summary of analyses for age by ability ............. 84 10. Summary of analyses for skill by age ............... 85 11. Summary of analyses for skill by gender ............ 86 Vii LIST OF FIGURES Figure Page 1. The competitive process ............................. 26 2. Competitive A-trait as a mediator between competitive stimulus and response ................... 28 3. Model of competitive anxiety. ....................... 28 viii CHAPTER I Introduction Sport scientists agree that participation in athletic events is mediated by psychological and physical effects (Martens, Vealey, & Burton, 1990). One psychological effect that has been studied.extensively in sport is anxiety; Anxiety has been conceptualized as having state and trait components (Spielberger, 1966). Specifically, trait anxiety has been defined as perceiving certain situations as threatening and responding with varying levels of state anxiety; whereas state anxiety is characterized by an existing or immediate emotional state of apprehension and/or tension (Spielberger, 1966). In any given situation an individual’s anxiety will beidetermined by an interaction of general levels of anxiety relevant to the situation (trait anxiety) and.specific situational constraints (state anxiety) (Spielberger, 1966). Martens (1977) developed a conceptual model specifically for studying anxiety in sport competition. Embedded in this model are the concepts of competitive trait and state anxiety. Competitive trait anxiety is described as an individual’s tendency to perceive competitive situations as threatening and to respond to these situations with state anxiety. According to Martens, both competitive trait and.state anxiety influence the quality of performance. Martens further'hypothesized.that l 2 high competitive trait anxiety led to higher state anxiety within a competitive situation. Because competitive trait anxiety has been shown to be a relatively stable intrapersonal factor and a mediator of state anxiety (Martens, 1977; Martens et al., 1990), the assessment of competitive trait anxiety became very important. Thus, Martens (1977) develOped the Sport Competition Anxiety Test (SCAT) for the purpose of providing a reliable and valid measure of the amount of anxiety an athlete generally feels about competing in sport. Since the conceptualization of competitive anxiety and the development of SCAT, there have been numerous studies confirming the reliability, internal consiStency, and validity of the instrument (Martens, 1977; Martens et al., 1990). Additionally, SCAT has been shown to be predictive of performance as well as cognitive, somatic, and competitive state anxiety (Martens, Burton, Rivkin, & Simon, 1980; Martens, Burton, Vealey, Bump, & Smith, 1982). Although SCAT has withstood investigations into its reliability, construct validity, and internal consistency rather well, some research has provided ambivalent findings in SCAT's relationship to sport and motor performance as mediated by gender, age, and athletic ability (Martens et al., 1990). For instance, some researchers have found a direct relationship between SCAT and performance in such tasks as overanm throwing (Weinberg, 1977), basketball (Sonstroem & 3 Bernardo, 1982), golf (Weinberg & Genuchi, 1980), and team batting average (Gerson & Deshaies, 1978) . However, nonsignificant findings were reported in studies that examined the relationship'between.SCAfl'and performance in such tasks as a computerized dice game (Broughton & Perlstrom, 1986), golf putting (Murphy &.Woolfolk, 1987), a flexed—arnihang-strength task (Poteet & Weinberg, 1980), rugby (Maynard & Howe, 1987), and marathon running (McKelvie, Valliant, & Asu, 1985). These ambivalent findings do not necessarily invalidate the construct validity of SCAT; they may be due to task, individual differences, or methodological problems (Martens et al., 1990). Because some sports yield significant findings and other sports do not, possible.mediating effects of the type of sport on the competitive trait anxiety/sport performance relationship warrant further investigation. In an attempt to clarify whether different sports affect the competitive trait anxiety/sport performance relationship disparately, the present investigation categorized the various sports into those requiring open skills (e.g., sports that are externally paced, are distinguishable by varying environmental conditions, and are characterized by flexibility in movement responses) and those requiring closed skills (e.g., sports that are self paced, where the environment remains the same, and consistency in movement is required)(Sage, 1984){ The rationale for dividing sports according to open and 4 closed skill requirements is based upon arousal theory (Oxendine, 1970). Neiss (1988) believed that anxiety is related to arousal because it is a broad construct equivalent to states of fear, anger, surprise, joy, and sadness. Arousal theory further stipulates that higher levels of arousal will provide energy and positively affect the athlete involved in sports such as football, and basketball (Oxendine, 1970). However, lower levels of arousal have been shown to be more beneficial in such sports as fencing, archery, and.golf (Fenz, 1972; Iso-Ahola & Hatfield, 1986; Kauss, 1980; Nideffer, 1976; Oxendine, 1970): Because open skill sports are characterized by changing environmental conditions and direct responses to an opponent's actions, a higher level of arousal is needed for optimal performance toloccur as opposed to sports that require self—paced action under unchanging environmental conditions. This categorization of sports works well within the confines of the present review, and a division of sports according to their respective sport skill requirements could. possibly clarify how certain sports may moderate the competitive trait anxiety/sport performance relationship. In terms of individual differences, equivocal findings exist between males and females and their levels of competitive trait anxiety. During the initial development of SCAT, the means for competitive trait anxiety were found to be higher for females than for males across all age-groups 5 (Martens, 1977). Martens and his colleagues (1990) point out that, although at the youth sport level females tend to be higher in competitive trait anxiety than.males, this trend is reversed for high school and college subjects. Additional investigations conducted by Hogg (1980), Krotee (1980), and Gill (1988) found females to be higher in competitive trait anxiety than males. Others, however, found no gender differences in competitive trait anxiety (Burhans, Richman, & Bergey 1988; Feltz & Albrecht 1986; Rainey, Conklin, & Rainey 1987; Rainey & Cunningham 1988; Smith, 1983). In researching the competitive trait anxiety levels in elite runners, Durtshi and Weiss (1984) report the means for females to be significantly lower than for males. Additional investigations into the link between competitive trait anxiety, gender, and sport performance suggest a rather strong negative relationship for females (Gerson & Deshaies, 1978; Hisanaga, 1982; Lanning & Hisanaga, 1983) as opposed to a positive relationship for males (Weinberg, 1979; Weinberg & Genuchi, 1980). These studies imply that as anxiety increases, a female’s performance will decrease; whereas for males, an increase in anxiety will yield.an increase in performance. The present synthesis of studies investigated this controversy'and attempted to determine whether SCAT predicts performance differently for males and females. Other individual factors that have contributed to the equivocal findings over competitive trait anxiety are age and 6 ability levels of subjects. In regard to age, during the development of SCAT, normative data compiled by Martens (1977) indicated that competitive trait anxiety increased slightly with age. More recent research supports these initial findings by demonstrating that younger athletes are indeed lower in competitive trait anxiety than older athletes (Hogg, 1980; Power, 1982; Watson, 1986). Hogg’s (1980) research. also established that a sample of female youth swimmers under the age of 11 were lower in competitive trait anxiety than all other age groups. In contrast, studies conductedwwith.children aged 10 to 13 (Smith, 1983) and 9 to 15 (Feltz & Albrecht, 1986) found no differences in competitive trait anxiety. Adding to the confusion, Gould (1983) reported that his study conducted with 13 to 19 year old elite‘wrestlers showed higher levels of competitive trait anxiety for the younger wrestlers as opposed to the older ones. To understand the ambivalence regarding the relationship between competitive trait anxiety and sport performance, as mediated by the athletes’ ages, the following factors need to be considered: First, what are the age ranges of the various samples? Second, at what level do the subjects compete (e.g., elite or recreational)? Third, how important do the subjects rate the contest? Fourth, how nmch exposure to competitive situations have the athletes had in the past (Martens et al., 1990). The relationships among competitive trait anxiety; age, and 7 sport performance have not been investigated with these issues in mind. An overall analysis of the research on competitive trait anxiety and Sport performance by age could show to what degree age might moderate the competitive trait anxiety/sport performance relationship. In terms of ability, Gravelle, Searle, and St.Jean (1982) investigated the relationship between sport ability and competitive trait anxiety. Their investigation was conducted with candidates for theeCanadian national volleyball teanh The mean SCAT scores indicated that all subjects scored below the 50th percentile, indicating low levels of competitive trait anxiety for these subjects. Additionally, research conducted with athletes chosen for an all—star game showed that their levels of competitive trait anxiety were significantly lower than that of those athletes designated as substitute players (Smith, 1983). However, research conducted by Power (1982), Gould et al. (1983) , Miller and Miller (1985), and Passer (1983) provided no evidence as to the differences in competitive trait anxiety between successful and less successful athletes, thus adding ambivalence to research on athletic ability and competitive trait anxiety. The mediating effect of an athlete’s ability on the competitive trait anxiety/sport performance relationship is not clear. However, because the research in this particular area cannot support a specific influence that ability'may have on competitive trait anxiety, an investigation to explore to 8 what degree ability is involved in predicting sport performance would provide a first step in developing new ideas and showing what the overall relationship may be. The frequent disparity between SCAT and the variables causing equivocal findings (i.e., type of task.and individual differences) in terms of competitive anxiety have been identified. The present synthesis of studies attempted to explore the controversies, but more importantly, investigated the direct link between competitive trait anxiety and performance by looking at the degree to which type of sport, gender, age, and.ability'may moderate the relationship between competitive trait anxiety and performance. As this study is primarily concerned with the relationship between competitive trait anxiety and performance, identifying the most reliable statistical procedure becomes crucial. Research in any area of interest should be able to produce an integrated statement of the findings. Because results from a single study are inconclusive, meta-analysis has been used to integrate the findings of many studies. This technique has also been used effectively in the past in the sport and exercise domain (Crews & Landers, 1987; Feltz &'Landers, 1983; Lirgg, 1991; Thomas & French, 1985). Meta-analysis is defined as an analysis of analyses, where many individual statistical findings are combined into a single analysis (Glass, 1976, 1977). Consequently, :meta- analysis is considered a review of studies, with shared or 9 common conceptual hypotheses, rather than a single statistical analysis. In the present meta—analysis, SCAT is assumed to be a measure of A-trait that can predict performance. In addition, studies that use SCAT are assumed to share the hypothesis that a relationship exists between trait anxiety and sport performance. The procedure to test whether the studies used share the same hypothesis is a test of homogeneity. If this test is found to be significant, the correlations are considered heterogeneous (different across studies); if the test is found to be nonsignificant, the correlations are considered homogeneous, that is, they come from the same population of studies addressing a similar hypothesis. Therefore, in using meta-analytic techniques, a clearer understanding of the relationship between competitive trait anxiety and sport performance can be achieved. Statement of the Problem The purpose of this research synthesis was to investigate the'overall relationship between.competitive trait.anxiety,.as measured by SCAT, and sport performance. In addition, several variables (type of sport, gender, age, .and ability) hypothesized to affect the anxiety/performance relationship were examined. The procedure used in this investigation was meta-analysis. 10 Questions that Prompted this Study 1. What is the overall average correlation between competitive trait anxiety and performance as measured by SCAT? 2. Do any of the following sport-related features influence the relationship between competitive trait anxiety and performance as measured by SCAT? a. Gender b. Ability c. Age d. Type of Sport (open vs closed skills) 3. Are the correlations for each relationship homogeneous? Delimitations This study was delimited to studies that looked at performance in terms of athletic competition. Additionally, only those studies that provided a Pearson correlation coefficient between SCAT and sport performance, or a statistic that could be converted to a Pearson correlation coefficient, were included. 11 Definitions The following definitions apply to this investigation: 1. Trait anxiety: This is a predisposition to perceive certain environmental stimuli as threatening or nonthreatening. These stimuli are then responded to with varying levels of state anxiety (Martens et al., 1990). 2. State anxiety: Emotional states are characterized by feelings of apprehension and/or tension and are linked with negative affect to the specific situational constraints of the event (Silva & Weinberg, 1984). 3. Competitive trait anxiety} This.is anrathlete’s tendency to perceive competitive situations as threatening or nonthreatening and to respond with varying levels of state anxiety (Martens et al., 1990). 4. Sport Competition Anxiety Test: This measurement tool, developed by Martens (1977), provides a reliable and valid indication of an athlete's level of competitive trait anxiety. This test is abbreviated as SCAT. 12 5. Meta—analysis: This statistical procedure provides a way to review and derive conclusions for research with a common theoretical and statistical base. Limitations This study was limited by the small number of correlations due to a relative lack of studies . Additionally, methodological problems which may be inherent in the studies used were not addressed statistically. CHAPTER II A REVIEW OF THE LITERATURE Introduction Most athletes and coaches believe that the relationship between arousal, anxiety, and sport performance is worth investigating. Although early research in the field of sport personality' produced. equivocal results regarding' the arousal/anxiety performance relationship, more recent studies indicate that being aroused or anxious is a complex emotion that is not necessarily performance—enhancing. Furthermore, the concepts of arousal and anxiety have not been clearly differentiated in past research. In order to clarify how these emotions can affect an athlete’s performance, this chapter presents definitions of arousal and.anxiety. Additionally, several theories inrarousal and anxiety are reviewed. These theories are drive theory, the inverted—U hypothesis, a cognitive model, and the optimal level theory. SCAT, an instrument used to measure competition anxiety, is also examined. Lastly, a discussion of meta— analysis, a research synthesis procedure, is presented to aid in bringing an understanding of how the encompassing body of literature in anxiety can be statistically synthesized in order to interpret contradictory findings. 13 14 Defining Arousal and Anxiety Although there are many different definitions of arousal, the simplest description of arousal is a physiological response to a situation that motivates, directs, and integrates a person's behavior. In scientific research, the construct of arousal has been used interchangeably with terms such as tension (Duffy, 1962; Kauss, 1980), activation (Duffy, 1962), drive (Passer, 1984), and anxiety (Landers, 1980). The need to distinguish between arousal and anxiety is important because these constructs are not synonymous. Arousal is the cumulative product of all emotions, or the entire continuum of an individual's psychological activation, and is often used to describe the combined effects of an individual's feelings at any one time (Sonstroem, 1984). As such, the psychological activity labeled arousal may vary from deep sleep to extreme excitement. In terms of athletics, arousal is often viewed as a motivating emotion where mid-range arousal is associated with top performance (Landers, 1980; Sonstroem, 1984). A certain level of arousal is always present and it is that level of arousal that dictates the intensity at which we function. Iso—Ahola and Hatfield (1986) define arousal as the level of physical or mental activity ranging from deep sleep to highly agitated, emotion-stricken terror. Conversely, anxiety has been described as a feeling of uncertainty, discomfort, and an overall negative sense of being (Zuckerman, 15 1979). Anxiety has also been defined as excessive arousal or a heightened state of physiological activation to include negative feelings (Iso-Ahola & Hatfield, 1986). Although the difference between arousal and anxiety has been established, the reader should note that most research in sport psychology still uses the terms arousal and anxiety interchangeably. Several theories that explain the relationship between arousal and performance have been proposed. Among these are (a) drive theory, (b) the inverted-U' hypothesis, (c) a cognitive model, and (d) the optimum level theory. All of these theories attempt to explain the impact of arousal on performance. Arousal Theories Drive Theory In drive theory, a direct relationship between arousal and performance is proposed. This theory hypothesizes that as arousal increases, so will performance (Singer, 1975). Drive theory proposes three factors in arousal. The first factor is drive (D), defined as physiological arousal. Second, habit (H) is described as the dominance of correct or incorrect responses. Both habit and drive will have significant effects on the third factor, performance (P) (Hull, 1943; Spence & Spence, 1966). The relationship of P, H, and.D has been placed 16 into the mathematical equation: P = H x D (Hull, 1943). One criticism. of drive theory is its difficulty in assessing habit strength (Spence, Farber, & McFann, 1956) . The dominance of correct or incorrect responses is especially difficult to determine with motor tasks. Therefore, some researchers believe that drive theory cannot be adequately tested.in motor performance (Carron, 1980). Also, in regard to motor tasks, some researchers suggest that drive theory is applicable only for simple motor tasks, such as weightlifting and the shotput (Carron, 1980; Singer, 1975). The Inverted—U Hypothesis The inverted-U hypothesis is not really a theory, in that it only describes a relationship between arousal and jperformance; it does not explain.why this relationship exists. This hypothesis states that when individuals become 'over psyched“ or too motivated and aroused, performance drops steadily (Duffy, 1962; Landers, 1980; Martens, 1974; Rushall, 1979). Like drive theory, the inverted-U hypothesis also recognizes that as arousal increases, performance will increase. However, unlike drive theory, a critical point will be reached at which time arousal and performance are at their optimum. If the level of arousal is still steadily increased, performance will decrease proportionately (Duffy, 1962). If represented.iJ1ea graph, this curvilinear relationship will look like an inverted-U, explaining the peculiar name of this 17 hypothesis. The inverted-U hypothesis finds support from empirical evidence and from.its intuitive appeal. Support has been.shown in both lab experiments (e.g., Castaneda, 1956) and field experiments (e.g., Fenz & Jones, 1972). However, one problem in examining the inverted-U hypothesis is the difficulty in verifying the different levels of arousal, because a minimum of three different levels of arousal must be observed (low, optimal, high). Furthermore, the definition of optimal level of arousal is difficult to determine when considering individual differences in physiological responses to arousal (Iso-Ahola & Hatfield, 1986). A Cognitive Model Both drive theory and the inverted-U hypothesis have been extensively researched in examining the effects of arousal on performance. Rather than dismissing these models of arousal, researchers have found that drive theory and the inverted-U hypothesis act together with cognitions (Burton, 1988). Thus, affect, cognition, and anxiety have become valuable concepts in arousal research. Sport psychologists now recognize that arousal is a complex emotion with various components that may affect an individual’s performance in many ways. In addition to complex physiological differences that characterize various states of the inverted-U hypothesis, recent studies have examined affect and cognition in terms of their influence on 18 performance (Neiss, 1988). In a study conducted with swimmers, Burton (1988) found that one can differentiate between cognitive and somatic anxiety. Thus, a multidimensional concept of anxiety should be considered. Cognitive anxiety is described as all the negative feelings about competition that are characterized by worry, negative self-talk, and unpleasant visual imagery. Cognitive anxiety arises when expectations of success become negative. Cognitive anxiety has also been shown to have a negative linear relationship with sport performance (Burton, 1988). Somatic anxiety is the physiological component of anxiety that is directly related to autonomic arousal. Somatic anxiety is typically short in duration.and manifests itself as anxiety during lockerroom preparations, pre-contest warm ups, and when spectators are present (Burton, 1988). Somatic anxiety characteristically increases prior to competition and reaches its peak as the competition begins. Then, during competition, it decreases rapidly. Somatic anxiety may, therefore, be expected to affect performance only in the initial stages of the competition. The Optimum.Level Theory In its simplest terms, the optimum level theory (Berlynes, 1960) proposes that every individual has his/her own optimal level of stimulation, complexity, and arousal (Iso-Ahola & Hatfield, 1986; Carron, 1980). The diurnal cycle shows that 19 throughout the day, the optimal level of arousal varies with the level of arousal that is regulated by the intrinsic diurnal rhythms. As an example, drowsiness which is below optimum during the day is Optimum at night before sleep. Thus, the optimal level of arousal throughout the day must track the diurnal levels, but the level of arousal also depends on the realities of the task demands (Zuckerman, 1979). When a state of optimal arousal has been reached, one has the ability to discriminate between many available situational cues and select the most relevant ones to which to respond (Singer, 1975). In non—athletic situations, only the amount of arousal necessary to complete the task is employed (Duffy, 1962). As Zuckerman (1979) suggested, arousal potential represents an overall power to excite the nervous system, to command attention, and to influence behavior. Landers (1980), in.building upon.Easterbrook’s (1959) early researCh on arousal, stated that as the levels of arousal increase, a narrowing of attention will result. Nideffer (1976) saw this occurrence as a positive response that helps an athlete focus on his or her task. Attentional narrowing is not viewed as a positive side effect by Landers, however, because as he pointed out, it is accompanied by a loss of sensitivity to environmental cues. Thus, the proper level of arousal would. depend. on the (attentional level necessary (utilization of all necessary cues) to complete a task. This "tunneling effect," as Carron (1980) referred to it, mostly 20 affects the performance in sports involving complex tasks. As arousal increases, attention decreases and is accompanied by a narrowing of the visual field. The higher the breadth of perceptual cues necessary to complete a task, the more an individual's attention. will suffer' under' high levels of arousal. Thus, sports demanding a narrower level of attentional focus .may tolerate ihigher levels of arousal because there are fewer task cues and less chance of task relevant cues being eliminated through the perceptual narrowing process (Duffy, 1962; Iso-Ahola & Hatfield, 1986; Rushall, 1979). Summagy A great many laboratory experiments and field studies have been conducted in an attempt to better understand how performance may benefit, or suffer, from varying levels of arousal. Arousal is generally depicted as a positive force in sport when examined in terms of an individual being "psyched. " However, in very broad terms and across all sports, high arousal does not necessarily help an athlete perform better. It may give the athlete a sudden surge of energy, but also be increased to a point where the benefit of additional energy will adversely affect the athlete's performance (Nideffer, 1976; Oxendine, 1970). A.great deal of research relating to motor tasks (e.g., the level of difficulty and the necessary level of arousal needed 21 in order to complete a task) has been conducted and has led to some general conclusions. First, high levels of arousal are Optimal for performance in gross motor activities that feature strength, speed, and endurance, such as weightlifting and various track and field events (Oxendine, 1970). Second, high levels of arousal are considered countersproductive when complex skills, fine muscle movement, coordination, steadiness, and concentration are necessary to complete the task (e.g., gymnastics and archery) (Oxendine, 1970). Lastly, a slightly above-average level of arousal for normal or sub- normal motor tasks (daily chores) is recommended (Fenz & Jones, 1972; Iso-Ahola & Hatfield, 1986; Kauss, 1980; Martens, 1974; Oxendine, 1970; Sonstroem, 1984). Com etitive Anxiet As in many areas of research, the individuality factor must be considered when investigating the anxiety/sport performance relationship (Kauss, 1980). Specifically, in terms of sport, the phenomenon of each individual’s anxiety level must be considered. In sport, anxiety is referred to as competitive anxiety and defined as a tendency to perceive competitive situations as threatening (Martens, 1977). Recognition must.be given to each individual having his or her own characteristic, optimal level of stimulation in arousal, anxiety, cognitive 22 activity, and motor activity. Some individuals may perform better with.high levels of anxiety and some with low levels of anxiety. The optimal level of arousal may vary with age, learning' experiences, recent levels. of stimulation, task demands, and diurnal cycle. Additionally, individuals may differ in anxiety for genetic or environmental reasons that, in turn, 1mg! affect their temperaments (Zuckerman, 1970). Individual differences in the level of arousal could also be due to variations in environmental demands, type of nervous and endocrine system, and genetics (Duffy, 1962; Oxendine, 1970). In an attempt to define individualism.and place individual levels of arousal and anxiety in a more measurable context, Spielberger (1966) conceptualized arousal in terms of trait anxiety and state anxiety. In Spielberger’s words, anxiety states are characterized by subjective, consciously perceived feelings of apprehension and tension, accompanied by or associated with activation or arousal of the ANS. (p. 17) 23 Trait anxiety on the other hand is: a motive or acquired behavioral disposition that predisposes an individual to perceive a wide range of objectively non—dangerous circumstances as threatening and to respond to these with state anxiety reactions disproportionate in intensity to the magnitude of the objective danger. (p. 17) Martens (1974), further" elaborating' on ‘trait anxiety, believed that some people experience some situations as more arousing than others. However, these individuals are not necessarily higher in arousal. Generally, individuals high in trait anxiety show higher state anxiety as opposed to people low in trait anxiety (Martens, 1977). Hence, the interactions between state anxiety and trait anxiety may be viewed as codeterminants of behavior without specifying either as a primary or secondary source:of arousal (Martens, 1977; Passer, 1984). Martens (1977) developed the concept of competitive trait anxiety to highlight the anxiety athletes experience in sport settings. This construct is described as an individual’s tendency to perceive competitive situations as threatening and to respond with state anxiety to these perceived threats. Because the construct of competitive trait anxiety was 24 hypothesized to have varying effects on sport performance (Martens, 1977) , the need to measure competitive trait anxiety became apparent. The development of SCAT provided a measure of competitive trait anxiety with which to assess the specific anxiety that occurs in competitive situations (Martens, 1977; Martens et al., 1990). The Competitive Anxiety Model When discussing the influence of trait anxiety on sport performance one must not neglect the parallels to behavioral influence and prediction. In personality research, traits are assumed to be fundamental units of each person’s personality (Spielberger, 1966) . Additional assumptions presume that these traits are stable and consistent attributes. In trying to bring a greater understanding to the phenomenon of anxiety and behavior, three different paradigms have been proposed. The first is the trait approach. Proponents of this theory believe that behavioral differences lie within the person (e.g., Eysenck, 1952). Second, situationisni implies that human behavior is largely determined by situation specific factors which may be identified within the environment (Mishel, 1968, 1973). Thirdly, interactionism suggests that behavior is a product of both personality and situations, and that if both variables are taken into account, behavioral prediction becomes more accurate (Bowers, 1973; Carson, 1969; Magnuson & Endler, 1977; Vale & Vale, 1969). 25 Through the assessment of inter-individual responses and particular situations, the essential elements of interactionism were implemented in Martens' (1977) development of a competitive anxiety model (Martens et al., 1990). In addition, Martens explained that the following four theoretical developments in the field of personality guided the development of the competitive anxiety model: 1. The adoption of an interactional theory of personality that predicted behavior better than did trait or situational paradigms. 2. The trait-state theory of anxiety, which distinguished between trait anxiety and state anxiety. 3. The development of situation-specific trait anxiety instruments that had superior predictive power compared to general trait anxiety scales. 4. The development of a conceptual model for the study of competition as a social process (Martens et al., 1990). Embedded within this model lies the competitive process, which includes the following elements: (a) the objective competitive situation, (b) the subjective competitive situation, (c) the response, and (d) the consequence of engaging in competition. Although the interactionist theory is presented with logic and is intuitively appealing, the model lacks the inclusion of cognitions. A subsequent model (see Figure 1) with emphasis on the cognitive process was suggested by Martens (1977). Consequences Figure l. 26 shuafion Objective competitive /7‘° " \\ 0 Of. 6’ [$56 {yo ‘9‘”) \ O . g . Y 6),]. 44' Qualities 006‘ of the person \I‘e ’6‘“ \" I / L Response The Competitive Process Subjective competitive situation Note. From Competitive Anxiety in Sport (pp. 18, 70) by R. Martens, R.S. Vealey, and D. Burton, Champaign, 1990. IL: Human Kinetics. Copyright 1990 by Human Kinetics Publishers. Reprinted by perm1551on. 27 To briefly illustrate the four components of the model, one must initially consider the objective competitive situation which specifies the environmental demands of the competition such as the type of task, difficulty of opponents, playing conditions and/or rules, and availability of extrinsic rewards. How a person perceives the components of the objective competitive situation may be indicated by his/her level of trait anxiety and.is ultimately termed.the subjective competitive situation. The subjective competitive situation is mediated by personality dispositions, attitudes and abilities . The next factor, response, is largely determined by the subjective competitive situation and manifests itself as behavioral responses, physiological responses, and psychological responses. The consequence is largely, but not exclusively, determined by the outcome of the competition. Because the studies using trait anxiety'measures wereroften testing various theoretical predictions, the need to categorize these studies became apparent. A categorization according to the theoretical predictions that were tested becomes possible by combining the models of the competitive process and competitive trait anxiety (see Figures 1 and 2). Figure 3 is a representation of this combination. The four-link process of the model of competitive anxiety in Figure 3 begins in Link.l, where situational factors in the objective competitive situation (taken from Figures 1 and 2) and intrapersonal factors (e.g., competitive trait anxiety) Objective competitive situation 28 Competitive A-trait V Perception oi threat A-state reaction Figure 2. Competitive trait anxiety as a mediator between competitive stimulus and response. Note. From Competitive Anxiety in Spgrt (PP. 18. 70) by R. Martens, R.S. Vealey, and D. Burton, 1990, Champaign, IL: Human Kinetics. Copyright 1990 by Human Kinetics Publishers. Reprinted by permission. 0 Pedorma'nce outcomee Consequences) 4/ Situationd (actors HIS) lntrapersonal tactore (A-trait) State respomee (A-state) ' Pertormance Perception ot threat (868) Figure 3. Model of competitive anxiety. Note. From Competitive Anxiety in Sport (p. 70) by R. Martens, R.S. Vealey, and D Burton. 1990. Champaign, IL: Human Kinetics. Copyright 1990 by Human Kinetics Publishers. Reprinted by permission. 29 interact to create a perception of threat (taken from Figure 1). This perception of threat, that is part of the subjective competitive situation (taken from Figure 1), then interacts in Link:2 with intrapersonal factors to influence an individual’s state responses. Different performance outcomes, or consequences are then created through the interaction of cognitive, somatic, and behavioral responses with intrapersonal factors (Link 3). The cycle is then completed with the reciprocal influence of performance outcomes on intrapersonal factors (Link 4). The intrapersonal factors, which are the central part of this model, are identified as sport-specific personality dispositions, such as fear of failure and.the tendency to feel threatened (Gill, Dzewatowski, & Deeter, 1988; Vealey, 1986; Willis, 1982). Additional intrapersonal factors, such as general personality dispositions (e.g., test anxiety, gender, age, ability, and. participation status), have also ‘been studied. In Link.1 of the competitive anxiety model, the interaction of situational and intrapersonal factors are responsible for an ensuing perception of threat within a competitive situation. Specifically, perception of threat is characterized by fear of failure, fear of evaluation, ego threat, outcome uncertainty, negative outcome certainty, external control, and perceived importance of the task. In terms of fear and.threat, research has established that high competitive trait anxiety 30 individuals are threatened more in a competitive situation than are low competitive trait anxiety individuals (Brustad, 1988; Brustad & Weiss, 1987; Feltz & Albrecht, 1986; Gould, Horn, & Spreeman, 1983; Passer, 1983; Rainey, Conklin, & Rainey, 1987). Link 2 represents the influence of intrapersonal factors and perception of threat on state anxiety responses in terms of performance. Several laboratory and field studies have been conducted to shed light upon this relationship. This research has found SCAT to be a significant predictor of general state anxiety, cognitive, somatic, and competitive state anxiety (Gill & Martens, 1977; Martens & Gill, 1976; Martens, Burton, Rivkin, & Simon, 1980; Murphy & Woolfolk, 1987; Poteet & Weinberg, 1980; Scanlan, 1975, 1978; Scanlan & Passer, 1977, 1978, 1979b; Watson, 1986; Weinberg, 1978; Weinberg 8: Genuchi, 1980). According to interactional theory, performance is best predicted tar state anxiety responses. This relationship is displayed in Link 3 of the competition anxiety model. According tx: the competitive anxiety npdel, trait anxiety should not be able to predict performance. However, the trait anxiety/sport performance relationship has been examined in various laboratory and field studies. This research, although mostly equivocal in terms of predicting performance for a specific sport, has indicated that in situations where performance is based on complex motor skills or is 31 qualitatively measured, SCAT can make significant predictions (Cooley, 1987; Hubard & McKelvie, 1986; Scanlan, 1978, Scanlan & Ragan, 1978; Sonstroem & Bernardo, 1982) . The ambivalence of this research is demonstrated when the findings conducted with specific sports are examined. Here, significant relationships with SCAT have been reported in such tasks as overarm throwing, basketball, golf, and team batting average (Gerson & Desaies, 1978; Sonstroem & Bernardo, 1982; Weinberg, 1977; Weinberg & Genuchi, 1980). Nonsignificant relationships were reported in studies investigating the trait anxiety/performance relationship using a computerized dice game, golf putting, a flexed-arm hang-strength task, rugby, and marathon running (Broughton 8: Perlstrom, 1986; Maynard & Howe, 1987; McKelvie, Valliant, & Asu, 1985; Murphy & Woolfolk, 1987; Poteet & Weinberg, 1980). Therefore, the overall mediating effects of a particular sport on competitive trait anxiety cannot be determined by SCAT and predictions into subsequent levels of performance have been poor. After competition, individuals have different thoughts, feelings, and behaviors that are based not only on the individual’s performance but also on intrapersonal factors that influence how a person may perceive his or her performance. Research examining the effects of trait anxiety and situational factors on thoughts, feelings, and behaviors of athletes has provided the following information. First, competitive trait anxiety may influence postcompetition state 32 anxiety (Scanlan & Lewthwaite, 1984); however, the effects are minimal when compared to performance influences (Simon & Martens, 1977). Second, postcompetition expectancies are not influenced by trait anxiety (Scanlan & Passer, 1979a, 1981), and when achievement motivation levels are considered, postcompetition social comparison preferences based on competitive trait anxiety emerge (Scanlan & Ragan, 1978). Third, significance has been reported between competitive trait anxiety and perceived performance outcomes and perceived impact of tension on performance (Hubard & McKelvie, 1986). However, competitive trait anxiety and causal attributions have been found to be unrelated (Passer, 1983). The influence of performance outcomes on intrapersonal factors, specifically trait anxiety, is represented in Link 4 of the competition anxiety model. The cycle of reciprocal interactionisniis completed.asione learns that an individual’s level of competitive trait anxiety is based on the consequences of all prior experiences in sport. This link represents a process which occurs over time. Although no research has been conducted to test the link between the influence of different performance outcomes or consequences on competitive trait anxiety, a longitudinal approach would be the most appropriate way to shed light upon this relationship (Martens et al., 1990). 33 Development of SCAT Competitive trait anxiety has been identified as a construct that describes individual differences in the perception of threat, and/or the state anxiety responses to the perceived threat. Because levels of perceived threat often vary from individual to individual, SCAT was developed to provide a valid and reliable measure of competitive trait anxiety' (Martens, 1977). Spielberger’s (1973) State-Trait Anxiety Inventory for Children (STAIC) served as a model for the format to be used with SCAT’s child (10—14 years of age) and adult (15 years and.older) versions. Additionally, several selection criteria. were established. for* developing SCAT. First, SCAI'was to be an objective, as opposed to a projective scale. Second, response bias had to be minimized. Third, procedures for administering SCAT had to be unambiguous and the completion time had to be short. Finally, SCAT had to be easy to score. In SCAT’s initial development, a pool of 75 items were selected and modified from Taylor's (1953) Manifest Anxiety Scale, Spielberger’s (1973) STAIC, and Sarason, Davidson, Lighthall, Waite, and Ruebush’s (1960) General Anxiety Scales. These items were then rated by six judges on face or content validity and clarity of sentence structure. After three versions and various item, statistical, and discriminant analyses, the SCAT-C (for children) was derived and satisfied 34 all acceptance criteria. Martens also changed the SCAT—A (for adults) by modifying the instructions of SCAT—C to be more appropriate for adults and by changing one word of one item. SCAT-A.was also found to have exceeded the required selection criteria. The problem of bias was addressed by (a) not referring to SCAT as an anxiety scale, (b) including spurious items, and (c) reversing some of the test items. The problem of social desirability was not addressed because the “lie scales” were found to suffer from the same weakness that they were supposed to detect (Martens et al., 1990). By reversing some test items, however, response bias was shown to be minimal (Smith, 1969; Spielberger, Gorsuch, & Lushene, 1970). The two forms, SCAT-Cland.SCAT-Aq are self—administered.and can be taken alone or in groups. The questionnaire contains 15 questions and simply asks people how they usually feel when competing in sport. Each item is scored on a 3-point Likert scale (often, sometimes, hardly ever), with total points ranging from 10 (low competitive trait anxiety) to 30 (high competitive trait anxiety). Generally, the test takes only about 5 ndnutes to complete and is most often administered under the title of “The Illinois Competition Questionnaire." Reliability and Validity Issues SCAT’s reliability was established by several test—retest techniques with various samples and analyses of variance 35 (ANOVA). These procedures resulted in high test-retest reliability (mean = 0.77) and ANOVA reliability of 0.81 for SCAT-C and 0.85 for SCAT-A. The 10 final items yielded mean item analysis coefficients of 0.64 for high competitive trait anxiety and 0.58 for low competitive trait anxiety, a triserial correlation of 0.68, and a ‘mean discriminant function coefficient of 1.39. The degree of internal consistency (the homogeneity of scale items) was assessed using the Kuder-Richardson formula 20 (KR-20). This analysis resulted in KR—20 coefficients of.; = 0.95 and .1.“ = 0.97 for SCAT—A and SCAT-C respectively, demonstrating a high degree of internal consistency. Reliability is essential for validity to occur; however, reliability does not ensure validity. Content validity was provided by having six qualified researchers evaluate the initial 75 items of SCAT. Only those items rated.highly by all judges were retained. Concurrent validity was obtained by correlating SCAT with other personality constructs to assess how well empirical relationships matched theoretical predictions. Concurrent validity was demonstrated through a significant relationship between competitive trait anxiety, as measured by SCAT, and the following personality constructs: The'ChildrenfslManifest Anxiety Scale Short Form.(Levy, 1958), the General Anxiety Scale for Children (Sarason.et al., 1960), the Trait Anxiety Inventory for Children (Spielberger, 1973), the Junior-Senior High School Personality Questionnaire 36 (Cattell & Cattell, 1969), the Social Avoidance and Distress Scale and the Fear of Negative Evaluation Scale Internal-External Control Scale (Watson & for Friend, 1969), the Children (Bialer, 1961), and the Mehrabian Achievement Mbtivation Scale (Mehrabian, 1968). construct validity, evidence that To demonstrate competitive trait anxiety is related to other constructs consistent with its theoretical predictions must be shown. In theory, SCAT should measure competitive trait anxiety and predict state.anxietyu Martens (1977) based this validation on six specific theoretical predictions from the model. 1. High SCAT subjects will manifest higher state anxiety than will low SCAT subjects in competitive but not in non-competitive situations. SCAT’will predict precompetition but not noncompetition state anxiety better than will other person, task, and situation.variables. SCAT will correlate more strongly with state anxiety in competitive than noncompetitive situations. SCAT“will correlatezmore strongly with competitive state anxiety as the level of situational threat increases. SCAT will predict competitive but not noncompetitive state anxiety better than will other trait anxiety measures. High SCAT subjects will perform poorer than will low SCAT subjects in competitive but not in noncompetitive 37 situations. Many researchers conducted experimental and field studies to test these six points (e.g., Martens & Gill, 1976; Martens & Simon, 1976; Scanlan, 1975; Simon & Martens, 1977; Weinberg, 1977). Various investigative techniques were used to provide data on SCAT's construct validity. This research found.SCAfl'to be a valid measure of competitive trait anxiety within its theoretical predictions. Specifically, these investigations established that high SCAT subjects had higher state anxiety than low SCAT subjects in competitive situations but not in noncompetitive situations, that SCAT was a better predictor of state anxiety than other person, task, and situational variables in competitive situations, and that SCAT correlated more strongly with state anxiety in competitive situations than in noncompetitive situations. In summary, the experimental and field studies conducted support SCAT's reliability and validity in.accordance with the theoretical predictions.of competitive trait anxiety; In terms of validity, SCAT has proven to be notable in its measurement of competitive trait anxiety as demonstrated by SCAT's prediction of pregame state anxiety scores (;_= 0.64) compared to STAI trait anxiety (g = 0.30) and coaches ratings (g = 0.12) (Martens & Simon, 1976). SCAT has also been successful and reliable in assessing an athlete’s perception of threat. For example, high competitive trait anxiety subjects perceive competitive situations as more threatening than do low 38 competitive trait anxiety subjects (Cooley, 1987; Huband & McKelvie, 1986; Murphy & WOolfolk, 1987; Poteet & Weinberg, 1980; Scanlan, 1978; Wandzilak, Potter' & Lorenzen 1982; Watson, 1986; Weinberg, 1978, 1979; 1980). Weinberg & Gennuchi, SCAT was also identified as an important predictor of the following perceived sources of threat: fear of failure/evaluation (Crocker et al., 1988), ego threat (Kanteroliatis & Gill, 1987), outcome uncertainty (Gould et al., 1983), and external control (Scanlan & Passer, 1979a, 1981). Few studies have been conducted to test the last prediction that high SCAT subjects will perform poorer than low SCAT subjects in competitive but not in noncompetitive situations. In fact, the strength of the relationship between SCAT and performance in competitive situations is varied. The present research synthesis investigated the nature and strength of this relationship through a meta—analysis. A Review of Meta-Analysis Introduction The biological, physical, and.hatural sciences are clearly defined and examined by standardized, quantifiable techniques. In social sciences, however, human behavior is more complex and difficult to explain, research environments are more 39 difficult to control, common definitions are not universally agreed. upon, and. :methods, techniques, and sampling characteristics vary frequently from study to study (Wolf, 1990). These research problems lead to conflicting results which necessitate further research. Thus, establishing guidelines for reliable and valid reviews, integrations, and syntheses becomes critical. Meta-analysis is one way in which to synthesize the literatures and allows one to address the following potential problems with traditional literature reviews: 1. Selectively including studies, 2. Differential subjective weighing of studies, 3. Misleading interpretations of study findings, 4. Not using the characteristics of the studies as potential explanations for disparate or consistent results across studies, and 5. Not being able to examine moderating variables (Wolf, 1990). The basic concepts underlying meta-analysis were first employed by Thorndike (1933) and, subsequently, by Ghiselli (1949). In the early 1970’s, Levine, Romashko, and Fleishman (1973) and Levine, Kramer, and Levine (1975) also employed these concepts. However, these researchers failed to solve the probleni of integrating findings across studies to jyield cumulative knowledge. Glass (1976) is credited for solving this problem.and for first coining the term I'meta—analysis." 40 Meta-analysis is defined by Glass as “a rigorous alternative to the casual, narrative~discussions of research studies which typify our attempts to make sense of the rapidly expanding research literature“ (1976, p. 3). The review of a large body of literature is most efficiently executed when quantitative methods are employed. Meta-analysis allows stronger conclusions to be drawn because studies are not prejudged a priori and/or excluded because of a particular research design. Thus, the meta-analytic method of revieW' is more objective than traditional literature reviews. Meta-analysis may also highlight gaps in the literature, provide insight into new directions for research, and identify mediating, interactional relationships, and/or trends. Finally, verification of heterogeneity (different, or unrelated.studies and/or hypotheses) allows for identification of outliers and.may lead to an increased understanding of old and new hypotheses (Glass, 1983; Green & Hall, 1984; Light & Pillemer, 1984; Wolf, 1990). In meta-analysis, many studies are combined in order to interpret the effects of two or more variables upon each other. The underlying assumption in conducting a meta-analysis is that each study provides a sample estimate of the size of effect that is representative of the population effect size. In other words, the assumption of meta-analysis is that the studies are testing the same hypothesis (i.e., they are homogeneous). As a consequence, meta-analysis is used to 41 integrate the findings of many studies, compare correlations, correct the observed variance of the studies to eliminate sampling error, and finally to correct the mean variance of the population values for errors in neasurement and range variation (Hunter et al., 1982). Hence, an investigation using meta-analysis provides a method by which an understanding of the study features can be quantifiably interpreted. Additionally, the meta-analytic review is not limited to one specific type of statistical resolution when quantifiable interpretation is sought. Such assessment may be achieved in experimental studies (effect of an independent variable on a dependent variable), correlational studies (strength and direction of a relationship between.two variables), or surveys (a simple rate or incidence figure) (Glass, McGaw, & Smith, 1981). Stages of Meta-Analysis Several steps must be followed in order for meta-analysis to take on the rigorous and systematic approach necessary for reviewing research (Cooper, 1984). An explanation of these steps follows. Problem formation. Here, research questions are outlined. The questions are guided by certain conceptual and operational definitions within one’s meta—analytic review. In the present review the underlying question is: How effective is trait 42 anxiety, as measured by SCAT, in predicting sport performance? Data pollpction. This second step identifies studies that can be included into a specific :meta-analysis. In this analysis, those studies that provided correlations (or a way to calculate correlations) between SCAT scores and sport performance were included. pa:ta evaluation. In this stage, many study features of interest are coded for proper analysis. For example, two study features that were coded in the present analysis were subject’s age and.ability level. Guidance as to‘which features should be investigated is provided in the Problem Formation step. Data analysis and interpretations This fourth.step includes the selection of statistical tools for analysis based upon the statistical format of the research studies. In this meta— analysis, study findings were based on correlational coefficients. Public presentation of results. In this last step, the results are discussed. Depending upon the number of features addressed in the meta—analysis, the results may be more, or less encompassing. However, all aspects of the study need to be discussed in order to clarify or explain various obscure 43 findings as they may appear. Measuring Study Findings When employing meta-analytic techniques, the findings of the empirical studies should be integrated on a conunon scale. The study findings are then treated as the dependent variable in the data analysis stages of the meta-analysis. Three categories for integrating study findings on a common scale have been identified: significant test findings, standardized mean differences, and outcomes of correlational studies. Methods which can be used from studies that report significant test statistics include vote—counting and the combination of significant levels into a joint null hypothesis test. These methods, however, have several limitations, especially when the number of studies is small (Glass et al., 1981). Two scale—free indices used in meta—analysis are the standard mean difference (effect size) and the product-moment correlation. The effect size is defined as the population value of the standardized mean difference (Hedges, 1982), or the difference between the means of the experimental group and the control group divided by the control group standard deviation (Glass, 1977) . It is used extensively for expressing and combining results of studies that assess the effectiveness of an experimental treatment. The other index of effect magnitude, the product-moment correlation coefficient, is most often used in summarizing the relationship between two 44 continuous variables (Hedges & Olkin, 1985). In studies that do not report the means and standard deviations necessary for effect size calculations, one can derive the effect size from one of the formulas described by Glass, McGaw, and Smith (1981). In the analysis of effect sizes (correlations or proportions), the effect sizes could be analyzed using procedures such as ANOVA (Glass, 1976). According to Hedges (1981), though, this approach could be problematic. He points out that in using such an approach, the homoscedasticity assumption.cannot be satisfied. Thus, a new technique~designed for effect size examination was introduced (Hedges, 1981, 1982; Hedges & Olkin, 1983, 1985). This technique involved a test of homogeneity to determine whether the studies have a single population effect size and whether all studies incorporated into the meta-analysis share a common hypothesis. Hence, each study is assumed to provide a sample estimate of the size of effect that is representative of the population effect size. If the homogeneity test is non—significant, the various studies are testing the same hypothesis; in other words, they are homogeneous. Limitations and Criticisms One of the greatest criticisms of meta—analysis is that it compares quite different studies; i.e., it mixes apples and oranges (Wolf, 1990). However, there would be no need to compare studies that have the same findings. By necessity, 45 some study distinctions must be ignored, but this discarding of information must be done with distinctions that make no important difference and must be executed in an orderly fashion (Glass et al . , 1981) . Additionally, the oversimplification of results by down—playing mediating or interaction effects can be detrimental to accurate result interpretation. Therefore, potential mediating factors and interactions should be built into the design of a meta— analysis by coding characteristics of studies in order to empirically examine whether interactions exist. It is also difficult to draw logical conclusions when comparing and aggregating studies that include different measuring techniques, definitions of variables, and subjects. Additional criticisms of meta-analytic reviews point out that the results of a meta-analysis are uninterpretable because the results from poorly designed studies are included along with those of good studies. Additionally, meta-analysis tends to lend itself to bias because all published research is in favor of significant findings. Also, the effects may not be independent because multiple results from the same studies are often used (Wolf, 1990). These points should be addressed within each meta-analysis and the proper statistical and empirical steps taken to ensure reliability. In further addressing the criticisms of meta-analysis, one must attend to various techniques for reducing bias. For instance, the ''apples" and "oranges'I criticism is reduced when 46 each variable is coded properly, so that one can empirically test whether and how these variables are similar or different from each other. Also, when considering that some studies included into a meta-analysis are based upon different sample sizes (some may be small and unrepresentative of the population, others may be large and follow randomized control group designs), and the problem of bias in terms of assigning all the studies the same weight, weighing each of the standard normal deviates (Q's) by the size of the sample on which it is based should be considered. In doing so, the less representative studies (those with small n’s) will not contribute to the results of the meta—analysis as much as the more well-designed and more representative (large n’s) studies (Wolf, 1990) . Often a meta-analysis is biased because of the inclusion of only published studies. A study that reports nonsignificant findings will probably not be published. Hence, those studies that are published are biased towards statistical significance. Two reasons for publication bias may be that (a) an unpublished study was poorly designed and therefore did not meet the publication standards required, or (b) some studies may not have been submitted for publication in the first place because of relatively weak significance (Feltz & Landers, 1983) . This problem can be addressed, however, by employing a method that allows the researcher to establish how many studies confirming the null hypothesis would be needed in 47 order to reverse a decision that a significant relationship exists. This procedure is called the ''Fail Safe N." (Wolf, 1990). Summapy Many athletes and coaches have reported that sport performance is adversely affected when one is overly aroused or anxious. In response to such reports, researchers have examined how, and to what degree, arousal and anxiety may affect an athlete's performance. Several theories have been developed to provide a stepping stone for detailed research. in this area. Drive theory suggests that a positive linear relationship between arousal and performance exists; that is, as arousal increases, so does performance. Proponents of the inverted-U hypothesis suggest that if a person’s arousal is continuously increased, a point will be reached when the positive effect on performance is reversed and performance actually decreases. With the introduction (ME cognitions into arousal research, a multidimensional concept of anxiety was suggested that includes the concepts of cognitive anxiety (negative feelings about a given situation), and somatic anxiety (physiological responses). The optimum level theory accounted for individual differences in optimal stimulation and suggested that not one 48 single level of arousal will yield optimum performance for all athletes. Three terms are used to clarify the emotion of athletes in competition. First, anxiety is defined as the emotion encountered in athletics when the competition poses a threat to the athlete. Thus, the term “competitive anxiety“ was coined. Furthermore, an athlete's predisposition to perceive various circumstances (trait anxiety) is separated from those feelings that are specific t1) a. given situation (state anxiety). In order to lay a foundation and bring a clearer understanding to competitive anxiety, the model of the competitive process was developed (Martens, 1977). This model identified the concepts of the objective competitive situation, the subjective competitive situation, the response, and the consequence of engaging in competition. Through the inclusitmi of distinctions 'between. personality traits and states, and a need for categorization of research, the model of competitive anxiety outlines four separate, jyet interrelated links surrounding and affected by intrapersonal factors. The present review identifies the variables of age, gender, ability, and.type of sport as moderating the competitive trait anxiety/sport performance relationship. The aim of this investigation is to clarify the degree to which competitive trait anxiety, as measured by SCAT, is affected by those 49 'moderating variables outlined and will be able to aid in sport performance predictions. In addition to SCAT’s reliability and 'validity as a measurement tool of competitive trait anxiety, many experimental and field studies support SCAT's ability to predict trait anxiety in accordance with all theoretical predictions imposed on SCAT by the competitive trait anxiety model. In trying to bring a clearer understanding to the ;phenomenon of competitive trait anxiety and.to properly review the relationship of the variables to be studied, this investigation identifies meta—analysis as a method that could provide a quantifiable interpretation of the magnitude of influence among the various variables to be considered. CHAPTER I I I Method The degree to which trait anxiety, as measured by SCAT, can predict the level ofjperformancerof athletes in various sports was investigated using meta~analytic techniques. This chapter outlines the procedures implemented in study identification, coding, and data treatment. Collection of Studies The literature searCh was conducted manually. The publication Competitive Anxiety in Sport (Martens, Vealey, & Burton, 1990) was the primary reference used in the identification of research that utilized SCAT. In addition to this publication, the Social Science Citation Index was used to find additional studies or experiments that implemented SCAT. A computer search was deemed unnecessary because the afore mentioned publications provided a complete listing of research implementing SCAT without any disparities. The search identified 89 studies that used SCAT. Also, the Dissertation Abstracts International (Sciences and Engineering) was used to identify doctoral dissertations that implemented SCAT. This search identified 35 studies/experiments. This manual search 50 51 resulted in a total of 124 studies that used SCAT. One hundred and twelve of these studies were excluded, however, because they supplied insufficient data or because the dependent variabLe was not an individuals level of performance in a given sport but rather state anxiety, intrapersonal and.social factors, gender, stress, imagery, and a host of other variables. Additionally, several authors were contacted by phone in an attempt to acquire data. At the conclusion of the data search 12 studies (yielding 14 correlations) were recognized as meeting all criteria for this synthesis. Coding of Study Features Data for this meta-analysis were selected from studies that specifically examined the relationship between trait anxiety, as measured by SCAT, and sport performance. From a review of the literature, several variables (study features) were identified as possibly moderating the anxiety/performance relationship. These variables were categorized under subject variables and task variables. A list of these study features and their corresponding classes used in this investigation are presented in Table 1. Subject Variables The three subject variables identified were gender, age, 52 Table 1 Fpatures of Studies -~-—-—~_-C-I-u--c—--’—--.---q-~-.e.—-‘b-—c.—‘e————~--t—--—~.—--_------~--- --—v-—--’—-———-‘--—-————-——-—---—-——‘—_—c—o—---——-————----———--- Subject Variables Gender Male Female Male and Female Age Up to 18 years of age 18 and over Ability Elite, Varsity Non-Elite, Non-Varsity Task Variables Sport Skills Open Skills Closed Skills 53 and ability. These variables were further categorized into several classes. The classes represented under gender were male and female subjects. A third class comprised males and females and was added for those studies that did not discriminate between gender. The age of subjects was categorized by those under and those over 18 years of age. This particular age was chosen, because most studies included in the present investigation used either high school or college age subjects and 18 years of age was determined to be the most logical cut-off age. The ability variable was further categorized into those studies using elite or varsity athletes, and studies using non-elite or non—varsity athletes. The subjects were assigned.to either class as indicated.by the original research article. Task Variables The task variables consisted of the type of sport skills necessary to perform the task. The type of skills were categorized into open versus closed skills (Sage, 1984). Sports that required consistency of movement, an unchanging environment, and self-paced action were coded as closed skill sports (i.e., billiard, golf, volleyball, marathon running, and throwing for accuracy). Those sports characterized by flexibility in movement response, varying environmental conditions, and externally paced action were coded as Open skill sports (i.e., tennis, softball, basketball, and rugby). 54 Treatment of the Data Each study included in this review used a Pearson correlation coefficient, or another statistic that could be converted to a Pearson correlation coefficient, as the common statistic between trait anxiety and sport performance. Data conversion was executed following various steps as outlined by Horton (1978). Then using Fisher's {p to p" transformation, the data were normalized. z scores were computed for as many relationships between trait anxiety and sport performance as provided by each study. Thus, a single study could provide several correlations. The correlations representing each trait anxiety/sport performance relationship were then analyzed to test for their consistency in regard to a single underlying population correlation. This test of homogeneity followed Hedges' (1981, 1982) model with the following formula: Em 1/(EL-3)[(ZL‘Z)(ZELZ)] where 1/33 - 3 represents the variance of each p; 25 is each individual correlation; g_represents the average correlation weighted by sample size; and the fig statistic indicates the sample correlation variabilityu g5 values were then compared to a chi-square distribution with 5 — 1 degrees of freedom, where g represents the number of correlations. The correlations are considered to be homogeneous when the 31., value is less than the chi-square critical value. When the 55 test of homogeneity is significant (i.e., more than the chi square critical value), the correlation representing the trait anxiety/sport performance relationship is not consistent and, thus, the average correlation statistic cannot be generalized across studies. In the event of a significant test of homogeneity, the correlations were grouped together by a particular study feature. Each study feature was then examined in terms of between—group differences and the homogeneity of the within group correlations. By this method, study features could be identified that displayed a modifying effect on the correlations. The relationship .113 = .‘LIT ’ Eu allowed for ANOVA-like comparisons; where H5 represents the total homogeneity' value across the correlation for each SCAT/performance relationship; 13, represents the total within- group homogeneity; pp represents the difference between the total value of the SCAT/performance relationshipiand.the total within study-feature group 3,, value. Hg was then compared to the chi-square table with the appropriate degrees of freedom (df = number of classes within a study feature, minus one). A nonsignificant within group and significant between group analysis indicated that the study feature had a modifying effect on the correlations. CHAPTER IV Results The purpose of this research synthesis was to investigate the relationship between competitive trait anxiety, as measured by SCAT, and sport performance. Several variables bypothesized.to:moderate the anxiety/performance relationship were examined through the use of meta-analytic procedures. The following questions guided this study: 1. What is the overall average correlation between competitive trait anxiety and performance as measured by SCAT? 2. Do any of the following sport—related features influence the relationship between trait anxiety and performance as measured by SCAT? a. Gender b. Ability c. Age d. Type of Sport (open vs closed skills) 3. Are the correlations for each relationship homogeneous? This chapter is divided into results of the overall test of homogeneity, results of tests of homogeneity for those variables (study features) hypothesized to have a moderating effect on the competitive trait anxiety/sport performance relationship, and results of various study feature 56 combinations that may also have a moderating effect upon that relationship. All statistical results are reported at the .05 level of significance. From. the 12 studies used in this meta-analysis, 14 correlations were Obtained. A summary of the characteristics for these 12 studies is presented.in Table 2. Included in this table are the number of subjects in each study, subject sex, subject age, sport, subject ability level, type of sport, and correlation coefficients. Tests of Homogeneity Overall Test of Homogeneity An overall test.of homogeneity (Hedges, 1982) was conducted using the 14 correlations that were obtained. The average correlation between anxiety and performance, calculated from these 14 correlations was —0.21. The overall homogeneity test :5 value, 77.41, was compared to the chi-square value of 22.36 with 5-1 = 13 degrees of freedom. The 31., value was greater than the chi—square value; therefore, the homogeneity test for the 14 correlations was significant. This implies that the correlations were not consistent across all studies, and the average ‘weighted. correlation. value of —0.21 cannot ‘be representative of them. Heterogeneity of these correlations was not surprising because the individual correlations ranged from.-0.79 to 0.25. A modifying influence from one or more of 57 Table 2 58 Summapy of Characteristics and Correlations for Competitive Trait Anxiety/Sport Performance Studies --—‘—--——---’---——----------—-—---—-------—-———————----’--——-— Study N Cooley 53 (1987) Cox 157 (1991) Gerson, 107 & Deshaies (1978) Hisanaga 24 (1982) Krane, 100 Williams, & Feltz (1991) Lanning, 24 & Hisanaga (1983) Maynard, 22 & Howe 13 (1987) McKelvie, 67 Valliant, & Asu (1985) Sonstroem, 30 & Bernardo (1982) Stadulis 156 (1977) 92 Weinberg 60 (1979) Weinberg, 63 & Genuchi, (1980) ”13 16—54 mean 29 >18 18-22 14-18 >18 14-18 19-24 19-24 >18 >18 >18 >18 >18 Tennis Volley- ball Soft- ball Volley- ball Golf Volley— ball Rugby Rugby Marathon running Basket- ball Nonelite Elite Elite Elite Nonelite Elite Elite Elite Nonelite Elite Billiard Nonelite Billiard Nonelite Throwing Nonelite for acc- uracy Golf Elite Open Closed Open Closed Closed Closed Open Open Closed Open Closed Closed Closed Closed .47 .78 .12 .79 -0 -O .14 .35 0.00 0.23 -0.22 0.11 0.25 59 the study features may explain the variation in correlations. The Homogeneity Test of Study Features To find an explanation for the heterogeneity of the 14 correlations within the anxiety/performance relationship, the effect of the following study features were considered: gender, ability, age, and type of sport. 5 values were calculated for each class within each study feature, and ANOVA-like comparisons were calculated for each study feature to determine whether between—class differences existed, or whether each class had a modifying effect on the overall Hm value of the anxiety/performance relationship. Gender. When categorized by subject gender, 7 of the 14 correlations were based on female subjects, 5 were based on male subjects, and 2 on both genders. H values were calculated for the male, female, and both gender classes. Additionally, 25 and fig'values were calculated for the gender feature. The ANOVA-like comparison analysis is presented in Table 3. The average correlation for males (p; -0.18) proved to be heterogeneous, as did.the average correlation for females (p: -0.23), and that for both genders (p; -0.18) indicating great variability of the correlations within each of the categories. The between-group analysis of males and females, omitting the class of both genders, provided a nonsignificant homogeneity test indicating that there was no difference between the average correlations for males and females. Therefore, gender 60 Table 3 Summapy of Analysis for Gender .----——-----------—-----——---~-—*—--—‘———-d--——-----—*————_—-- Source K df Homogeneity Average Confidence Test Correlation Interval Total 14 13 19 = 77.41* -0.21 -0.27 to -0.15 Between 2 1.13 = 0 . 53 Within 11 :5 = 76.88* Females 7 6 54.80* -0.23 -0.31 to -0.15 Males 5 4 17.04* -0.18 -0.29 to —0.07 Both 2 1 5.04* -0.18 —0.35 to 0.00 .——-——-_-——--_—--——-—-—-—-———---——--——--———————————--—-——-—-—-— 61 does not appear to have a modifying effect upon the competitive trait anxiety/sport performance relationship. Ability. When the 14 average correlations were grouped by level of ability, 8 correlations were provided by studies utilizing elite performances (p = -0.35) and 6 by studies utilizing nonelite performances (; = —0.08) . Homogeneity tests were performed for each class and Table 4 presents the ANOVA- like comparison analysis. The between-group analysis yielded a HB18 Male, 5 4 17.04* -.18 -.29 to —.07 >18 Both, 2 1 5.04* —.18 -.35 to .00 >18 * p < 05 82 Table 8 §ummagy of Analyses for Gender by Ability 83 Average Correlation Total Between Within Female, elite Male, elite Female, nonelite Male, nonelite Both, nonelite Homogeneity Test ET = 77.41 g3 = 20.81 kg = 56.60 37.83 1.72* 2.48* 9.53 5.04 Confidence Interval -.27 to -.15 —.44 to -.25 -.54 to —.19 -.15 to .13 —.22 to .04 -.35 to .00 84 Table 9 §ummary;of Analyses for Age by Ability Source K df Homogeneity Average Confidence Test Correlation Interval Total 14 13 £5 = 77.41* —.21 -.27 to -.15 Between 2 :5 = 40.76* Within ll :5 = 36.65* Elite, 2 1 .01 -.79 -.88 to -.64 <18 Elite, 6 5 17.37* —.28 -.37 to -.19 >18 Nonelite, 6 5 19.27* -.08 -.17 to .00 >18 85 Table 10 Summary of Analyses fgr Skill by Age Source K df Homogeneity Average Confidence Test Correlation Interval Total 14 13 :5 = 77.41* -.21 —.27 to -.15 Between 2 :5 = 40.73* Within 11 £5 = 36.68* <18, CS 2 l .01* -.79 —.88 to -.64 >18, OS 5 4 12.98* -.34 -.45 to -.21 >18, CS 7 6 23.69* -.11 —.19 to -.04 86 Table 11 Summagy of Analyses for Skill by Gender -—--——--——---.-—--—-—----———---——-—————-——————-——-—-—-———---—-— Source K df Homogeneity Average Confidence Test Correlation Interval Total 14 13 £5 = 77.41* - 21 -.27 to -.15 Between 5 H3 = 8 . 3 1 Within 8 fig = 69.10 Male, OS 2 l .33* -.21 -.52 to .14 Male, CS 3 2 16.67 -.18 -.29 to -.06 Female, 2 1 11.87 —.34 -.2 to .1 nonelite 5.04 —.18 -.35 to .00 APPENDIX C Studies not included in this meta-analysis Albrecht, R.R., & Feltz, D.L. (1987). Generality and specificity of attention related to competitive anxiety and sport performance. Journal of Sport Psychology, 2” 231-248. Anderson, MgB., & Williams, J.M. (1987). Gender role and sport competition anxiety: A re-examination. Research Quarterly for Exercise and Sport, pg, 52-56. Auvergne, S. (1983). Motivation and causal attribution for high and low achieving athletes. International Journal of Sport Psychology, 14, 85-91. Betts, E. (1982). Relation of locus of control to aspiration level and to competitive anxiety. Psychological Reports, pg, 71—76. Biddle, S.J.H., & Jandeson, K.I. (1988). Attribution dimensions: Conceptual clarification and.moderator variables. International Journal of Sport Psychology, 12, 47—59. Blachsmith, W.A. (1977). The effect of systematic desensitization on pre-match anxiety states among collegiate wrestlers (Doctoral dissertation, West Virginia University, 1977). Dissertation Abstracts International, 38, 1974A. Blais, M.R., & Vallerand, R.J. (1986). Multimodal effects of electomyographic biofeedback: Looking at children’s ability to control precompetitive anxiety. Journal of Sport Psychology, .8. 283-303. Broughton, R.S., & Perlstrom,.J.R. (1986). PK.experiments with a competitive computer game. Journal of Parapsyghology, pp, 193—211. Brustad, R.J. (1988). Affective outcomes in competitive youth sport: The influence of intrapersonal and socialization factors. Journal of Sport and Exercise Psychology, pp, 307— 321. ‘ 87 88 Brustad, R.J., & Weiss, M.R. (1987). Competence perceptions and sources of worry in high, medium, and low competitive trait—anxious young athletes. Journal_of Sport Psychology,.2, 97-105. Burhans, R.S., Richman, C.L., & Bergey, D.B. (1988). Mental imagery training: Effects on running speed performance. International Journal of Sport Psychology, 12, 26-37. Cheatham, T., & Rosentswieg, J. (1982). Validation of the Sport Competition Anxiety Test. Perceptual and Motor Skills, .§§. 1343-1346. Colley, A., Roberts, N., & Chipps, A. (1985). Sex-role identity, personality, and participation in team and individual sports by males and females. International Journal of Sport Psychology, 1p, 103-112. Crocker, P.R.E., Alderman, R.B., & Smith, F.M.R. (1988). Cognitive-affective stress management training' with. high performance jyouth. volleyball players: Effects on .affect, cognition, and performance. Journal of Sport and Exercise Psychology, _1_o, 448-460. Dothwaite, P.K., & Armstrong, M.R. (1984). An investigation into the anxiety levels of soccer players. International Journal of Sport Psychology, lg, 149-159. Durtschi, S.K., & Weiss, M.R. (1984). Psychological characteristics of elite and nonelite marathon runners. In D.M. Landers (Ed.), Sport and elite performers (pp. 73-80). Champaign, Il.: Human Kinetics. Ettenger, R.H. (1985). The effects of feedback on competitive anxiety anui motor performance (Doctoral dissertation, University of Southern California, 1985). Dissertation Abstracts International, pp, 352A. Feltz, ILIh, & Albrecht, RqR. (1986). Psychological implications of competitive running. In M.R. Weiss & D. Gould (Eds.), Sport for children and gypuths (pp. 225-230). Champaign, Il.: Human Kinetics. Fisher, A.C., & Zwart, E.F. (1982). Psychological analysis of athletes' anxiety responses. Journal of Sport Psychology, A” 139-158. Furst, D.M., & Tenebaum, G. (1984). The relationship between worry, emotionality, and sport performance, In D.M. Landers (Ed.), Sport and elite performers (pp. 89-96). Champaign, Il.: Human Kinetics. 89 Gill, D.L. (1988). Gender differences in competitive orientation and sport participation. International Journal of Sport Psychology, i2, 145-159. Gill, D.L., Dzewaltowski, D.A., & Deeter, T.E. (1988). The relationship of competitiveness and achievement orientation to participation in sport and non-sport activities. Journal of Sport and Exercise Psychology, 1Q, 139-150. Gill, D.L., & Martens, R. (1977). The role of task type and success-failure in group competition. International Journal of Sport Psychology, p, 160-177. Gould, D., Horn, T., & Spreeman, J. (1983). Competitive anxiety in junior elite wrestlers. Journal of Sport Psychology, p, 58-71. Gould, D., Horn, T., & Spreeman, J. (1983). Sources of stress in junior elite wrestlers. Journal of Sport Psychology, p, 159-171. Gould, D., Petlichoff, L., & Weinberg, R.S. (1984). Antecedents of, temporal changes in,(and:relationshipsibetween CSAI-Z subcomponents. Journal.of Sport Psychologyg p, 289-304. Gravelli, L., Searle, R., & St. Jean, P. (1982) Personality profiles of the Canadian women’s national volleyball team. Volleyball Technical Journal, 2, 13-17. Gray, J.J., & Haring, M.J. (1984) Mental rehearsal for sport performance: Exploring the relaxation-imagery paradigm. Journal of Spor; Behavior, 1, 68-78. Hamilton, S.A. (1986). Cognitive-behavioral training of basketball players: Professional and peer training (Doctoral dissertation, West Virginia University, 1985). Dissertation épstracts International, g1, 835A. Hanin, Y.L. (1982). Adaptation. of the Sport Competition Anxiety Test. Journal of Sport Psychology, gJ 308-309. Hanson, T.W., & Gould, D. (1988). Factors affecting the ability of coaches to estimate their athletes’ trait and.state anxiety levels. The Sport Psychologist, 2, 298-313. Hellstedt, J.C. (1987). Sport psychology at a ski academy: Teaching mental skills to young athletes. The Sport Psychologist, 1, 56-68. Hogg. J.M. (1980). Anxiety and. the! competitive swimmer. Canadian Journal of Applied Sport Sciences, §J 183—187. 9O Huband, E.D., &.McKelvie, J.S. (1986). Pre and.post game state anxiety in team athletes high and low in competitive trait anxiety. International Journal of Sport Psychology, 11, 191- 198. Jennings, L.M. (1989). The differences in the variables of attention, anxiety and mental imagery in golf and tennis players (Doctoral dissertation, Fordham university, 1988). Dissertation Abstracts Intepnational, pp, 28588. Karteroliotis, C., & Gill, D.L. (1987). Temporal changes in psychological and physiological components of state anxiety. Journal of Sport Psychology, 2, 261-274. Koyama, S., Inomata, K., & Takeda, T. (1982). Competitive anxiety of tennis players in varying situations. International Journal of Sport Psychology, 1;, 278-279. Krane, V., & Williams, J.M. (1987). Performance and somatic anxiety, cognitive anxiety, and confidence changes prior to competition. Journal of Sport Behavior, 1Q, 47-56. Krotee, M.L. (1980). The effects of various physical activity situational settings on the anxiety level of children, Journal of Sport Behavior, 3, 158-164. Lewthwaite, R. (1988). Threat perception in competitive trait anxiety: The endangerment of important goals. Manuscript submitted for publication. Magill, R.An, &.Ash, M.J. (1979). Academic, psycho-social, and motor characteristics of participants and non-participants in children’s sport. Researcthuarterl , 50, 230-240. Martens, R., Burton, D., Rivkin, F., & Simon, J. (1980). Reliability and validity of the Competitive State Anxiety Inventory (CSAI).In C.H. Nadeau, W.C. Halliwell, K.Mg Newell, & G.C. Roberts (Eds.), Psychology of motor behavior and sport - 1979 (pp. 91-99). Champaign, Il.: Human Kinetics. Martens, R., & Gill, D.L. (1976). State anxiety among successful and unsuccessful competitors who differ in competitive trait anxiety.; Research Quarterly, 41, 698-708. Martens, R., Rivkin, F., & Burton, D. (1980). Who predicts anxiety better: Coaches of athletes? In C.H. Nadeau, W.R. Halliwell, K.M. Newell, & G.CD. Roberts (Eds.), Psychology of motor behavior and sport - 1979 (pp. 84-90). Champaign, Il.: Human Kinetics. 91 ZMartens, R., & Simon, J.A. (1976). Comparison of three predictors of state anxiety in competitive situations. Research Quarterly. 41, 381-387. McClean, E. (1984). The relationship of attentional style and anxiety patterns to performance tendencies in intercollegiate basketball competition (Doctoral dissertation, University of Tennessee, 1984). Dissertation Abstracts International, 3;, 1103A. McKelvie, S.J., & Huband, D.E. (1980). Locus of control and anxiety in college athletes and non-athletes. Perceptual and Motor Skills, pp, 819-822. McSweeney, F.P. (1984). Hemispheric EEG lateralization and achievement motivation in average and elite male and female adolescent distance runners (Doctoral dissertation, Rutgers University, 1983). Dissertation Abstracts International, 4;, 359B. Miller, P.B., & Miller, A.J. (1985). Psychological correlates of success 1J1 elite sportswomen. International Journal of Sport Psychology, gg, 289-295. lMillhouse, J.A. (1981). Cognitive, control of competitive anxiety' (Doctoral. dissertation, ‘West. Virginia.‘University, 1980). Dissertation Abstracts International, 4;, 4640B. Ostrow, A.C., & Ziegler, S.G. (1978). Psychometric properties of the Sport Competition Anxiety Test. In B. Kerr (Ed.), Human pppformance and behavior (pp. 139-142). Calgary, AB: University of Calgary. Owie, I. (1981). Influence of sex-role standards in sport competition anxiety. International Journal of Sport Psychology, 1;, 289-292. Passer, M.w. (1983). Fear of failure, fear of evaluation, perceived competence, and self-esteem in competitive-trait- anxious children. Journal of Sport Psychology, 5. 172-188. Passer, M.w., & Seese, M.D. (1983). Life stress and athletic injury: Examination of positive versus negative events and three moderation variables. Journal.of Human Stress, g, 11—16. Petlichkoff, L.M. (1986, October). Psycholggical characteristics of successful and less successful elite sailors. Paper presented at the Annual Meeting of the Association for the Advancement of Applied Sport Psychology, Jekyll Island, GA. 92 Poteet, D., & Weinberg, R. (1980). Competition trait anxiety, state anxiety, and performance. Perceptual and Motor Skills, 59. 651-654. Power, S.L. (1982). An analysis of anxiety levels in track and field athletes of varying ages and abilities. International Journal of Sport Psychology, pp, 258-267. Rainey, D.W., Conklin, W.R., & Rainey, K.W. (1987). Competitive trait anxiety among male and female junior high school athletes. International Journal of Sport Psychology, 18. 171-180. Rainey, D.W., & Cunningham, H. (1988). Competitive trait anxiety in male and female college athletes. Research Quarterly for exercise and Sport, pg, 244-247. Riddick, C.C. (1984). Comparative psychological profiles of three groups of female: collegians: Competitive swimmers, recreational swimmers, and inactive swimmers. Journal.of Sport Behavior, 1, 160-174. Rinehart, R.E. (1981) . Relationship of anticipatory anxiety to swimming performance as measured by catecholamine levels and psychological inventory. Unpublished. master’s thesis, University of Arizona, Tucson. Robinson, T.T., & Carron, A.V. (1982). Personal and situational factors associated with dropping out versus maintaining participation in competitive sport. Journal of Sport Psychology, 3, 364-378. Rupnow, A., & Ludwig, D.A. (1981). Psychometric note on the reliability of the Sport Competition Anxiety Test: Form C. Research Quarterly for Exercise and Sport, pg, 35-37. Scanlan, T.K. (1977). The effects of competition trait anxiety and success-failure on the perception of threat in a competitive situation. Research Quarterly, pp, 144-153. Scanlan, T.K. (1978). Perceptions and responses of high- and low-competitive trait-anxious males to competition. Research Quarterly, pp, 520-527. Scanlan, T.K., & Lewthwaite, R. (1984). Social psychological aspects of competition for male youth Sport participants: I. Predictors.of competitive stress. Journal.of Sport Psycholpgy, é, 208-226. 93 Scanlan, T.K., & Lewthwaite, R. (1985). Social psychological aspects of competition for male youth sport participants: III. Determinants of personal performance expectancies. Journal of Spprt Psychology, 1, 389-399. Scanlan, T.K., & Lewthwaite, R. (1986). Social psychological aspects of competition for male youth sport participants: IV. Predictors of enjoyment. Journal of Sport Psychology, p, 25- 35. Scanlan, T.K., & Lewthwaite, R. (1988). From. stress to enjoyment: Parental and coach influences on young participants. In E.W; Brown & C.F. Branta (Eds.), Competitive sports for children and youth (pp. 41-48). Champaign, Il.: Human Kinetics. Scanlan, T.K., Lewthwaite, R., & Jackson, B.L. (1984). Social psychological aspects of competition for male youth sport participants: II. Predictors of performance outcomes. Journal of Sport Psychology, p, 422-429. Scanlan, T.K., & Passer, M.W. (1977). The effects of competition trait anxiety and game win-loss on perceived threat in a natural competitive setting. In D.M. Landers & R.w. Christina (Eds.), Psychology of motor behavior and sport - 1976 (pp. 157-160). Champaign, IL: Human Kinetics. Scanlan, T.K., & Passer, M.W. (1978). Factors related to competitive stress among male ‘youth sport participants. Medicine and Science in Sport, iQ, 103-108. Scanlan, T.K., & Passer, M.W. (1979). Factors influencing the competitive performance expectancies of young female athletes. Journal of Sport Psychploqy, 1, 212-220. Scanlan, T.K., & Passer, M.W. (1979). Sources of competitive stress in young female athletes. Journal of Sport Psychology, .1, 151-159. Scanlan, T.K., & Passer, M.W. (1981). Determinants of competitive performance expectancies of young male athletes. Journal of Personality, 32, 60-74. Scanlan, T.K., & Ragan, J.T. (1978). Achievement motivation and. competition: Perceptions and. responses. 'Medicine .and Science in Sports, 1Q, 276-281. Schloder, M.E. (1988). Effects of teaching styles on performance, anxiety, and attitude toward instruction of college novice swimmers (Doctoral dissertation, Arizona State University, 1987). Dissertation Abstracts International, 52, 458A. 94 Seabourne, T.G. (1984). The effect of individualized, non- individualized, and package cognitive intervention strategies on karate performance (Doctoral dissertation, North Texas State University, 1983) . Dissertation Abstracts International, 44, 3332A. Segal, J.D., & Weinberg, R.S. (1984). Sex, sex role orientation, and competitive trait anxiety. Journal of_§port Behavior, 1, 153-159. Simon, J.A., & Martens, R. (1977). SCAT as a predictor of A- stated in varying competitive situations. In D.M. Landers & R.W. Christina (Eds.), Psychology of motor behavior and sport - 1976 Vol. 2, pp. 146—156). Champaign, Il.: Human Kinetics. Smith, R.E. (1980). A cognitive-affective approach to stress management training for athletes. In C.H. Nadeau, W.R. Halliwell, K.M. Newell, & G.C. Roberts (Eds.), Psycholpgyyof motor behavior and sport - 1979 (pp. 54-72). Champaign, Il.: Human Kinetics. Smith, T. (1983). Competition trait anxiety in youth sport: Differences according to age, sex, race and playing status. Perceptual and Motor Skills, p1, 1235-1238. Stadulis, R.E. (1977). Need. achievement, competition preference and evaluation seeking. In D.M. Landers & R.W. Christina (Eds.), Psychology of motor behavior and sport - 1976 (pp. 113-122). Champaign, IL: Human Kinetics. Thirer, J., Knowlton, R., Sawka, M., & Chang, T.J. (1978). Relationship of psycho—physiological characteristics to perceived. exertion and levels of anxiety in competitive swimmers. Journal of Sport Behavior, 1, 169-173. Thirer, J., & O’Donnell. L.A. (1980). Female intercollegiate athletes' trait-anxiety level and performance in a game. Perceptual and Motor Skills, pg, 18. Vealey, R.S. (1986) . Conceptualization of sport-confidence and competitive cmientation: Preliminary investigation and instrument development. Journal of Sport Psychology, p, 221- 246. Wagner, D.B. (1984). Sports anxiety reduction techniques to improve golf performance (Doctoral dissertation, Hofstra University, 1983). Dissertation Abstracts International, 35, 1275B. Wandzilak, T., Potter, G., & Lorentzen, D. (1982). Factors related to predictability of pre-game state anxiety. International Journal of Sport Psychology, is, 31-42. 95 Wark, K.A., & Wittig, A.F. (1979). Sex role and sport competition anxiety. Journal_of Sport Psychology, ;, 248-250. ‘Watson, G.G. (1986). Approach-avoidance Ibehavior' in team sports: An application to leading Australian national hockey players. International Journal of_§port Psychology, 11, 136- 155. Weinberg, R.S. (1978). The effects of success and failure on the patterning of neuromuscular energy. Journal of Motor Behavior, 19, 53-61. Willis, J.D. (1982). Three scales to measure competition- related motives in sport. Journal of Sport PsypholOgXJ.é, 338- 353. Wittig, A.F. (1984). Sport competition anxiety and sex role. Sex Roles, 1Q, 469-472. Wittig, A.F., Duncan, S.L., & Schurr, K.T. (1987). The relationship of gender, gender—role endorsement and.perceived physical self—efficacy to sport competition anxiety. Journal of Sport Behavior, 11, 192-199. Wolfe, J.E. (1980). Anthropometrical, physiological and psychological measures to predict performance in cross-country skiing (Doctoral dissertation, University of Northern Colorado, 1979). Efissertation Abstracts International, 4Q, 4478A. LI ST OF REFERENCES REFERENCES Allen, M.J., & Yen, W;M. (1979). Introduction to measurement theopy. Monterey, CA: Brooks/Cole. Berlyne, D.E. (1960). Conflict, arousal. and curiosity. New York: McGaw-Hill. Bialer, I. (1961). Conceptualization.of success and failure in mentally retarded and normal children. Journal of Personality, 22, 303-320. Bowers, K.S. (1973). Situationism.in psychology: An analysis and a critique. Psycholggical Review, 8Q, 307-336. Broughton, R.S., & Perlstrom,.J.R. (1986). PK experiments with a competitive computer game. Journal of Parapsychology, 52, 193-211. Brustad, R.J. (1988) . Affective outcomes in competitive youth: The influence of intrapersonal and socialization factors. Journal of Sport and Exercise Psychology,.lfl: 307—321. Brustad, R.J., & Weiss, M.R. (1987). Competence perceptions and sources of worry in high, medium, and low competitive trait-anxious young athletes. Journal of Sport Psychology,.9, 97-105. Burhans, R.S., Richman, C.L., & Bergey, D.B. (1988). Mental imagery training: Effects on running speed. performance. International Journal of Sport Psychology, lg, 26-37. Burton, D. (1988). Do anxious swimmers swim slower? Reexamining the elusive anxiety-performance relationship. Journal of Sport and Exercise Psychology, lg, 45-61. Carron, A.Vg (1980). Socialypsycholpgy’of sport. Ithaca, N.Y.: Mouvement. Carson, R.C. (1969). Interaction concepts ofJersonality. Chicago: Aldine. Castaneda, A. (1956). Effects of stress on complex learning and performance. Journal of Experimental Psychology, _5_2_, 9-12. 96 97 Cattell, R.B., & Cattell, M.D. (1969). Junior-senior high school personality gpestionnaire. Champaign, Il.: Institute for Personality and Ability Testing. Colley, A., Roberts, N., & Chipps, A. (1985). Sex-role identity, personality, and participation in team and individual sports by males and females. International Journal of Sport Psychology, 1p, 103-112. Cooley, E.J. (1987). Situational and trait determinants of competitive state anxiety. Perceptual and Motor Skills, pp, 767-773. Cooper, H.M. (1984). The integration review. Beverly Hills, CA: Sage Publication. Cox, R.H. (1991). Relationship between performance in women's volleyball and selected psychological measures. In L. VanderVelden & J.H. Humphrey (Eds.), Psychology and sociology of sport: Current and selected research. New York: AMS Press. Crews, D.J., & Landers, D.M. (1987). A.meta-analytic review'of aerobic fitness and reactivity to psychosocial stressors. Medicine and Science in Sports and Exercise, 12, 5114-5120. Crocker, P., Alderman, R., & Smith, F. (1988). Cognitive- affective stress management training with high performance youth volleyball players: Effects on affect, cognition, and performance. Journal 0: Sport and Exercise Psychology, 1Q, 448-460. Duffy, E. (1962). Activation and behavior. John Wiley & Sons, Inc., N.Y. Durtschi, S.K., & Weiss, M.R. (1984). Psychological characteristics of elite and nonelite marathon runners. In D.M. Landers (Ed.), Sport and elite performers (pp. 73-80). Champaign, IL: Human Kinetics. Eysenck, H.J. (1952). The scientific study of personality. London: Routledge & Kegan Paul. Feltz, D.L. , & Albrecht, R.R. (1986) . Psychological implications of competitive running. In M.R. Weiss & D. Gould (Eds.), Sport for children and youths (pp. 225-230). Champaign, IL: Human Kinetics. Feltz, D.L., & Landers, D.M. (1983). The effects of mental practice on motor skill learning and performance: A meta- analysis. Journal of Sport Psychology, p, 25-57. 98 Fenz, W., & Jones, G. (1972). Individual differences in psychologic arousal and performance in sport parachutists. Psychosomatic Medicine, 12, 1-8. Gerson, R., & Deshaies, P. (1978). Competitive trait anxiety and performance as predictors of precompetitive state anxiety. International Journal of Sport Psychology, p, 16-26. Ghiselli, E.E. (1949). The validity of occupational aptitude tests. New York: John Wiley. Gill, D.L. (1988). Gender differences in competitive orientation and sport participation. International Journal of Sport Psychology, 12, 145—159. Gill, D.L., Dzewatowski, D.A., & Deeter, T.E. (1988) . The relationship of competitiveness and achievement orientation to participation in sport and non-sport activities. Journal 0; Sport and Exercise Psychology, 1Q, 139-150. Gill, D.L., & Martens, R. (1977). The role of task type and success-failure in group competition. International Journal of Sport Psychology, s, 160-177. Glass, G.V. (1976). Primary secondary, and meta-analysis of research. Educational Researcher, p, 3-8. Glass, G.V. (1977). Integrating findings: Thenmeta—analysis of research. Review of Research in Education, p, 351-379. Glass, G.V. (1983). Synthesizing empirical research: Meta- analysis. In S.A, Ward. and L.J. Reed (Eds.), Knowledge structure and use: Implications for synthesis and interpretation. Philadelphia: Temple University Press. Glass, G.V., McGaw, B., & Smith, M. (1981). Meta-analysis in Social Research. Beverly Hills, CA: Sage Publications. Gould, D., Horn, 'T., 8: Spreeman, J. (1983). Competitive anxiety in junior elite wrestlers. Journal of Sport Psychology, p, 58-71. Gravelle, L., Searle, R., & St. Jean, P. (1982). Personality profiles of the Canadian women’s national volleyball team. Volleyball Technical Journal, 1, 13-17. Green, B., & Hall, J. (1984). Quantitative methods for literature review. Annual Review of Psychology, 1;, 37-53. Hedges, L.V. (1981). Distribution.theory forlslass’s estimator of effect size and related estimators. Journal of Educational Statistics, p, 107-128. 99 Hedges, L.V. (1982). Fitting categorical models to effect sizes from a series of experiments. Journal of Educational Statistics, 1, 119-137. Hedges, L.VQ, & Olkin, I. (1983). Regression. models in research synthesis. American Statistician..;l: 137-140. Hedges, L.V., & Olkin, I. (1985). Statistical methods for meta-analysis. New York: Academic Press. Hisanaga, B. (1982). Coaches, relaxation trainingI and sport performance: A field study. Unpublished Doctoral Dissertation, University of Indiana. Hogg, J.Mg (1980). Anxiety’ and. the competitive swimmer. Canadian Journal of Applied Sport Sciences, p, 183-187. Horton, R.L. (1978). The general linear model: Data analysis in the social and behavioral sciences. London, McGaw Hill. Hubard, E.D., & McKelvie, J.S. (1986). Pre and.post game state anxiety in team athletes high and low in competitive trait anxiety. International Journal of Sport Psychology, 11, 191- 198. Hull, C.L. (1943). Principles of behavior; New York: Appleton- Century-Crafts. Hunter, J.E., Schmidt, F.L., & Jackson, G.B. (1982). Meta- analysis: Cumulating'research.findings.across studies. Beverly Hills, CA: Sage Publications. Iso-Ahola, S.E., & Hatfield, B. (1986). Psychology of sports. Dubuque, Iowa: Wm.C. Brown Publishers. Kanteroliatis, (LU, & Gill, D. (1987). Temporal changes in psychological and physiological components of state anxiety. Journal of Sport Psycholggy, 3, 261-274. Kauss, D.R. (1980). Peak performance. N.J.: Prentice-Hall, Inc. Krain, V., Williams, J. & Feltz, D. (in press). Path analysis examining relationships among cognitive anxiety, state confidence, self-efficacy, and golf performance. Journal of Sport Behavior. Krotee, M.L. (1980). The effects of various physical activity situational settings on the anxiety level of children. Journal of Sport Behavior, 1, 158-164. 100 Lanning, W., & Hisanaga, B. (1983). A study of the relation between the reduction of competition anxiety and an increase in athletic performance. International Journal of Sport Psychology, 11, 219-227. Landers, D. (1980). The arousal-performance relationship revisited. Research Quarterly, 21, 77-90. Levine, J.Mg, Kramer, G.G., & Levine, E.N. (1975). Effects of alcohol on human performance: An integration of research findings based on abilities classification. Journal of Applied Psyghology, pp, 285-293. Levine, J.M., Romashko, T., & Fleishman, E.A. (1973). Evaluation of an abilities classification system for integration and generalizing human performance research findings: An. application to 'vigilance tasks. Journal of Applied Psychology, 22, 149-157. Levy, NQA. (1958). A short form of the children’s manifest anxiety scale. Child Development, 22, 153-154. Light, R.J., & Pillemer, D.B. (1984). Summing up: The science of reviewing research, Cambridge, ZMA; Harvard. University Press. Lirgg. C. (1991). Gender differences in self-confidence in physical activity: A.meta-analysis of recent studies. Journal of Exercise and Sport Psychology, 12, 294-310. Magnuson, D., & Endler, N.S. (1977). Personality at the crossroads: Current issues in interactional psychology. Hillsdale, N.J.: Erlbaum. Martens, R. (1974). Arousal and motor performance. In J.H. Wilmore (Ed.) Exercise and sport sciences review, (V01 2. pp. 155-180). New York: Academic Press. Martens, R. (1977). Sport competition.anxiety;test. Champaign, Il.: Human Kinetics. Martens, R., Burton, D., Rivkin, F., & Simon, J. (1980). Reliability and validity of the Competitive State Anxiety Inventory (CSAI). In.C.H. Nadeau, W.C. Halliwell, K.MkiNewell, & G.C. Roberts (Eds.) , Psychology of motor behavior and sport- 1979 (pp. 91-99). Champaign, Il.: Human Kinetics. Martens, R., Burton, D., Vealey, R.S., Bump, L.A., & Smith, D.E. (1982). Competitive state anxiety inventopy-Z. Symposium conducted at the meeting of the North American Society for the Psychology of Sport and Physical Activity (NASPSPA), College Park, MD. 101 Martens, R,, &.Gill, D. (1976). State anxiety among successful and unsuccessful competitors who differ in competitive trait anxiety. Research Quarterly, 11, 698-708. Martens, R. & Simon, J.A. (1976). Comparison of three predictors of state anxiety in competitive situations. Research Quarterly, 11, 381-387. Martens, R., Vealey, R.S., & Burton, D. (1990). Competitive anxiety in sport. Champaign, Il.: Human Kinetics Books. Maynard, I.W., & Howe, B.L. (1987). Interrelations of trait and state anxiety with game performance of rugby Players. Perceptual and Motor Skills, pp, 599-602. McKelvie, S.J., Valliant, P.M., & Asu, M.E. (1985). Physical training and personality factors as predictors of marathon time and training injury. Perceptual and Motor Skills, pp, 551-566. Mehrabian, A. (1968). Male and female scales of the tendency to achieve. Educational and Psychological Measurement, 22, 493-502. Mischel, VL (1968). PersonaliUL and assessment. New Yerk: Wiley. Mischel, W. (1973). Toward a cognitive social learning reconceptualization.of personality. Psyphological Review,.22, 252-283. Miller, B.P., & Miller, A.J. (1985). Psychological correlates of success 1J1 elite sportswomen. International Journal of Sport Psychology, 1p, 289-295. Murphy, S.M., & Woolfolk, R.L. (1987). The effects of cognitive interventions on competitive anxiety and performance on a fine motor skill accuracy task. International Journal of Sport Psychology, 12, 152-166. Neiss, R. (1988). Reconceptualizing arousal: Psychobiological states in motor performance. Psychological Bulletin, 103, 345- 366. Nideffer, R.M. (1976). The inner athlete: Mind plus.muscle for winning. N.Y.: Thomas Y. Crowell Co. Nideffer, R.M. (1980). Athletes guide to mental training. Champaign, Il.: Human Kinetics Publishers. 102 Oxendine, J.B. (1970). Emotional arousal and. :motor performance. Ques 1.1;: 23-32. Passer, M.W. (1983). Fear of failure, fear of evaluation, perceived competence, and self-esteem in competitive-trait- anxious children. Journal of Sport Psychology, é, 172-188. Passer, M.W. (1984). Competitive trait anxiety in children.and adolescents. In J.M. Silva & R.S. Weinberg (Eds.), Psychological foundations of sport (pp. 130-144). Champaign, Il.: Human Kinetics. Poteet, D., & Weinberg, R. (1980). Competition trait anxiety, state anxiety, and performance. Perceptual and Motor Skills, .52. 651-654. Power, S.L. (1982). An analysis of anxiety levels in track and field athletes of varying ages and abilities. International Journal of Sport Psychology, 12, 258-267. Rainey, D.W., Conkling, W.E., & Rainey, K.W. (1987). Competitive trait anxiety among male and female junior high school athletes. International Journal of Sport Psycholggy, .23.. 171-180. Rainey, D.W., & Cunningham, H. (1988). Competitive trait anxiety in male and female college athletes. Research Quarterly for Exercise and Sport, 22, 244-247. Rushall, B.S. (1979). Psychology of sport. Pellham Books, London. Sage, G.H. (1984). Motor learning and control. Dubuque, Iowa: Wm. C. Brown Publishers. Sarason, S.B., Davidson, K.S., Lighthall, F.F., Waite, R.R., & Ruebush, B.K. (1960). Anxiety in elementary school children. New York: Wiley. Scanlan, T. (1975). The effect of coppetition trait anxiety and success-failure on the spergeption of threat in a competitive situation. Unpublished. Doctoral Dissertation, University of Illinois. Scanlan, T. (1978). Perceptions and responses of high- and low-competitive trait-anxious males to competition. Research Quarterly, 12, 520-527. Scanlan, It, (i Lewthwaite, R. (1984). Social psychological aspects of competition for male youth sport participants: I. Predictors of competitive stress. Journal.of Sport Psychology, 6, 208-226. 103 Scanlan, T., & Passer, M.W. (1977). The effects of competition trait anxiety and game win-loss on perceived threat in a natural competitive setting. In D.M. Landers & R.W. Christina (Eds.), Psychology of motor behavior and sport - 1976 (pp. 157-160). Champaign, Il.: Human Kinetics. Scanlan, T., & Passer, M.W. (1978). Factors related to competitive stress among' male ‘youth sport participants. Medicine and Science in Sport, 12, 103-108. Scanlan, T., & Passer, M.W. (1979a). Factors influencing the competitive performance expectancies of young female athletes. Journal of Sport Psychology, 1, 212-220. Scanlan, T., & Passer, M.W. (1979b). Determinants of competitive performance expectancies of young male athletes. Journal of Personality, 12, 60-74. Scanlan, TL, (& Passer, M.W. (1981). Determinants of competitive performance expectancies of young male athletes. Journal of Personality, 12, 60-74. Scanlan, T., & Ragan, J.T. (1978). Achievement motivation and competition: Perceptions and responses. Medicine and Science in Sports, 12, 276-281. Silva, J .M. , & Weinberg, R. S . (1984) . Psychological Foundations of Sport. Champaign, Il.: Human Kinetics. Simon, J.A., & Martens, R. (1977). SCAT as a predictor of A- states in varying competitive situational In D.M. Landers & R.W. Christina (Eds.), Psychology of motor behavior and sport - 1976 (Vol. 2, pp. 146-156). Champaign, Il.: Human Kinetics. Singer, R.N. (1975). Motor learning and human performance. MacMillan Publishing Co., N.Y. Smith, C.P. (1969). The origin and expression of achievement- related motives in children, In.C.P. Smith (Ed.), Achievement- related motives in children (pp. 102-150). New York: Russell Sage Foundation. Smith, T. (1983). Competition trait anxiety in youth sport: Differences according to age, sex, race and playing status. Perceptual and Motor Skills, 21, 1235-1238. Sonstroem, R.J., & Bernardo, P. (1982). Intraindividual pregame state anxiety and basketball performance: A re- examination of the inverted-U curve. Journal of Sport Psychology, 1, 235-245. 104 Sonstroem, R.J. (1984). A review of anxiety in Sport. In J.M. Silva & R.S. Weinberg (Eds.), Psychological foundations of sport (pp. 104-117). Champaign, Il.: Human Kinetics. Sparling, P.B. (1980). Meta-analysis of studies comparing maximal oxygen uptakes in men and women. Medicine and Science in Sports, p1, 542-552. Spence, K.W., Farber, I., & McFann, H. (1956). The relation of anxiety (drive) level to performance in competitional and non- competitional paired-associates learning. Journal of Eyperimpntal Psychology, 22, 296-305. Spence, J.T., & Spence, K.W. (1966). The motivational components of manifest anxiety: Drive and drive stimuli. In C.D. Spielberger (Ed.), Anxiety; and. behavior; New' York: Academic. Spielberger, C.D. (1966). Theory and research on anxiety. In C.D. Spielberger (Ed.), Anxiety and behavior (pp. 3-20). New York: Academic Press. Spielberger, C.D. (1973). Preliminary test manual forfithp State-Trait Anxiety Inventory for Children. Palo Alto, CA: Consulting Psychologists. Spielberger, C.D., Gorsuch, R.L., & Lushene, R.L. (1970). Manual for the State-Trait Anxiety Inventopy. Palo Alto, CA: Consulting Psychologists. Stadulis, R.E. (1977). Need. achievement, competition preference and evaluation seeking. In D.M. Landers & R.W. Christina (Eds.), Psychology of nptor behavior and sport - 1976 (pp. 113—122). Champaign, IL: Human Kinetics. Taylor, J.A. (1953). A personality scale of manifest anxiety. Journa1 of Abnormal and Social Psyphology, 12, 285-290. Thomas, J.R,, & French, K.E. (1985). Gender differences across age in motor performance: A meta-analysis. Psychological Bulletin, 22, 260- 282. Thomas, J.R., & French, K.E. (1986). The use of meta-analysis in exercise and sport: A tutorial. Research Quarterlg/ for Exercise and Sport, 21, 196-204. Thorndike, R.L. (1933). The effect of the interval between test and retest on the constancy of the I.Q. Journal of Educational Psychology, 22, 543-549. 105 Vale, J.W., & Vale, C.A. (1969). Individual differences and general laws in psychology: A reconciliation. American Psychologists, 21, 1093-1108. Vealey, R.S. (1986) . Conceptualization of sport-confidence and competitive orientation: Preliminary investigation and instrument development. Journal of sport Psychology, 2, 221- 246. Wandzilak, T., Potter, G., & Lorenzen, D. (1982). Factors related to predictability of pre-game state anxiety. International Journal of Sport Psychology, 12, 31-42. Watson, G.G. (1986). Approach-avoidance behavior in team sports: An application to leading Australian national hockey players. International Journal of Sport Psycholcgy, 12, 136- 155. Watson, D., & Friend, R. (1969). Measurement of social- evaluative anxiety. Journal of Consulting. and Clinical Psychology, Q: 448-457. Weinberg, R.S. (1977). Anxiety and motor behavior: A new direction. In R.W. Christina & D.M. Landers (Eds.), Psychology of motor behavior and sport - 1976 (pp. 132-139). Champaign, Il.: Human Kinetics. Weinberg, R.S. (1978). The effects of success and failure on the patterning of neuromuscular energy. Journal of Motor Behavior, 12, 53-61. Weinberg, R.S. (1979). Anxiety and motor performance: Drive theory vs. cognitive theory. International Journal of Sport Psychology, _1_C_)_, 112-121 . Weinberg, R.S., & Genuchi, M. (1980). Relationship between competitive trait anxiety, state anxiety, and golf performance: A field study. Journal of Sport Psychology, 2, 148-154. Willis, J.D. (1982). Three scales to measure competition- related motives in sport. Journal of Sport Psychology, 1, 338- 353. Wolf, F.M. (1990). Meta-analysis. Quantitative methods for research synthesis. Newbury Park, CA.: Sage Publication. Zung, V.T., Weltman, A., Glass, G.V., & Mood, D.P. (1983) . The effects of exercise on blood lipids and lipoproteins: A meta- analysis of studies. Medicine and Science in Sports ang Exercise, 12, 393-402. 106 Zuckerman, M. (1979). Sensation seeking: Beyond the optimal level of arousal. Lawrence Erlbaum, Hillsdale, N.J. 1|'WINHWIHHNIWIIWWI!”HIIIJWIH ll